Suggestion Next Release

[Diego Escobar]

Hi:

I have a suggestion about the default software for Fedora:

First: In Latin America most part of people does´t have a fast Internet
conection, and for this reason is difficult to install new software by YUM.
Second: An important point that experts evaluate on a S.O. is "ready out of
the box" and you supose how many changes you have to do to leave Fedora
ready.

With this ideas my suggestion is to include many software by default on
Fedora (Especially in Fedora LiveCD)


   - OpenOffice
   - Linux Terminal Server (tsclient)
   - Vino
   - Yumex
   - Fedora Frog (Easy front-end)
   - Nautilus (always use browser by default)

And why not? we can do easy including by default the most common
repositories for Fedora:


   - Livna
   - Freshrpm

I hope my suggestion will be taken into account. Remember, this idea is
designed to do Fedora more competitive and more easy for the final user.

Regards.

Diego Escobar
Fedora Ambassador for Colombia
Fedora 8 Kernel Linux 2.6.24 + Gnome

[Ian Weller]

On Fri, 21 Mar 2008, Diego Escobar wrote:

>  And why not? we can do easy including by default the most common
>  repositories for Fedora:
>
>
>    - Livna
>    - Freshrpm
>  
Unfortunately in that case, legal issues take precedence over ease of
use...

[Diego Escobar]

Hi:

I have a suggestion about the default software for Fedora:

First: In Latin America most part of people does´t have a fast Internet
conection, and for this reason is difficult to install new software by YUM.
Second: An important point that experts evaluate on a S.O. is "ready out of
the box" and you supose how many changes you have to do to leave Fedora
ready.

With this ideas my suggestion is to include many software by default on
Fedora (Especially in Fedora LiveCD)


   - OpenOffice
   - Linux Terminal Server (tsclient)
   - Vino
   - Yumex
   - aMSN
   - Fedora Frog (Easy front-end)
   - Nautilus (always use browser by default)

And why not? we can do easy including by default the most common
repositories for Fedora:


   - Livna
   - Freshrpm

I hope my suggestion will be taken into account. Remember, this idea is
designed to do Fedora more competitive and more easy for the final user.

Regards.

Diego Escobar
Fedora Ambassador for Colombia
Fedora 8 Kernel Linux 2.6.24 + Gnome

[Igor Pires Soares]

Em Sex, 2008-03-21 às 23:10 -0500, Diego Escobar escreveu:
>   
>  Hi:
>   
>  I have a suggestion about the default software for Fedora:
>   
>  First: In Latin America most part of people does´t have a fast
>  Internet conection, and for this reason is difficult to install new
>  software by YUM. 
>  Second: An important point that experts evaluate on a S.O. is "ready
>  out of the box" and you supose how many changes you have to do to
>  leave Fedora ready.
>   
>  With this ideas my suggestion is to include many software by
>  default on Fedora (Especially in Fedora LiveCD)
>   
>        * OpenOffice 
>        * Linux Terminal Server (tsclient)  
>        * Vino 
>        * Yumex 
>        * aMSN  
>        * Fedora Frog (Easy front-end) 
>        * Nautilus (always use browser by default)
>  And why not? we can do easy including by default the most common
>  repositories for Fedora:
>   
>        * Livna 
>        * Freshrpm
>  I hope my suggestion will be taken into account. Remember, this idea
>  is designed to do Fedora more competitive and more easy for the final
>  user. 
Diego,

The Fedora Project Brazil is creating a Brazilian Portuguese LiveCD to
fit our needs. It includes OpenOffice and Nautilus browser mode by
default. Maybe you're interested in doing something like this for
Colombia.

Please, don't include non-free stuff or unofficial packages in your
local spin, if so, you cannot call it Fedora.

You can find the .ks files here:
http://igor.fedorapeople.org/spin-pt_BR/

[Jef Spaleta]

On Sat, Mar 22, 2008 at 5:38 AM, Igor Pires Soares <igorsoares(a)gmail.com&gt;
wrote:

>  The Fedora Project Brazil is creating a Brazilian Portuguese LiveCD to
>  fit our needs. It includes OpenOffice and Nautilus browser mode by
>  default. Maybe you're interested in doing something like this for
>  Colombia.
>  

Making a different set of package selections I can understand. But, I have
some concerns about making deliberate changes to default settings like
nautilus's default mode.  We shouldn't  be proliferating a bunch of user
interface customizations.  Why need to be consistent about what the default
settings look like in the Fedora branded spins. If a default setting is the
wrong default.. then our duty is to work with upstream to change it. We
shouldn't be propogating all possible choices for a default just because the
option is there.

There's no compelling reason why any localized spin should be making a
change in nautilus's default behavior which contradicts the setting in the
official Desktop spin. You can do it, but I'm going to have a hard time
supporting it.  If its better for your users in your area, why aren't you
making the case to get the default changed in upstream?

-jef

[Brian Pepple]

On Sat, 2008-03-22 at 11:40 -0800, Jeff Spaleta wrote:
>  On Sat, Mar 22, 2008 at 5:38 AM, Igor Pires Soares
>  <igorsoares(a)gmail.com&gt; wrote:
>          
>          The Fedora Project Brazil is creating a Brazilian Portuguese
>          LiveCD to
>          fit our needs. It includes OpenOffice and Nautilus browser
>          mode by
>          default. Maybe you're interested in doing something like this
>          for
>          Colombia.
>  
>  
>  Making a different set of package selections I can understand. But, I
>  have some concerns about making deliberate changes to default settings
>  like nautilus's default mode.  We shouldn't  be proliferating a bunch
>  of user interface customizations.  Why need to be consistent about
>  what the default settings look like in the Fedora branded spins. If a
>  default setting is the wrong default.. then our duty is to work with
>  upstream to change it. We shouldn't be propogating all possible
>  choices for a default just because the option is there.
>  
>  There's no compelling reason why any localized spin should be making a
>  change in nautilus's default behavior which contradicts the setting in
>  the official Desktop spin. You can do it, but I'm going to have a hard
>  time supporting it.  If its better for your users in your area, why
>  aren't you making the case to get the default changed in upstream? 
+1

Later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Jakub 'Livio' Rusinek]

>  If its better for your users in your area, why
>  aren't you making the case to get the default changed in upstream? 
I don't think they will be satisfied. They also may want to change this
not.

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Kevin Kofler]

Jeff Spaleta <jspaleta <at> gmail.com> writes:
>  If its better for your users in your area, why aren't you making
> the case to
>  get the default changed in upstream?-jef 
Because upstream GNOME has repeatedly refused to make that change, they were 
even reluctant to expose that option in the UI, now they did that, but that was 
as far as they wanted to go.

As for changing the default in Fedora, as the ones responsible for the Fedora 
GNOME packages are, or at least used to be, the same ones who made the upstream 
decision, good luck there too!

        Kevin Kofler

[Jef Spaleta]

On Sat, Mar 22, 2008 at 3:44 PM, Kevin Kofler <kevin.kofler(a)chello.at&gt;
wrote:

>  Because upstream GNOME has repeatedly refused to make that change,
> they
>  were
>  even reluctant to expose that option in the UI, now they did that, but
>  that was
>  as far as they wanted to go.
>
>  As for changing the default in Fedora, as the ones responsible for the
>  Fedora
>  GNOME packages are, or at least used to be, the same ones who made the
>  upstream
>  decision, good luck there too!
>  
Either you make a persuasive argument to change the default and get it
changed across the board, or we learn to live with it.  I'm not happy about
seeing Fedora branding spread 7000 images over all possible combinations of
default settings just because we have the tools to make it drop dead easy to
do.   Just because its easy to do, doesn't mean its appropriate.

I consider this sort of stuff as downstream patch aggregation....which is
something we are suppose to be avoiding.  And these sorts of changes are
completely orthogonal to issues of needed changes for localization
variants.  These sorts of changes are arbitrary personal preferences that
are being changed because people are in disagreement with upstream.  If we
want to have officially localized variants, then I think its reasonable to
expect the people looking to pass off their localized derivatives
distributions as Fedora images stay as close as possible to the
non-localized Desktop spin. Making arbitrary changes to defaults, just
because you can, dilutes the Fedora brand.  If people working on localized
spins don't show some reasonable restraint on applying arbitrary changes, we
aren't going to see officially localized spins...ever... because there will
be no commonality between default behavior across images and that just isn't
going to fly.

If you want to make arbitrary customizations to the defaults and release
that as a distribution, because you think you know better than upstream,
fine go right ahead. I'm not going to try to stop you.  You have the
infinite freedom to do what you want because we've made the tools possible
for you to do it.  But please don't ask it to be call Fedora.  Because
Fedora as a project needs to continue to work with upstream...even in
situations where individual Fedora contributors personally disagree about
what the default setting should be.  We need the Fedora images we release to
behave consistently.

-jef

[Diego Escobar]

2008/3/21, Diego Escobar <dezone24(a)gmail.com&gt;:
>  Hi:
>
>  I have a suggestion about the default software for Fedora:
>
>  First: In Latin America most part of people does´t have a fast Internet
>  conection, and for this reason is difficult to install new software by YUM.
>  Second: An important point that experts evaluate on a S.O. is "ready out
>  of the box" and you supose how many changes you have to do to leave Fedora
>  ready.
>
>  With this ideas my suggestion is to include many software by default on
>  Fedora (Especially in Fedora LiveCD)
>
>
>     - OpenOffice
>     - Linux Terminal Server (tsclient)
>     - Vino
>     - Yumex
>     - aMSN
>     - Fedora Frog (Easy front-end)
>     - Nautilus (always use browser by default)
>
>   And why not? we can do easy including by default the most common
>  repositories for Fedora:
>
>
>     - Livna
>     - Freshrpm
>
>  I hope my suggestion will be taken into account. Remember, this idea is
>  designed to do Fedora more competitive and more easy for the final user.
>
>  Regards.
>
>  Diego Escobar
>  Fedora Ambassador for Colombia
>  Fedora 8 Kernel Linux 2.6.24 + Gnome
>  

Hi.

First, thank you for your responses. I understand about legal issues
including some packages or repositories on Fedora. And I understand why
Livna is not included. (Because is not official as others parts). But for
most part of people is mandatory to have those packages. So, trying to do
easy for the final user can we do our official "community" repository? Can
we do our official "Fedora Frog" to simplify the installation for
common programas? (I believe there are no problem because most part of that
sofware can ba downloaded without problems by the user).

Also, I will response some things:


   1. Yum is working good... but if it could be faster...(Especially
   solving dependencies) nobody will unhappy.
   2. I m understand too, the space problems on LiveCD. But how can the
   other distros to include that software by default?
   3. Abirwod and Gnumeric are nice, but we are talking about do Fedora
   competitive and OpenOffice is the way. Users needs a suite as powerful as
   Microsoft Office.
   4. Talking about Yumex and Pirut, I think that Pirut needs a new face.
   (Check other distros). By other side, I dont know why, but Yumex works
   faster than Pirut and provides a face similar to Synaptic.

Thanks to Igor Pires for the idea of Spin. If the only way is create a spin,
I will so interested working on that for Colombia.

Please, consider this as a "suggestion" thinking in final users. They are no
interested in "official, unofficial" and they are no interested in read
"how
to enable repositories", They wants a S.O. ready and complete for the PC.

Thanks.

Regards.

-- 
Diego Escobar
Fedora Ambassador for Colombia
Fedora Release 8 (Warewolf): Linux Kernel 2.6.24 + Gnome 2.20

[Jakub 'Livio' Rusinek]

Dnia 22-03-2008, sob o godzinie 13:42 -0500, Diego Escobar pisze:

>  Hi.
>   
>  First, thank you for your responses. I understand about legal issues
>  including some packages or repositories on Fedora. And I understand
>  why Livna is not included. (Because is not official as others parts).
>  But for most part of people is mandatory to have those packages. So,
>  trying to do easy for the final user can we do our official
>  "community" repository? Can we do our official "Fedora Frog" to
>  simplify the installation for common programas? (I believe there are
>  no problem because most part of that sofware can ba downloaded without
>  problems by the user). 
You don't understand. We can't make Livna/RPM Fusion official, because
of legal issues. And stop here. It's IMPOSSIBLE.

Fedora protects us from any problems with patents by removing
nonfreeness from repos.

That's why Livna exists. It provides us the packages we want, but Fedora
simply can't include.

Unofficial packages can land in Fedora, but Fedora Frog allows mostly
doing things, which Fedora protects us from.

>  Also, I will response some things:
>   
>       1. Yum is working good... but if it could be faster...(Especially
>          solving dependencies) nobody will unhappy.  
It's true, so work upstream :P .

>       1. I m understand too, the space problems on LiveCD. But how
> can
>          the other distros to include that software by default? 
Because of other compromises.

>       1. Abirwod and Gnumeric are nice, but we are talking about do
>          Fedora competitive and OpenOffice is the way. Users needs a
>          suite as powerful as Microsoft Office.  
And you're sure that EVERYONE needs such functionality?

>       1. Talking about Yumex and Pirut, I think that Pirut needs a
> new
>          face. (Check other distros). By other side, I dont know why,
>          but Yumex works faster than Pirut and provides a face similar
>          to Synaptic.  
Pirut can have new face, but it will be replaced by PackageKit. Did you
read my e-mail?

Yumex is complete different from Synaptic. Don't know where you get such
"association" from.

>  Thanks to Igor Pires for the idea of Spin. If the only way is create
> a
>  spin, I will so interested working on that for Colombia.
>   
>  Please, consider this as a "suggestion" thinking in final users. They
>  are no interested in "official, unofficial" and they are no interested
>  in read "how to enable repositories", They wants a S.O. ready and
>  complete for the PC.  
Read Fedora's description if you still understand.

"Best free and open source software"

BTW, Happy Easter.

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Les Mikesell]

Jakub 'Livio' Rusinek wrote:
> > First, thank you for your responses. I understand about legal issues
> > including some packages or repositories on Fedora. And I understand
> > why Livna is not included. (Because is not official as others parts).
> > But for most part of people is mandatory to have those packages. So,
> > trying to do easy for the final user can we do our official
> > "community" repository? Can we do our official "Fedora Frog" to
> > simplify the installation for common programas? (I believe there are
> > no problem because most part of that sofware can ba downloaded without
> > problems by the user).
>  
>  You don't understand. We can't make Livna/RPM Fusion official, because
>  of legal issues. And stop here. It's IMPOSSIBLE. 
Some of it.  Things like the Nvidia drivers wouldn't be impossible.

>  Fedora protects us from any problems with patents by removing
>  nonfreeness from repos. 
I'd say you have that backwards.  By avoiding the issue, fedora leaves 
every end user on his own when he needs a patented function. That's the 
opposite of protection.

>  
> >      1. Abirwod and Gnumeric are nice, but we are talking about do
> >         Fedora competitive and OpenOffice is the way. Users needs a
> >         suite as powerful as Microsoft Office. 
>  
>  And you're sure that EVERYONE needs such functionality? 
And a fully functional java.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Rahul Sundaram]

Les Mikesell wrote:
>  Jakub 'Livio' Rusinek wrote: 
> >
> > You don't understand. We can't make Livna/RPM Fusion official, because
> > of legal issues. And stop here. It's IMPOSSIBLE.
>  
>  Some of it.  Things like the Nvidia drivers wouldn't be impossible. 
>  
>  And a fully functional java. 
http://fedoraproject.org/wiki/Objectives
http://fedoraproject.org/wiki/ForbiddenItems

Rahul

[Les Mikesell]

Rahul Sundaram wrote:
>  Les Mikesell wrote:
> > Jakub 'Livio' Rusinek wrote:
>  
> >>
> >> You don't understand. We can't make Livna/RPM Fusion official, because
> >> of legal issues. And stop here. It's IMPOSSIBLE.
> >
> > Some of it.  Things like the Nvidia drivers wouldn't be impossible.
>  
> >
> > And a fully functional java.
>  
>  http://fedoraproject.org/wiki/Objectives 
There are some very obvious conflicts on this page.

>  http://fedoraproject.org/wiki/ForbiddenItems 
Some strange innuendo about not liking Nvidia and ATI here, nothing 
definitively illegal. Sun Java, Vmware, and several of the others listed 
could be much easier to install without including anything illegal.


-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Rahul Sundaram]

Les Mikesell wrote:
>  Rahul Sundaram wrote:
> > Les Mikesell wrote:
> >> Jakub 'Livio' Rusinek wrote:
> >
> >>>
> >>> You don't understand. We can't make Livna/RPM Fusion official,
> because
> >>> of legal issues. And stop here. It's IMPOSSIBLE.
> >>
> >> Some of it.  Things like the Nvidia drivers wouldn't be impossible.
> >
> >>
> >> And a fully functional java.
> >
> > http://fedoraproject.org/wiki/Objectives
>  
>  There are some very obvious conflicts on this page. 
That comment is vague and useless.

> > http://fedoraproject.org/wiki/ForbiddenItems
>  
>  Some strange innuendo about not liking Nvidia and ATI here, nothing 
>  definitively illegal. Sun Java, Vmware, and several of the others listed 
>  could be much easier to install without including anything illegal 
If the legality is not clear, it is a risk including it and Fedora has 
chosen not to take the risk in this instance. Read the link more 
carefully. Just because it is legal doesn't meant Fedora will include 
it. There are several other considerations.

Rahul

[Les Mikesell]

Rahul Sundaram wrote:
>  Les Mikesell wrote:
> > Rahul Sundaram wrote:
> >> Les Mikesell wrote:
> >>> Jakub 'Livio' Rusinek wrote:
> >>
> >>>>
> >>>> You don't understand. We can't make Livna/RPM Fusion official,
> because
> >>>> of legal issues. And stop here. It's IMPOSSIBLE.
> >>>
> >>> Some of it.  Things like the Nvidia drivers wouldn't be impossible.
> >>
> >>>
> >>> And a fully functional java.
> >>
> >> http://fedoraproject.org/wiki/Objectives
> >
> > There are some very obvious conflicts on this page.
>  
>  That comment is vague and useless. 
As is a page that embraces usability and "just works" on the one hand 
and rejects the vendor-supplied drivers that would make it possible on 
the other.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Rahul Sundaram]

Les Mikesell wrote:
>  Rahul Sundaram wrote:
> > Les Mikesell wrote:
> >> Rahul Sundaram wrote:
> >>> Les Mikesell wrote:
> >>>> Jakub 'Livio' Rusinek wrote:
> >>>
> >>>>>
> >>>>> You don't understand. We can't make Livna/RPM Fusion
> official, 
> >>>>> because
> >>>>> of legal issues. And stop here. It's IMPOSSIBLE.
> >>>>
> >>>> Some of it.  Things like the Nvidia drivers wouldn't be impossible.
> >>>
> >>>>
> >>>> And a fully functional java.
> >>>
> >>> http://fedoraproject.org/wiki/Objectives
> >>
> >> There are some very obvious conflicts on this page.
> >
> > That comment is vague and useless.
>  
>  As is a page that embraces usability and "just works" on the one hand 
>  and rejects the vendor-supplied drivers that would make it possible on 
>  the other. 
Objectives are a list of priorities and balances. Besides this is out of 
context quote. Objectives encourage usability in default configuration. 
Default configuration doesn't include proprietary kernel drivers and 
those don't encourage usability nor do they just work anyway.

Rahul

[Jakub 'Livio' Rusinek]

>  I'd say you have that backwards.  By avoiding the issue, fedora
> leaves 
>  every end user on his own when he needs a patented function. That's the 
>  opposite of protection. 
Fedora doesn't care if this is harder or easier. I proposed here many
things, which could improve Fedora, but they were rejected.

Fedora doesn't care about end users. I don't know who it cares about...

That's the reason I'll switch to openSUSE quickly. Probably after
release of 11. Big improvement from 10.3. From the start of openSUSE,
it's getting better than Fedora and leaves the stereotypes in the past.

>  >>      1. Abirwod and Gnumeric are nice, but we are talking
> about do
>  >>         Fedora competitive and OpenOffice is the way. Users needs a
>  >>         suite as powerful as Microsoft Office. 
>  > 
>  > And you're sure that EVERYONE needs such functionality?
>  
>  And a fully functional java. 
I don't need both Java and OO.org.

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[drago01]

2008/3/22 Jakub 'Livio' Rusinek <jakub.rusinek(a)gmail.com&gt;:
>  > I'd say you have that backwards.  By avoiding the issue,
> fedora leaves
>   > every end user on his own when he needs a patented function. That's the
>   > opposite of protection.
>
>   Fedora doesn't care if this is harder or easier. I proposed here many
>   things, which could improve Fedora, but they were rejected.
>
>   Fedora doesn't care about end users. I don't know who it cares about... 
Can you please stop posting useless comments like this?
Just because you think that we don't care about end users it isn't
true (and your suggestions have been rejected for reasons, not because
"fedora does not care about end users")

[Jef Spaleta]

2008/3/22 Jakub 'Livio' Rusinek <jakub.rusinek(a)gmail.com&gt;:

>  That's the reason I'll switch to openSUSE quickly. Probably
> after
>  release of 11. Big improvement from 10.3. From the start of openSUSE,
>  it's getting better than Fedora and leaves the stereotypes in the past.
>  

I sincerely hope that the opensuse can find a way to make use of your
particular... talents, which you feel we've squandered.  I'll make sure to
keep an eye on the impact you make in that community.

-jef

[Peter Gordon]

On Sat, 2008-03-22 at 21:22 +0100, Jakub 'Livio' Rusinek wrote:
>  Fedora doesn't care if this is harder or easier. I proposed here
> many
>  things, which could improve Fedora, but they were rejected. 
And they all were rejected with reasons and explanations of why they are
not suitable for Fedora - be it legal issues or deviation from
upstreams, etc. These directly counter the goals of the Fedora Project
as a whole. 

Granted, they almost certainly would improve Fedora in the short-term
for many; but they would just as certainly harm us as a whole.

We want to make things easy and simple, yes. In fact, we LOVE doing
this. But doing so in a manner that is legally-dubious (proprietary
drivers or non-Free codecs, etc.) or pushes us _away_ from proper
community involvement (such as significant upstream deviations) is
simply not feasible for us.

>  That's the reason I'll switch to openSUSE quickly. Probably
> after
>  release of 11. Big improvement from 10.3. From the start of openSUSE,
>  it's getting better than Fedora and leaves the stereotypes in the past. 
I've heard good things about OpenSUSE, but from my trials with it in KVM
last year, it's quite a beastly thing at first. To each his own, I
suppose. =)
-- 
Peter Gordon (codergeek42)
GnuPG Public Key ID: 0xFFC19479 / Fingerprint:
  DD68 A414 56BD 6368 D957 9666 4268 CB7A FFC1 9479

[Kevin Kofler]

Les Mikesell <lesmikesell <at> gmail.com> writes:
>  Some of it.  Things like the Nvidia drivers wouldn't be
> impossible. 
They're illegally infringing on the kernel's GPLv2 license.

And even if you think they aren't, it would still be against Fedora's primary 
objectives to include proprietary drivers.

>  And a fully functional java. 
That's what OpenJDK/IcedTea is for.

        Kevin Kofler

[Rahul Sundaram]

Kevin Kofler wrote:
>  Les Mikesell <lesmikesell <at> gmail.com> writes:
> > Some of it.  Things like the Nvidia drivers wouldn't be impossible.
>  
>  They're illegally infringing on the kernel's GPLv2 license.
>  
>  And even if you think they aren't, it would still be against Fedora's primary 
>  objectives to include proprietary drivers. 
s/include/exclude or to be more specific only include free and open 
source drivers.

Rahul

[Kevin Kofler]

Rahul Sundaram <sundaram <at> fedoraproject.org> writes:

>  Kevin Kofler wrote:
>
>  > And even if you think they aren't, it would still be against Fedora's
>  > primary objectives to include proprietary drivers.
>  
>  s/include/exclude or to be more specific only include free and open 
>  source drivers. 
Oops, I guess my sentence was ambiguous. I meant:
And even if you think they aren't, it would still [be against Fedora's primary 
objectives] to include proprietary drivers.
not:
And even if you think they aren't, it would still be against [Fedora's primary 
objectives to include proprietary drivers].
which as you pointed out would be nonsense.

        Kevin Kofler

[Diego Escobar]

Hi...

I think is enough and a I will close this "suggestion" with a final comment.
I agree with Jeff, is not a good idea have a lot of Fedora
images customized, is not good for the community, but when something doesn´t
have what we want, simply change it.

The question is not about to do "arbitrary customizations" else listen what
the real users wants. Remember a developer works for the users and he can´t
ignore their requeriments. So, I think is a good idea to have a little space
where users can vote and decided what is the default setting or software on
Fedora (obviously inside the Fedora rules), and by this way most part of
people will be happy knowing that his vote counts. (Please, this is
different of ProposedFeatures).

Thanks.
-- 
Diego Escobar
Fedora Ambassador for Colombia
Fedora Release 8 (Warewolf): Linux Kernel 2.6.24 + Gnome 2.20

[Igor Pires Soares]

Em Sáb, 2008-03-22 às 22:08 -0500, Diego Escobar escreveu:
>  
>  Hi...
>   
>  I think is enough and a I will close this "suggestion" with a
>  final comment. I agree with Jeff, is not a good idea have a lot of
>  Fedora images customized, is not good for the community, but when
>  something doesn´t have what we want, simply change it.
>   
>  The question is not about to do "arbitrary customizations" else listen
>  what the real users wants. Remember a developer works for the users
>  and he can´t ignore their requeriments. So, I think is a good idea to
>  have a little space where users can vote and decided what is the
>  default setting or software on Fedora (obviously inside the Fedora
>  rules), and by this way most part of people will be happy knowing that
>  his vote counts. (Please, this is different of ProposedFeatures).
>   
>  Thanks. 
+1

I really understand and agree why we should not include proprietary and
patented stuff in our spins, but I don't think changing the nautilus
window behavior is bad and I think some people is making a big deal of
this.

I'am sure that we need to have a limit for customization, but in this
case I don't think we crossed that line. We started out from a .ks file
provided by Rahul Sundaram but we had to make some small changes. For
instance, we had to include a command to fix the installation icon,
because it wasn't appearing on the desktop. It was totally necessary.

We also need to have some space for customization. The translation
itself is a customization and the local teams give support for that.

If we need to drop the Nautilus browser mode to make the spin official
there is no problem by me, but our users will not like it.

[Jef Spaleta]

On Sat, Mar 22, 2008 at 7:43 PM, Igor Pires Soares <igorsoares(a)gmail.com&gt;
wrote:

>  If we need to drop the Nautilus browser mode to make the spin
> official
>  there is no problem by me, but our users will not like it.
>  

The question isn't will some users not like or not a specific customization
away from a default.  The question is, is it something specific to the scope
of your image concept.. or is it a generally applicable customization that
people have a disagreement with?  Yes the nautilus browser mode default that
some segment of the users think should be changed, but that desire isn't
locale specific, so it shouldn't be changed in a locale specific spin.  If
its going to be changed it needs to be changed across the board in how our
nautilus package behaves by default or not.

This goes for pretty much every potential customization. There's needs to be
a very good technical reason to override a default.. it needs to be 'broken'
and not just 'unpopular'.

-jef

[Andrew Farris]

Igor Pires Soares wrote:
>  I really understand and agree why we should not include proprietary
> and
>  patented stuff in our spins, but I don't think changing the nautilus
>  window behavior is bad and I think some people is making a big deal of
>  this. 
I'm curious whether you've had a real discussion/vote/request from a large group 
of potential users to do this, or whether it is the personal preference of the 
few who are planning to do the spin?  While its a small change to make, and you 
could say others are making a big deal out of it... it is only one checkbox 
easily accessible to every user in mere seconds after they login.  I don't see 
the point of anyone (on either side of the preference) making a big deal out of it.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Henrique Junior]

Hello *,

All this discussion shows important things, such as the fact that everyone
agrees with the issue of non-free content, but leads me to think on the
"infinite freedom" of our choices here.
I can't remember how long Nautilus has been distributed in spatial mode by
default, but by the users that I know, I believe that most of then prefer
the browser mode. In all these years of Fedora I (really) never seen an user
keeping the spatial mode and I want to ask to everyone here: who really the
uses the spatial mode?

By the way, if only changing the Nautilus mode may mean that the localized
spins can no longer be called Fedora, how far go our freedom to add a
wallpaper with the flag of our country under the theme Fedora, or choose a
collection of most popular icons such as Tango? Will this change the purpose
of a Fedora upstream too?


2008/3/23, Andrew Farris <lordmorgul(a)gmail.com&gt;:
>  Igor Pires Soares wrote:
>  > I really understand and agree why we should not include proprietary and
>  > patented stuff in our spins, but I don't think changing the nautilus
>  > window behavior is bad and I think some people is making a big deal of
>  > this.
>
>
>  I'm curious whether you've had a real discussion/vote/request from a large
>  group
>  of potential users to do this, or whether it is the personal preference of
>  the
>  few who are planning to do the spin?  While its a small change to make,
>  and you
>  could say others are making a big deal out of it... it is only one
>  checkbox
>  easily accessible to every user in mere seconds after they login.  I don't
>  see
>  the point of anyone (on either side of the preference) making a big deal
>  out of it.
>
>
>  --
>  Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
>    gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B
>  1DF3
>  No one now has, and no one will ever again get, the big picture. - Daniel
>  Geer
>  ----
>  ----
>
>
>  --
>  fedora-devel-list mailing list
>  fedora-devel-list(a)redhat.com
>  https://www.redhat.com/mailman/listinfo/fedora-devel-list
>  


-- 
Henrique "LonelySpooky" Junior
-------------------------------------------------------------
"In a world without walls and fences, who needs windows and gates?!"

[Andrew Farris]

Henrique Junior wrote:
>  Hello *,
>  
>  All this discussion shows important things, such as the fact that everyone
>  agrees with the issue of non-free content, but leads me to think on the
>  "infinite freedom" of our choices here.
>  I can't remember how long Nautilus has been distributed in spatial mode by
>  default, but by the users that I know, I believe that most of then prefer
>  the browser mode. In all these years of Fedora I (really) never seen an user
>  keeping the spatial mode and I want to ask to everyone here: who really the
>  uses the spatial mode? 
I for one never use browser mode, ever.  I am a long time Apple OS user though 
so I was accustomed to spatial browsing from the behavior of OS6 on.  Since 
Gnome introduced it again I've used nautilus that way.

As for the default, the chances of getting upstream to accept a return to 
browser mode is about nil.  As I remember it even getting an option in the 
general preferences to allow using browser mode was a major fight for people. 
When they first shifted to this (it was Gnome 2.0?), and the only way to get 
nautilus open in browser mode was using a CLI argument.

Personally I wouldn't care which behavior is default, it takes 2 seconds to 
change.  If the preference option went away, I'd care.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Jesse Keating]

On Sun, 2008-03-23 at 01:55 -0300, Henrique Junior wrote:
>  I want to ask to everyone here: who really the
>  uses the spatial mode? 
I rather like the spatial mode and I keep it that way.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Les Mikesell]

Jesse Keating wrote:
>  On Sun, 2008-03-23 at 01:55 -0300, Henrique Junior wrote:
> > I want to ask to everyone here: who really the
> > uses the spatial mode?
>  
>  I rather like the spatial mode and I keep it that way. 
I don't see how anyone with directories more than 2 levels deep or uses 
it for system file access can tolerate the useless trail of intermediate 
windows it leaves behind.  Have I missed something basic about how to 
get rid of them?

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Brian Pepple]

On Sun, 2008-03-23 at 10:13 -0500, Les Mikesell wrote:
>  
>  I don't see how anyone with directories more than 2 levels deep or uses 
>  it for system file access can tolerate the useless trail of intermediate 
>  windows it leaves behind.  Have I missed something basic about how to 
>  get rid of them? 
A mouse-middle-key double-click opens a directory, and closes it's
parent.

Later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Dimi Paun]

On Sun, 2008-03-23 at 13:09 -0400, Brian Pepple wrote:
>  A mouse-middle-key double-click opens a directory, and closes
> it's
>  parent. 
Perfectly mysterious, you can't expect anyone to know that, right?
We are talking about a default that would keep newbies comfortable,
not people in-the-know.

Not only is this rather obscure, but for a long time the middle button
has been a pain to access, as X seem to map it to any random button
when the mouse has more then a few buttons. 

And when it doesn't, typically the middle button is the wheel which is
hard to click to begin with, let alone double click it.

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Brian Pepple]

On Sun, 2008-03-23 at 13:25 -0400, Dimi Paun wrote:
>  On Sun, 2008-03-23 at 13:09 -0400, Brian Pepple wrote:
>  > A mouse-middle-key double-click opens a directory, and closes it's
>  > parent.
>  
>  Perfectly mysterious, you can't expect anyone to know that, right?
>  We are talking about a default that would keep newbies comfortable,
>  not people in-the-know.
>  
>  Not only is this rather obscure, but for a long time the middle button
>  has been a pain to access, as X seem to map it to any random button
>  when the mouse has more then a few buttons. 
>  
>  And when it doesn't, typically the middle button is the wheel which is
>  hard to click to begin with, let alone double click it. 
The help documentation (which I'm sure you've read, right?) also points
out that the user can also hold down Shift while double-clicking to get
the same affect.

Later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Hans de Goede]

Brian Pepple wrote:
>  On Sun, 2008-03-23 at 10:13 -0500, Les Mikesell wrote:
> > I don't see how anyone with directories more than 2 levels deep or uses 
> > it for system file access can tolerate the useless trail of intermediate 
> > windows it leaves behind.  Have I missed something basic about how to 
> > get rid of them?
>  
>  A mouse-middle-key double-click opens a directory, and closes it's
>  parent.
>   
Amazing what genius came up with that, try mouse middle button double clicking 
with middle butten emulation, or with a slightly flaky scrollwheel (iow most).

And yet the people in favor of spatial view keep claiming this was designed by 
a usability expert. As soon as someone says doubleclick with a different button 
then left, he loses all credibility as usability expert.

+1 to dimi for having a poll about changing the default to open all links in
    browser mode.

And +1 to the poll in favor of changing the default.

Regards,

Hans

[Dennis Jacobfeuerborn]

Hans de Goede wrote:
>  Brian Pepple wrote:
> > On Sun, 2008-03-23 at 10:13 -0500, Les Mikesell wrote:
> >> I don't see how anyone with directories more than 2 levels deep or
> >> uses it for system file access can tolerate the useless trail of
> >> intermediate windows it leaves behind. Have I missed something basic
> >> about how to get rid of them?
> >
> > A mouse-middle-key double-click opens a directory, and closes it's
> > parent.
> >
>
>  Amazing what genius came up with that, try mouse middle button double
>  clicking with middle butten emulation, or with a slightly flaky
>  scrollwheel (iow most).
>
>  And yet the people in favor of spatial view keep claiming this was
>  designed by a usability expert. As soon as someone says doubleclick with
>  a different button then left, he loses all credibility as usability expert.
>
>  +1 to dimi for having a poll about changing the default to open all
>  links in
>  browser mode.
>
>  And +1 to the poll in favor of changing the default. 
+1

Maybe I could get used to the spatial mode but the middle-double-click 
makes its use impossible for me. Who exactly was this usability expert 
thinking this was a good idea? I'd really like to know.

Regards,
   Dennis

[Jesse Keating]

On Sun, 2008-03-23 at 19:43 +0100, Dennis Jacobfeuerborn wrote:
>  
>  
>  Maybe I could get used to the spatial mode but the middle-double-click 
>  makes its use impossible for me. Who exactly was this usability expert 
>  thinking this was a good idea? I'd really like to know. 
Actually middle click is the alternative.  The actual default method is
to shift->double click.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Denis Leroy]

Hans de Goede wrote:
>  Brian Pepple wrote:
> > On Sun, 2008-03-23 at 10:13 -0500, Les Mikesell wrote:
> >> I don't see how anyone with directories more than 2 levels deep or 
> >> uses it for system file access can tolerate the useless trail of 
> >> intermediate windows it leaves behind.  Have I missed something basic 
> >> about how to get rid of them?
> >
> > A mouse-middle-key double-click opens a directory, and closes it's
> > parent.
> >
>  
>  Amazing what genius came up with that, try mouse middle button double 
>  clicking with middle butten emulation, or with a slightly flaky 
>  scrollwheel (iow most). 
Middle-click was always a problem with Gnome. I filed this

https://bugzilla.redhat.com/show_bug.cgi?id=437280

for metacity (see upstream bug), but these type of bugs are almost 
systematically ignored.

+1 for changing the default (not because of middle click, but because 
I'm convinced any semi-serious poll would show spatial mode to not be 
popular at all).

[Nicolas Mailhot]

Le dimanche 23 mars 2008 à 23:53 +0100, Denis Leroy a écrit :
>  Middle-click was always a problem with Gnome. I filed this
>  
>  https://bugzilla.redhat.com/show_bug.cgi?id=437280
>  
>  for metacity (see upstream bug), but these type of bugs are almost 
>  systematically ignored.
>  
>  +1 for changing the default (not because of middle click, but because 
>  I'm convinced any semi-serious poll would show spatial mode to not be 
>  popular at all). 
The default is a keyboard+mice combo
You ask for another keyboard+mice combo

And at the same time the GNOME UI experts insist the shortcut gnome
applet must be limited to keyboard-only, and mixing input devices makes
no sense.

Am I the only one to see the irony?

http://bugzilla.gnome.org/show_bug.cgi?id=373934

-- 
Nicolas Mailhot

[Toshio Kuratomi]

Les Mikesell wrote:
>  Jesse Keating wrote:
> > On Sun, 2008-03-23 at 01:55 -0300, Henrique Junior wrote:
> >> I want to ask to everyone here: who really the
> >> uses the spatial mode?
> >
> > I rather like the spatial mode and I keep it that way.
>  
>  I don't see how anyone with directories more than 2 levels deep or uses 
>  it for system file access can tolerate the useless trail of intermediate 
>  windows it leaves behind.  Have I missed something basic about how to 
>  get rid of them?
>   
<Shift> + <Double-Click>

-Toshio

[Les Mikesell]

Toshio Kuratomi wrote:
>  Les Mikesell wrote:
> > Jesse Keating wrote:
> >> On Sun, 2008-03-23 at 01:55 -0300, Henrique Junior wrote:
> >>> I want to ask to everyone here: who really the
> >>> uses the spatial mode?
> >>
> >> I rather like the spatial mode and I keep it that way.
> >
> > I don't see how anyone with directories more than 2 levels deep or 
> > uses it for system file access can tolerate the useless trail of 
> > intermediate windows it leaves behind.  Have I missed something basic 
> > about how to get rid of them?
> >
>  <Shift> + <Double-Click> 
Is there a way to reverse this default so I can do the weird combo in 
the rare event that I want an extra window left open as a side effect of 
navigating to where I intended?

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Andrew Farris]

Les Mikesell wrote:
>  Toshio Kuratomi wrote:
> > <Shift> + <Double-Click>
>  
>  Is there a way to reverse this default so I can do the weird combo in 
>  the rare event that I want an extra window left open as a side effect of 
>  navigating to where I intended? 
I've looked for how that keystroke is defined and I cannot find any way to swap 
these.  I would love to see an option added to change them, especially to fully 
customize it (if I want to double-right click that should be ok).  I've wondered 
whether it can be changed for awhile but never found info on it yet.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Jef Spaleta]

On Sun, Mar 23, 2008 at 7:13 AM, Les Mikesell <lesmikesell(a)gmail.com&gt; wrote:

>  I don't see how anyone with directories more than 2 levels deep
> or uses
>  it for system file access can tolerate the useless trail of intermediate
>  windows it leaves behind.  Have I missed something basic about how to
>  get rid of them?
>  
double-middle-click  will make the parent window close when you open the new
folder window.
I use double-middle-click religiously.

-jef

[Denis Leroy]

Jeff Spaleta wrote:
>  double-middle-click  will make the parent window close when you open
> the 
>  new folder window.
>  I use double-middle-click religiously. 
I take it your mouse doesn't have a scroll-wheel :-)

[Jesse Keating]

On Sun, 2008-03-23 at 23:55 +0100, Denis Leroy wrote:
>  I take it your mouse doesn't have a scroll-wheel :-) 
Actually one of mine does, and I continue to use it there.  Also one of
my mice is an Apple MightyMouse and has a ball for the middle button and
I continue to use double middle click there.  The only place I don't is
on my laptop where there is no button and there I use shift middle
click.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Denis Leroy]

Jesse Keating wrote:
>  On Sun, 2008-03-23 at 23:55 +0100, Denis Leroy wrote:
> > I take it your mouse doesn't have a scroll-wheel :-)
>  
>  Actually one of mine does, and I continue to use it there.  Also one of
>  my mice is an Apple MightyMouse and has a ball for the middle button and
>  I continue to use double middle click there.  The only place I don't is
>  on my laptop where there is no button and there I use shift middle
>  click. 
The Logitech mice I have have always made any middle-click operation 
somewhat awkward...

Jesse out of curiosity, when you open a folder, at what ratio do you 
decide to open into a new window versus opening in the same window (with 
  either shift or middle-clicks) ? (I'm sure you can see where I'm going 
with that question :-) ).

[Jesse Keating]

On Mon, 2008-03-24 at 00:18 +0100, Denis Leroy wrote:
>  The Logitech mice I have have always made any middle-click operation
>
>  somewhat awkward...
>  
>  Jesse out of curiosity, when you open a folder, at what ratio do you 
>  decide to open into a new window versus opening in the same window (with 
>    either shift or middle-clicks) ? (I'm sure you can see where I'm going 
>  with that question :-) ). 
Depends on what I'm doing.  Here is a common scenario:

I open up the bookmark for the Samba share on the machine I do Fedora
composes on.  That's a normal double click.  I then middle click through
a couple folders to get to the top level compose dir.  Then it's a
normal double click into the specific compose dir, and one of each of
the arch folders, then middle click again into the iso/ dir.  That
generally leaves me with a folder for the composes, a folder for the
arches, and a folder for the isos all open, so that I can go back and
look at other arches, or other composes easily.  Also the windows all
open in convenient locations so that each is viewable and don't overlap.
Then I can easily right click on .iso files and burn them via nautilus.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Dimi Paun]

On Sun, 2008-03-23 at 19:46 -0400, Jesse Keating wrote:
>  Depends on what I'm doing.  Here is a common scenario:
>  
>  I open up the bookmark for the Samba share on the machine I do Fedora
>  composes on.  That's a normal double click.  I then middle click
>  through
>  a couple folders to get to the top level compose dir.  Then it's a
>  normal double click into the specific compose dir, and one of each of
>  the arch folders, then middle click again into the iso/ dir.  That
>  generally leaves me with a folder for the composes, a folder for the
>  arches, and a folder for the isos all open, so that I can go back and
>  look at other arches, or other composes easily.  Also the windows all
>  open in convenient locations so that each is viewable and don't
>  overlap. 
Great, but your scenario would work the same or better (depending on the
depth of your tree) if you had to Shift-DblClick to open a new window.
And it would have the huge advantage of not violating the principle of
least surprise in that right now for vast majority of newbies (read
users used to Windows), you have to do something special just to perform
a seemingly trivial operation.

Folks, I'm not sure people understand what a brutal experience it is for
someone to switch to Linux: there are problems left-right-and-center [1]
For a newbie that has battled recreating their work environment for
hours or days, the last thing they want is to be shown how clueless they
are because they can't even browse to the files the way they were used
to! We are just shooting ourselves in the foot, for seemingly no good
purpose.

[1] This is not an attack on anybody. We've made great strides in the
last few years, and it shows. Quite a few people should be extremely
proud of the results. But we're not out of the woods. I've recently
migrated my wife's Laptop (ThinkPag X60) to Linux. The first part
was so easy and nice it was a revelation: parted and installing Fedora.
These steps were so nice that she remarked on how smooth everything
went. But past that it was a bloody nightmare, trying to recreate
her working environment: flash+sound problems, Evolution bugs, Palm
synchronization, Bluetooth, etc.

We spent the entire day together trying to get it to work, and by
1am she was crying her eyes out in frustration and desperation. It
was sad, and despite my best efforts I couldn't help her. It took
another 2 days of fscking around with ppp and all sorts of weird
daemons, and enlisting the help of one of my kernel-hacking friends
just to get her Palm to sync! Brutal doesn't even begin to describe it. 

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Callum Lerwick]

On Sun, 2008-03-23 at 20:33 -0400, Dimi Paun wrote:
>  Great, but your scenario would work the same or better (depending on
>  the
>  depth of your tree) if you had to Shift-DblClick to open a new window.
>  And it would have the huge advantage of not violating the principle of
>  least surprise in that right now for vast majority of newbies (read
>  users used to Windows), you have to do something special just to
>  perform
>  a seemingly trivial operation. 
Funny, I've never heard a single complaint about spacial mode from my
wife. And I use it myself as well. Of course, though neither of us have
ever owned Macs at home, we both grew up in school districts dominated
by Macs. (And Apple II's if you go back far enough. Mind you, we both
were entering college by the time OSX was released... Read up on MECC
(R.I.P.), Minnesota and Apple go waaaaaay back...) Hell, any half-way
advanced Mac OS user already knows about the shift-doubleclick shortcut.

If you truly care about UI design, you'll have read this already:

http://www.joelonsoftware.com/uibook/fog0000000249.html

Towards the middle of Chapter 1, there's a bit about Windows vs Mac that
seems appropriate here. Unfortunately it doesn't really provide any
direct advice as to how to reconcile the two. GNOME seems to have gone
down the path of cribbing from Apple the past 5 years or so. When you
care enough to steal the very best...

http://micro.magnet.fsu.edu/creatures/pages/russians.html

[Brian Pepple]

On Sun, 2008-03-23 at 23:55 +0100, Denis Leroy wrote:
>  Jeff Spaleta wrote:
>  > double-middle-click  will make the parent window close when you open the 
>  > new folder window.
>  > I use double-middle-click religiously.
>  
>  I take it your mouse doesn't have a scroll-wheel :-) 
Nope, all my machines have a scroll-wheel mouse except for my laptop.

Later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Jef Spaleta]

On Sun, Mar 23, 2008 at 2:55 PM, Denis Leroy <denis(a)poolshark.org&gt; wrote:

>  Jeff Spaleta wrote:
>  > double-middle-click  will make the parent window close when you open the
>  > new folder window.
>  > I use double-middle-click religiously.
>
>  I take it your mouse doesn't have a scroll-wheel :-)
>  
Some of them do, some of them don't.

Not making use of the middle click for anything because some mouse
manufacturers have decided to overload the function of the mechanical system
related to traditional 2nd mouse button interaction with a addtional scroll
functionality that is too sensitive.. is patently stupid.  The problem you
have is with the design of the mechanical system of the device you have.  If
your mouse can not be operated such that its easy to both click and scroll
with the middle button, then it is a failure of mechanical design of the
device.  The same problem would exist if the boneheaded manufacturers
decided to add a scroll function to the 1st or 3rd mouse buttons.  Its the
same sort of design problem that people have with trackpads with double
tapping for clicking enabled. Overloading a physical apperatus with multiple
functions descrimated by similar physical interactions makes it harder to
consistently use all exposed functions of the device.

-jef"If you can't suck it up and deal with your scroll wheel as a second
mouse button, enable chording of the 1st and 3rd and just forget that your
pos mouse device actually has a physical 2nd button and just chord when you
need to"spaleta

[Dimi Paun]

On Sun, 2008-03-23 at 17:07 -0800, Jeff Spaleta wrote:
>  -jef"If you can't suck it up and deal with your scroll wheel
> as a
>  second mouse button, enable chording of the 1st and 3rd and just
>  forget that your pos mouse device actually has a physical 2nd button
>  and just chord when you need to"spaleta   
But what would be the point?!? Do you expect newbies to know how to do
that? In fact, I don't know, and I'm not a newbie. In fact, most people
wouldn't probably know that you could even *do* such a thing, let alone
know how to do it.

Like it or not, most mice these days have a mouse wheel that doubles as
middle button. In fact, out of the 6 mice I have around me right now,
all fall in that category. And no matter how boneheaded the manufacturer
is, this is what we have to work with.

All I heard up to this point are complicated/unreasonable/unfeasible
solutions for newbies for a problem that we created and it's easy to
fix. Wouldn't it be a lot more reasonable to give people by default what
they are used to, and let power-users flip a switch? 

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Jef Spaleta]

On Sun, Mar 23, 2008 at 5:20 PM, Dimi Paun <dimi(a)lattica.com&gt; wrote:

>  All I heard up to this point are complicated/unreasonable/unfeasible
>  solutions for newbies for a problem that we created and it's easy to
>  fix. Wouldn't it be a lot more reasonable to give people by default what
>  they are used to, and let power-users flip a switch?
>  
I assure you WE didnt create the 2nd mouse button.. it has existed
historically for decades now. WE just make use of it, because its there to
be made use of. I've worked on vms and unix system well before windows
3.1was available where the 2nd mouse button was as useful as the
left-most
button.  If you can't use your scroll-wheel and the middle button equally
effectively.. then you have a piece of junk mouse and you should demand your
money back. It's a basic design flaw in the hardware, no worse than buying a
car and finding out that the break and the accelerator are the same pedal
and being told that if you have to touch it in different ways to get the car
to either go faster or to stop.  That being said, I use my 2nd button on my
scroll wheel mouse just fine.
.
How about we just assume that all pointing devices have a single button and
nothing else, one big button....with no double-clicking either.  I know some
apple mice that are real close to this.  You code software assuming that,
and you should work around pretty much any physical device design flaws.
And such software should pretty much work for the most common
denominator...exposing all functionality by requiring the use of keyboard
and mouse to expose additional functionality.   And in the case of the
functionality which started this gripefest, you have that very solution at
hand... middle click is exposed via the shift key as previously explained.

-jef

[Les Mikesell]

Jeff Spaleta wrote:
>  On Sun, Mar 23, 2008 at 5:20 PM, Dimi Paun <dimi(a)lattica.com&gt;
> wrote:
>  
> > All I heard up to this point are complicated/unreasonable/unfeasible
> > solutions for newbies for a problem that we created and it's easy to
> > fix. Wouldn't it be a lot more reasonable to give people by default what
> > they are used to, and let power-users flip a switch?
> >
>  
>  I assure you WE didnt create the 2nd mouse button.. it has existed
>  historically for decades now. WE just make use of it, because its there to
>  be made use of.  
Yes, but that has nothing to do with the assumption that navigating with 
the file manager should open new windows as a surprising and 
difficult-to-avoid side effect.  Applications are supposed to have an 
'open new window' option in the 'File' menu so it isn't hard to get
one 
when you want it.  There's no reason to leave a trail of them as an 
accidental side effect of a normal double-click or right-mouse/open.
It isn't that the oddball action is impossible, it is just wrong.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Andrew Farris]

Les Mikesell wrote:
>  Jeff Spaleta wrote:
> > On Sun, Mar 23, 2008 at 5:20 PM, Dimi Paun <dimi(a)lattica.com&gt; wrote:
> >
> >> All I heard up to this point are complicated/unreasonable/unfeasible
> >> solutions for newbies for a problem that we created and it's easy to
> >> fix. Wouldn't it be a lot more reasonable to give people by default what
> >> they are used to, and let power-users flip a switch?
> >>
> >
> > I assure you WE didnt create the 2nd mouse button.. it has existed
> > historically for decades now. WE just make use of it, because its 
> > there to
> > be made use of. 
>  
>  Yes, but that has nothing to do with the assumption that navigating with 
>  the file manager should open new windows as a surprising and 
>  difficult-to-avoid side effect.  Applications are supposed to have an 
>  'open new window' option in the 'File' menu so it isn't hard to get
> one 
>  when you want it.  There's no reason to leave a trail of them as an 
>  accidental side effect of a normal double-click or right-mouse/open.
>  It isn't that the oddball action is impossible, it is just wrong. 
Wrong?  There is nothing wrong with spatial browsing...

Surprising to you?  Surprising to whom?  A person new to linux who has no 
exposure to Windows does not find this behavior surprising, I assure you.  A 
window gets opened, you then open another folder... one without any prior 
expectations will observe that opening a different folder should open a new 
window.  Look at your (physical) desktop, with file folders on it and papers 
inside for the reason behind it.

A person new to nautilus spatial browsing but who has experience with linux may 
find it surprising, and so will someone who has experience with Windows Explorer 
but who has never seen Apple OS.  If this argument is going to be used lets get 
to the root of it... is conforming to the 'de facto' standard (folder trees) in 
browsing files in the Windows world important?  There is no wrong.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Les Mikesell]

Andrew Farris wrote:
>  
> > It isn't that the oddball action is impossible, it is just wrong.
>  
>  Wrong?  There is nothing wrong with spatial browsing...
>  
>  Surprising to you?  Surprising to whom? 
To anyone who has used any other windowing system or any application 
that has a File/Open new window option and expects to get new windows 
only when they ask for them.

>  A person new to linux who has 
>  no exposure to Windows does not find this behavior surprising, I assure 
>  you.  
Why would someone only use one OS? And why should you have to be such a 
person to find this behavior tolerable?

>  A window gets opened, you then open another folder... one without
>  any prior expectations will observe that opening a different folder 
>  should open a new window. 
And when they go more than a few folders deep they'll still be annoyed 
at all the useless still-open windows left around even if they expect them.

>  A person new to nautilus spatial browsing but who has experience with
>
>  linux may find it surprising, and so will someone who has experience 
>  with Windows Explorer but who has never seen Apple OS.  
The finder in OS X doesn't clutter my screen that way - at least in 
10.5.  What Apple OS do you mean?

>  If this argument 
>  is going to be used lets get to the root of it... is conforming to the 
>  'de facto' standard (folder trees) in browsing files in the Windows 
>  world important?  There is no wrong. 
"Different for no reason" is wrong for anyone who works with more than 
one system.  And opening new windows all over the place is different 
even from earlier versions of nautilus itself.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Andrew Farris]

Les Mikesell wrote:
>  Andrew Farris wrote:
> >
> >> It isn't that the oddball action is impossible, it is just wrong.
> >
> > Wrong?  There is nothing wrong with spatial browsing...
> >
> > Surprising to you?  Surprising to whom?
>  
>  To anyone who has used any other windowing system or any application 
>  that has a File/Open new window option and expects to get new windows 
>  only when they ask for them.
>  
> > A person new to linux who has no exposure to Windows does not find 
> > this behavior surprising, I assure you. 
>  
>  Why would someone only use one OS? And why should you have to be such a 
>  person to find this behavior tolerable? 
I use many of them and its possible to configure Windows Vista/XP/2000 to behave 
almost exactly the same way.. with the exception that windows does not offer you 
the ability to shift-click your way through (closing parent windows) and it 
cannot remember screen location for each directory separately.  Because it lacks 
these things it is far less effective to use spatial browsing.. hence why I use 
tree views in Explorer.

>   > A window gets opened, you then open another folder... one
> without
> > any prior expectations will observe that opening a different folder 
> > should open a new window.
>  
>  And when they go more than a few folders deep they'll still be annoyed 
>  at all the useless still-open windows left around even if they expect them. 
Not necessarily.

> > A person new to nautilus spatial browsing but who has experience
> with 
> > linux may find it surprising, and so will someone who has experience 
> > with Windows Explorer but who has never seen Apple OS. 
>  
>  The finder in OS X doesn't clutter my screen that way - at least in 
>  10.5.  What Apple OS do you mean? 
Thats because Apple has chosen not to make it default behavior, not because it 
is not included.  Take a look in the Finder preferences and you'll find it right 
there (always open in new window) and then set your finder mode to icon view 
(cmd-1) and start browsing spatially.  OSX will behave very similarly to Gnome 
when you've done that, remembering window placement for any directory you have 
opened.

Whether it is the default behavior is not something I'm really concerned about, 
only the perception that it is somehow 'wrong' because you don't like it.

> > If this argument is going to be used lets get to the root of
> it... is 
> > conforming to the 'de facto' standard (folder trees) in browsing files 
> > in the Windows world important?  There is no wrong.
>  
>  "Different for no reason" is wrong for anyone who works with more than 
>  one system.  And opening new windows all over the place is different 
>  even from earlier versions of nautilus itself. 
I suggest you go have a look through gnome development mailing lists for 
discussion on spatial browsing if you really are interested (especially if you 
want to argue it should not be the default for upstream).  You might also find 
this [1] interesting (see point 6).

This thread is one more example of why HCI is still (3 years after this blog) in 
the stone ages.. because people continue to demand things work the way they 
first learned them to work even when it makes very little sense from a 
perspective of how a human might best work with a computer.

And interesting read [2] on why the 'desktop' itself is a poor interface 
destined to be forgotten and left behind as we learn to interact with our 
computers in far more complicated ways.

[1] 
http://juicability.blogspot.com/2005/09/top-8-reasons-hci-is-in-its-stone...
[2] http://www.temple.edu/ispr/examples/ex01_12_24.html

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Les Mikesell]

Andrew Farris wrote:
>  
> > And when they go more than a few folders deep they'll still be annoyed 
> > at all the useless still-open windows left around even if they expect 
> > them.
>  
>  Not necessarily. 
How can anyone possibly want all of the intermediate windows left open 
when they really just want to get to some deep path location?

> >> A person new to nautilus spatial browsing but who has
> experience with 
> >> linux may find it surprising, and so will someone who has experience 
> >> with Windows Explorer but who has never seen Apple OS. 
> >
> > The finder in OS X doesn't clutter my screen that way - at least in 
> > 10.5.  What Apple OS do you mean?
>  
>  Thats because Apple has chosen not to make it default behavior, not 
>  because it is not included. 
And that's because Apple makes some effort to give people a better 
experience.

>   Take a look in the Finder preferences and 
>  you'll find it right there (always open in new window) and then set your 
>  finder mode to icon view (cmd-1) and start browsing spatially. 
But I don't want to browse spatially.

>  OSX will 
>  behave very similarly to Gnome when you've done that, remembering window 
>  placement for any directory you have opened. 
I'll control that myself, thank you.

>  Whether it is the default behavior is not something I'm really
> concerned 
>  about, only the perception that it is somehow 'wrong' because you don't 
>  like it. 
No option is 'wrong' if a user sets it himself. In that case it isn't 
anyone else's business. The default behavior is the only one where you 
can pass judgment.

>  I suggest you go have a look through gnome development mailing lists
> for 
>  discussion on spatial browsing if you really are interested (especially 
>  if you want to argue it should not be the default for upstream).  You 
>  might also find this [1] interesting (see point 6). 
If I understand point 6 to mean that spatial browsing relates more 
closely to physical objects, that makes sense and is why I don't like 
it.  If I wanted things to be as inconvenient as physical objects I 
wouldn't be sitting at a desk using a computer.  I want the objects to 
come to me, not to be frozen in some inconvenient distant space.  And I 
want them to clean up after themselves better than things in the 
physical world.

>  This thread is one more example of why HCI is still (3 years after
> this 
>  blog) in the stone ages.. because people continue to demand things work 
>  the way they first learned them to work even when it makes very little 
>  sense from a perspective of how a human might best work with a computer. 
I'm not demanding things to work the same as I first learned them, I 
just want changes to be for the better, not worse.  The problem is not 
so much about the attributes of spatial windows, although I much prefer 
to control those attributes by my view instead of having them attached 
to the object itself, the real problem is that opening unwanted windows 
is a side effect of navigation.  I don't want to have to remember some 
unnatural action for navigation vs. end point choices and I want an 
explict 'open a new window' when I reach locations that I want left open.

>  And interesting read [2] on why the 'desktop' itself is a
> poor interface 
>  destined to be forgotten and left behind as we learn to interact with 
>  our computers in far more complicated ways. 
No argument there, but again, I want the objects/options to come to me, 
not to be hiding in locations distant from my mouse pointer, especially 
as screens get bigger.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Henrique Junior]

Hello, folks.

With this discussion we can see how the thinking of a developer may be
far from the thinking of an end user. Everything an end user wants is
to use, simply as that. He assumes the system is working and to use
should be a simple thing. None of end users wants to read the
documentation before using (or my mother of 60 years old have to
Google and read the documentation of Nautilus in order to use it
without a million of opened windows on the desktop?). This spatial
mode "that brings the reality of a real desktop" to the virtual
desktop is absurd, because what it brings is the disorganization of
the real desktop, full of stuff scattered throughout the screen, which
is stressful and unproductive.

I don't know why for so long spatial remains the default of GNOME,
this only shows that the community is not being heard properly. The
Fedora Project should conduct polls to be more accessible to their end
users and not just the groups inside the project. Let's remember that
one of our slogans is the "voice of the community" and not "the voice
of the developers" or "the voice of ambassadors", because only for a
developer I could hear that 'OpenOffice is ugly and AbiWord is cool'.
AbiWord is cool inside its limitations, but it is not a good choice
for offices. A FACT so obvious that don't deserve any discussion.

Forgive me if some words sounds more heavy that I want, but my English
is very poor and it difficults a lot.

-- 
Henrique "LonelySpooky" Junior
-------------------------------------------------------------
"In a world without walls and fences, who needs windows and gates?!"

[Andrew Farris]

Henrique Junior wrote:
>  Hello, folks.
>  
>  With this discussion we can see how the thinking of a developer may be
>  far from the thinking of an end user. Everything an end user wants is
>  to use, simply as that. He assumes the system is working and to use
>  should be a simple thing. None of end users wants to read the
>  documentation before using (or my mother of 60 years old have to
>  Google and read the documentation of Nautilus in order to use it
>  without a million of opened windows on the desktop?). This spatial
>  mode "that brings the reality of a real desktop" to the virtual
>  desktop is absurd, because what it brings is the disorganization of
>  the real desktop, full of stuff scattered throughout the screen, which
>  is stressful and unproductive.
>  
>  I don't know why for so long spatial remains the default of GNOME,
>  this only shows that the community is not being heard properly. The
>  Fedora Project should conduct polls to be more accessible to their end
>  users and not just the groups inside the project. 
The proper place for this discussion is the gnome community as has been said 
above in the thread... because Fedora supports the adoption of upstream default 
behaviors.  Have you actually gone looking for the upstream discussion of this 
topic?  It will make the reason for the default very clear.

Here are a few places to start.  For anyone to say they don't know why only says 
you have not actually looked for the explanation.  I'm sorry, but thats just how 
it appears.  This is *NOT NEW* its been discussed more than 4 years in reference 
to nautilus.

[1] http://www.bytebot.net/geekdocs/spatial-nautilus.html
[2] http://developers.slashdot.org/article.pl?sid=04/06/13/175252
[3] http://osnews.com/story/7344
[4] http://arstechnica.com/articles/paedia/finder.ars

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Andrew Farris]

Andrew Farris wrote:
>  [4] http://arstechnica.com/articles/paedia/finder.ars 
While that is one of the earliest full articles on the subject that I'm aware 
of, there was gnome nautilus list discussion earlier... and in the interest of 
quality discussion here is one of the primary starting points (afaik).

http://mail.gnome.org/archives/nautilus-list/2002-September/msg00093.html

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Andrew Farris]

Les Mikesell wrote:
>  Andrew Farris wrote:
> >
> >> And when they go more than a few folders deep they'll still be 
> >> annoyed at all the useless still-open windows left around even if 
> >> they expect them.
> >
> > Not necessarily.
>  
>  How can anyone possibly want all of the intermediate windows left open 
>  when they really just want to get to some deep path location?
>  
> >>> A person new to nautilus spatial browsing but who has experience 
> >>> with linux may find it surprising, and so will someone who has 
> >>> experience with Windows Explorer but who has never seen Apple OS. 
> >>
> >> The finder in OS X doesn't clutter my screen that way - at least in 
> >> 10.5.  What Apple OS do you mean?
> >
> > Thats because Apple has chosen not to make it default behavior, not 
> > because it is not included.
>  
>  And that's because Apple makes some effort to give people a better 
>  experience. 
No.  Apple has offered options, one of which maintains the prior behavior of 
Apple OS 6.x to 9 (spatial browsing) while also offering other options such as 
the multi-pane view.  To say they chose not to make spatial browsing the default 
because 'they want a better experience' is to once again claim your preference 
is the better experience.  I beg to differ.  I have the theory they decided not 
to make spatial browsing the default because they wanted to make sure everyone 
learns multi-pane tree structure browsing... one of their innovations in UI 
(i.e. look how cool this is).  I'm a huge fan of OS X, but I'm not going to 
gloss over the self-promoting preference choice there.

BTW, Nowhere in OS X will you find the nautilus browser tree view however... 
you'll find modifications to that idea which Apple wanted to add.  There just 
simply is no Windows Explorer type tree to click on little triangles.  Its not 
there for a reason, to distinguish OS X from the Windows experience rather than 
to be just like it.

> >  Take a look in the Finder preferences and you'll find it
> right there 
> > (always open in new window) and then set your finder mode to icon view 
> > (cmd-1) and start browsing spatially.
>  
>  But I don't want to browse spatially. 
I wasn't suggesting you did... that would be rather stupid given the emails 
you've already posted.  I was pointing out that you apparently are unaware of 
the fact that this file access mode exists.. in OSX (which your prior email 
claimed not to be true).

> > OSX will behave very similarly to Gnome when you've done
> that, 
> > remembering window placement for any directory you have opened.
>  
>  I'll control that myself, thank you. 
Ok, good deal.  I'll let the computer help me.

> > Whether it is the default behavior is not something I'm
> really 
> > concerned about, only the perception that it is somehow 'wrong' 
> > because you don't like it.
>  
>  No option is 'wrong' if a user sets it himself. In that case it isn't 
>  anyone else's business. The default behavior is the only one where you 
>  can pass judgment.
>  
> > I suggest you go have a look through gnome development mailing lists 
> > for discussion on spatial browsing if you really are interested 
> > (especially if you want to argue it should not be the default for 
> > upstream).  You might also find this [1] interesting (see point 6).
>  
>  If I understand point 6 to mean that spatial browsing relates more 
>  closely to physical objects, that makes sense and is why I don't like 
>  it.  If I wanted things to be as inconvenient as physical objects I 
>  wouldn't be sitting at a desk using a computer.  I want the objects to 
>  come to me, not to be frozen in some inconvenient distant space.  And I 
>  want them to clean up after themselves better than things in the 
>  physical world. 
Uh huh, so you think that a tree browsing view is somehow much closer to 'not 
physical' than spatial browsing?  You feel that a hierarchy of deeply enclosed 
directories is easier to access when you're still going one level at a time into 
the tree than using the spatial browsing mode?  Tree views, and opening inside 
the same window all the time is modeled directly from File Folders within Filing 
Cabinets.

Both are fundamentally flawed ways that we access and organize our files in a 
computer.  They are both inadequate, and will need to change in the future.  The 
longer we take to realize this it will postpone real progress in human computer 
interaction while everyone keeps demanding that things work the way they used to.

You can argue that holding down shift, or using middle-doubleclick is 
inconvenient if you like (I believe it works just fine).  My windows clean up 
after themselves beautifully.

Point 6 was not saying spatial browsing is a holy grail for the future, its 
saying that human beings are spatially oriented and naturally think in terms 
much more akin to spatial browsing.  Memory of where a window, or the file 
within the window, is supposed to be is very natural for us to remember as we 
'browse' through files.  Those files are supposed to be where they last were, 
hence why having a window show up where you had it last time makes sense.

It is contrary to the way many people have learned to find their files, and if 
they don't like it thats understandable; they don't like change.  They don't 
have to embrace change.  But if they are going to claim the change is going 
backwards I'd like to see more than personal preference in the rhetoric.

> > This thread is one more example of why HCI is still (3 years
> after 
> > this blog) in the stone ages.. because people continue to demand 
> > things work the way they first learned them to work even when it makes 
> > very little sense from a perspective of how a human might best work 
> > with a computer.
>  
>  I'm not demanding things to work the same as I first learned them, I 
>  just want changes to be for the better, not worse.  The problem is not 
>  so much about the attributes of spatial windows, although I much prefer 
>  to control those attributes by my view instead of having them attached 
>  to the object itself, the real problem is that opening unwanted windows 
>  is a side effect of navigation.  I don't want to have to remember some 
>  unnatural action for navigation vs. end point choices and I want an 
>  explict 'open a new window' when I reach locations that I want left open. 
You prefer trees.. there are those of us that do not.  I definitely have not 
suggested that anyone remove nautilus browser mode from gnome.

> > And interesting read [2] on why the 'desktop' itself is a
> poor 
> > interface destined to be forgotten and left behind as we learn to 
> > interact with our computers in far more complicated ways.
>  
>  No argument there, but again, I want the objects/options to come to me, 
>  not to be hiding in locations distant from my mouse pointer, especially 
>  as screens get bigger. 
You're free not to use your body's motor memory to your advantage if you want.

What you prefer is fine for you; I will take offense to you calling spatial 
browsing wrong however... it is what I and many other people prefer.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Les Mikesell]

Andrew Farris wrote:
> >>
> >> Thats because Apple has chosen not to make it default behavior, not 
> >> because it is not included.
> >
> > And that's because Apple makes some effort to give people a better 
> > experience.
>  
>  No.  Apple has offered options, one of which maintains the prior 
>  behavior of Apple OS 6.x to 9 (spatial browsing) while also offering 
>  other options such as the multi-pane view. 
But I'm only concerned about the defaults.

> > I'll control that myself, thank you.
>  
>  Ok, good deal.  I'll let the computer help me. 
I don't consider putting things where I don't want them to be 'help'.

> >
> >> I suggest you go have a look through gnome development mailing lists 
> >> for discussion on spatial browsing if you really are interested 
> >> (especially if you want to argue it should not be the default for 
> >> upstream).  You might also find this [1] interesting (see point 6).
> >
> > If I understand point 6 to mean that spatial browsing relates more 
> > closely to physical objects, that makes sense and is why I don't like 
> > it.  If I wanted things to be as inconvenient as physical objects I 
> > wouldn't be sitting at a desk using a computer.  I want the objects to 
> > come to me, not to be frozen in some inconvenient distant space.  And 
> > I want them to clean up after themselves better than things in the 
> > physical world.
>  
>  Uh huh, so you think that a tree browsing view is somehow much closer to 
>  'not physical' than spatial browsing? 
It's closer to putting something where I want it. Spatial browsing is 
more like having to find it where it was left last time.

>  You feel that a hierarchy of 
>  deeply enclosed directories is easier to access when you're still going 
>  one level at a time into the tree than using the spatial browsing mode?  
>  Tree views, and opening inside the same window all the time is modeled 
>  directly from File Folders within Filing Cabinets. 
The navigation is the same either way, but the spatial mode is like 
rooms with 2 doors and one light switch.  As you open the door you turn 
the light on and it remains on when you exit the other door.  Then later 
you have to backtrack and clean up.  I want the light to go off as I 
exit the other door.

>  Both are fundamentally flawed ways that we access and organize our
> files 
>  in a computer.  They are both inadequate, and will need to change in the 
>  future.  The longer we take to realize this it will postpone real 
>  progress in human computer interaction while everyone keeps demanding 
>  that things work the way they used to. 
I agree here - give me a folder navigation method that doesn't involve 
intermediate windows or having to type names and I could live with the 
position/attributes of the destination windows being attached to the 
destination and not my view.  But I still want an explicit 'open new 
window here' command.

>  You can argue that holding down shift, or using middle-doubleclick is
>
>  inconvenient if you like (I believe it works just fine).  My windows 
>  clean up after themselves beautifully. 
So why not make the default a normal double-click with a preference 
setting for people who prefer something oddball?

>  Point 6 was not saying spatial browsing is a holy grail for the
> future, 
>  its saying that human beings are spatially oriented and naturally think 
>  in terms much more akin to spatial browsing.  Memory of where a window, 
>  or the file within the window, is supposed to be is very natural for us 
>  to remember as we 'browse' through files.  Those files are supposed to 
>  be where they last were, hence why having a window show up where you had 
>  it last time makes sense.
>  
>  It is contrary to the way many people have learned to find their files, 
>  and if they don't like it thats understandable; they don't like change.  
>  They don't have to embrace change.  But if they are going to claim the 
>  change is going backwards I'd like to see more than personal preference 
>  in the rhetoric. 
The point of a computer is not to emulate the real world, it is so you 
can do this better and easier.  And no one in the real world nests 
folders 10 deep anyway.

> > No argument there, but again, I want the objects/options to come
> to 
> > me, not to be hiding in locations distant from my mouse pointer, 
> > especially as screens get bigger.
>  
>  You're free not to use your body's motor memory to your advantage if you 
>  want.
>  
>  What you prefer is fine for you; I will take offense to you calling 
>  spatial browsing wrong however... it is what I and many other people 
>  prefer. 
I don't have any problem with preference settings to enable it.  There 
are just too many things that are inconvenient about it to have it as a 
default - and in any case there should be an option to flip the 
mouse/keboard commands for the choice of whether the window stays open 
or not.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Andrew Farris]

Les Mikesell wrote:
>  Andrew Farris wrote:
> > You can argue that holding down shift, or using middle-doubleclick is 
> > inconvenient if you like (I believe it works just fine).  My windows 
> > clean up after themselves beautifully.
>  
>  So why not make the default a normal double-click with a preference 
>  setting for people who prefer something oddball? 
Sounds like a very good thing to take upstream and get an answer about.  I would 
like to see that configurable as I said before, but its not in the shortcuts 
right now.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Kevin Kofler]

Andrew Farris <lordmorgul <at> gmail.com> writes:
>  No.  Apple has offered options, one of which maintains the prior
> behavior of 
>  Apple OS 6.x to 9 (spatial browsing) 
But that's what it is: backwards compatibility, not the default anymore.

>  I have the theory they decided not to make spatial browsing the
> default
>  because they wanted to make sure everyone learns multi-pane tree structure
>  browsing... one of their innovations in UI (i.e. look how cool this is). 
Because that system actually makes sense, which is why Dolphin in KDE 4 
supports this.

Dolphin also makes it easy to switch between Icons (browser mode with icons and 
thumbnails), Details (browser mode with a list view showing file dates and 
attributes) and Columns (what you call "multi-pane tree structure") modes 
through a single click in the toolbar. In KDE 4.1, the traditional tree view 
will also be available.

>  But if they are going to claim the change is going backwards I'd
> like to see
>  more than personal preference in the rhetoric. 
The reason we claim spatial mode is a step backwards is that both M$ and Apple 
have used something very similar in the past and both moved away from it.

That said, I personally don't really care about what Nautilus defaults to, I 
use KDE anyway. ;-)

        Kevin Kofler

[Matej Cepl]

On 2008-03-27, 06:27 GMT, Kevin Kofler wrote:
>  The reason we claim spatial mode is a step backwards is that 
>  both M$ and Apple have used something very similar in the past 
>  and both moved away from it. 
This is actually on the edge of being actionable libel -- File 
Manager in Windows 3.11 is to Apple Finder / Nautilus in spatial 
mode as edlin is to ed/ex -- half-baked, misunderstood copy of 
the interesting idea.

I like spatial mode actually (I hated konqueror as file manager 
while being in KDE and used Krusader instead for lack of better 
file manager). I thought that I will have to find a Gnome 
replacement for Krusader (there is one -- gnome-commander), but 
I found myself pretty happy with default nautilus, so I stick 
with it (plus a lot of command line commands in gnome-terminal).

Best,

Matej

[Kevin Kofler]

Matej Cepl <mcepl <at> redhat.com> writes:
>  I like spatial mode actually (I hated konqueror as file manager 
>  while being in KDE and used Krusader instead for lack of better 
>  file manager). I thought that I will have to find a Gnome 
>  replacement for Krusader (there is one -- gnome-commander), but 
>  I found myself pretty happy with default nautilus, so I stick 
>  with it (plus a lot of command line commands in gnome-terminal). 
That's interesting because Krusader and spatial Nautilus are very different 
beasts. For what it's worth, I agree Krusader is a better file manager than 
Konqueror (I use Krusader all the time), but I think Nautilus can't compete 
with either, especially not in the awful spatial mode. And "for lack of better 
file manager" is borderline insulting for a great file manager like 
Krusader. ;-) But different people have different tastes...

        Kevin Kofler

[Andrew Farris]

Kevin Kofler wrote:
>  Andrew Farris <lordmorgul <at> gmail.com> writes:
> > No.  Apple has offered options, one of which maintains the prior behavior of 
> > Apple OS 6.x to 9 (spatial browsing)
>  
>  But that's what it is: backwards compatibility, not the default anymore.
>  
> > I have the theory they decided not to make spatial browsing the default
> > because they wanted to make sure everyone learns multi-pane tree structure
> > browsing... one of their innovations in UI (i.e. look how cool this is).
>  
>  Because that system actually makes sense, which is why Dolphin in KDE 4 
>  supports this.
>  
>  Dolphin also makes it easy to switch between Icons (browser mode with icons and 
>  thumbnails), Details (browser mode with a list view showing file dates and 
>  attributes) and Columns (what you call "multi-pane tree structure") modes 
>  through a single click in the toolbar. In KDE 4.1, the traditional tree view 
>  will also be available.
>  
> > But if they are going to claim the change is going backwards I'd like to see
> > more than personal preference in the rhetoric.
>  
>  The reason we claim spatial mode is a step backwards is that both M$ and Apple 
>  have used something very similar in the past and both moved away from it. 
They moved away from it because user acceptance was problematic due to users being 
unable/unwilling to read the tiniest bits of documentation to find out they don't have
to 
end up with a billion windows open while using spatial mode... not because the system is 
innately flawed.

I agree columned mode is better than purely spatial, but working from a tree or partial 
tree partial spatial (i.e. nautilus browser/windows explorer) is the worst of both
worlds.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Les Mikesell]

Andrew Farris wrote:
>  
>  They moved away from it because user acceptance was problematic due to 
>  users being unable/unwilling to read the tiniest bits of documentation 
>  to find out they don't have to end up with a billion windows open while 
>  using spatial mode... not because the system is innately flawed. 
The problem is that the users don't _want_ all the windows down a path 
left open and they don't want to use some oddball action to avoid it as 
a side effect.

>  I agree columned mode is better than purely spatial, but working from
> a 
>  tree or partial tree partial spatial (i.e. nautilus browser/windows 
>  explorer) is the worst of both worlds. 
What's bad about them - assuming you open as many instances as you want? 
   It would be somewhat handier if they had 'File/New Window' options to 
duplicate the current window and 'right-mouse/Open in New Window' so you 
didn't have to start from the top every time but most of the paths I use 
diverge early anyway.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Andrew Farris]

Les Mikesell wrote:
>  Andrew Farris wrote:
> >
> > They moved away from it because user acceptance was problematic due to 
> > users being unable/unwilling to read the tiniest bits of documentation 
> > to find out they don't have to end up with a billion windows open 
> > while using spatial mode... not because the system is innately flawed.
>  
>  The problem is that the users don't _want_ all the windows down a path 
>  left open and they don't want to use some oddball action to avoid it as 
>  a side effect. 
Agreed... but spatial vs not spatial has nothing to do with that.  This is 
instead a default behavior issue with bindings, and I would agree with you that 
it would be better if the default was to close parent folders and then holding 
shift+click was the method to keep parents open.  But that is not an argument 
for not using spatial browsing.. its an  argument for sending a patch upstream 
to change those behaviors.

> > I agree columned mode is better than purely spatial, but working
> from 
> > a tree or partial tree partial spatial (i.e. nautilus browser/windows 
> > explorer) is the worst of both worlds.
>  
>  What's bad about them - assuming you open as many instances as you want? 
>    It would be somewhat handier if they had 'File/New Window' options to 
>  duplicate the current window and 'right-mouse/Open in New Window' so you 
>  didn't have to start from the top every time but most of the paths I use 
>  diverge early anyway. 
Two things:
1) it forces you to open two windows and navigate to them both to do the VAST 
majority of file operations; it is much more common to need access to drag a 
file between folders than it is to move it within the same folder.
2) because most paths *do* diverge early as you say, and I find the same to be 
true, the spatial browsing behavior often gets the window you'll need open on 
the way to the file in the first place, meaning you don't actually navigate 
twice as you do with two windows in a browser mode.  The only effective way to 
fight that is to sacrifice use of dragging to move files and instead use 
copy/cut/paste everywhere you handle files, meaning you're either using two 
hands or working with context menus anyway.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Kevin Kofler]

Andrew Farris <lordmorgul <at> gmail.com> writes:
>  1) it forces you to open two windows and navigate to them both to do
> the VAST 
>  majority of file operations; it is much more common to need access to drag a 
>  file between folders than it is to move it within the same folder. 
That's an argument for a 2-pane file manager (e.g. Krusader), not a spatial 
one. I don't understand why the 2-pane interface isn't more common.

http://www.krusader.org/screenshots.php

        Kevin Kofler

[Les Mikesell]

Kevin Kofler wrote:
>  Andrew Farris <lordmorgul <at> gmail.com> writes:
> > 1) it forces you to open two windows and navigate to them both to do the VAST 
> > majority of file operations; it is much more common to need access to drag a 
> > file between folders than it is to move it within the same folder.
>  
>  That's an argument for a 2-pane file manager (e.g. Krusader), not a spatial 
>  one. I don't understand why the 2-pane interface isn't more common.
>  
>  http://www.krusader.org/screenshots.php
>   
No, it's an argument for a general purpose window manager that lets you 
open as many windows containing as many instances of the same or 
different programs as you want.  But we've had that for ages.  Why do so 
many applications pretend they have to manage their own multiple windows 
in quirky and different ways too?

-- 
   Les Mikesell
     lesmikesell(a)gmail.com

[Kevin Kofler]

Les Mikesell <lesmikesell <at> gmail.com> writes:
>  No, it's an argument for a general purpose window manager that
> lets you 
>  open as many windows containing as many instances of the same or 
>  different programs as you want.  But we've had that for ages.  Why do so 
>  many applications pretend they have to manage their own multiple windows 
>  in quirky and different ways too? 
Because you need exactly 2 panes to do most file management operations (copy, 
move etc.) and a single one for the remaining ones, so having a dozen windows 
with folders you happened to pass while looking for those 2 folders is only 
annoying and forces you to play Duck Hunt with the X buttons, the extra windows 
serve no useful purpose.

        Kevin Kofler

[Les Mikesell]

Kevin Kofler wrote:
>  Les Mikesell <lesmikesell <at> gmail.com> writes:
> > No, it's an argument for a general purpose window manager that lets you 
> > open as many windows containing as many instances of the same or 
> > different programs as you want.  But we've had that for ages.  Why do so 
> > many applications pretend they have to manage their own multiple windows 
> > in quirky and different ways too?
>  
>  Because you need exactly 2 panes to do most file management operations (copy, 
>  move etc.) and a single one for the remaining ones, so having a dozen windows 
>  with folders you happened to pass while looking for those 2 folders is only 
>  annoying and forces you to play Duck Hunt with the X buttons, the extra windows 
>  serve no useful purpose. 
That's not the alternative I meant.  The sane way is to explicitly open 
2 instances of your file manager in browser mode and position both the 
window and the view of each independently.  No funky key combo's to deal 
with, no extra panes/windows that you don't need, just normal window 
management.  Works exactly the same way across Windows/Mac/Linux. 
Cut/copy/paste, drag/drop all do what you expect.  Since you'll probably 
need them frequently, you can just minimize these windows and leave them 
available to revive from the task bar.  The only down side is a bit of 
wasted screen space for browser mode but that's a reasonable tradeoff to 
be able to navigate more easily.

--
   Les Mikesell
    lesmikesell(a)gmail.com

[Kevin Kofler]

Andrew Farris <lordmorgul <at> gmail.com> writes:
>  Thats because Apple has chosen not to make it default behavior, not
> because
>  it is not included. 
Interesting... "Apple does it that way" was one of the reasons given for the 
default in GNOME. It looks like Apple realized it's a bad idea. M$ went through 
that long ago, Window$ 95 worked that way, they added browser mode and changed 
the default to it in Window$ 98. Why does GNOME have to copy ancient design 
flaws which both M$ and Apple recognized as such?

        Kevin Kofler

[Jakub 'Livio' Rusinek]

Dnia 27-03-2008, czw o godzinie 06:15 +0000, Kevin Kofler pisze:
>  Andrew Farris <lordmorgul <at> gmail.com> writes:
>  > Thats because Apple has chosen not to make it default behavior, not because
>  > it is not included.
>  
>  Interesting... "Apple does it that way" was one of the reasons given for the 
>  default in GNOME. It looks like Apple realized it's a bad idea. M$ went through 
>  that long ago, Window$ 95 worked that way, they added browser mode and changed 
>  the default to it in Window$ 98. Why does GNOME have to copy ancient design 
>  flaws which both M$ and Apple recognized as such? 
Don't forget to mention users' opinion. They do not like spatial mode
because of many windows, but it's noticeably faster than browser mode.

The main reason of browser mode's long start are the toolbars. Disable
main toolbar for example, and browser mode loads faster.

Quite interesting...

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Rodd Clarkson]

On Thu, 2008-03-27 at 07:26 +0100, Jakub 'Livio' Rusinek wrote:
>  Dnia 27-03-2008, czw o godzinie 06:15 +0000, Kevin Kofler pisze:
>  > Andrew Farris <lordmorgul <at> gmail.com> writes:
>  > > Thats because Apple has chosen not to make it default behavior, not because
>  > > it is not included.
>  > 
>  > Interesting... "Apple does it that way" was one of the reasons given for
> the 
>  > default in GNOME. It looks like Apple realized it's a bad idea. M$ went through
>
>  > that long ago, Window$ 95 worked that way, they added browser mode and changed 
>  > the default to it in Window$ 98. Why does GNOME have to copy ancient design 
>  > flaws which both M$ and Apple recognized as such?
>  
>  Don't forget to mention users' opinion. They do not like spatial mode
>  because of many windows, but it's noticeably faster than browser mode. 
I like spacial browsing, so maybe I'm strange. ;-]


R


-- 
"It's a fine line between denial and faith.
 It's much better on my side"

[Jakub 'Livio' Rusinek]

Dnia 27-03-2008, czw o godzinie 21:11 +1100, Rodd Clarkson pisze:
>  On Thu, 2008-03-27 at 07:26 +0100, Jakub 'Livio' Rusinek
> wrote:
>  > Dnia 27-03-2008, czw o godzinie 06:15 +0000, Kevin Kofler pisze:
>  > > Andrew Farris <lordmorgul <at> gmail.com> writes:
>  > > > Thats because Apple has chosen not to make it default behavior, not
> because
>  > > > it is not included.
>  > > 
>  > > Interesting... "Apple does it that way" was one of the reasons given
> for the 
>  > > default in GNOME. It looks like Apple realized it's a bad idea. M$ went
> through 
>  > > that long ago, Window$ 95 worked that way, they added browser mode and changed
>
>  > > the default to it in Window$ 98. Why does GNOME have to copy ancient design 
>  > > flaws which both M$ and Apple recognized as such?
>  > 
>  > Don't forget to mention users' opinion. They do not like spatial mode
>  > because of many windows, but it's noticeably faster than browser mode.
>  
>  I like spacial browsing, so maybe I'm strange. ;-] 
Yes, you're very strange, because you like spatial mode... Joking ;) .

Just your option ;) .

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Les Mikesell]

Rodd Clarkson wrote:
>  
> >>> Thats because Apple has chosen not to make it default behavior, not because
> >>> it is not included.
> >> Interesting... "Apple does it that way" was one of the reasons given
> for the 
> >> default in GNOME. It looks like Apple realized it's a bad idea. M$ went
> through 
> >> that long ago, Window$ 95 worked that way, they added browser mode and changed 
> >> the default to it in Window$ 98. Why does GNOME have to copy ancient design 
> >> flaws which both M$ and Apple recognized as such?
> > Don't forget to mention users' opinion. They do not like spatial mode
> > because of many windows, but it's noticeably faster than browser mode.
>  
>  I like spacial browsing, so maybe I'm strange. ;-] 
If you have very shallow directory trees and don't attempt system 
administration or navigating network shares it is sort-of tolerable. 
But what good is a file manager that makes it too inconvenient to use 
for real work?

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Jesse Keating]

On Thu, 2008-03-27 at 08:11 -0500, Les Mikesell wrote:
>  If you have very shallow directory trees and don't attempt system
>
>  administration or navigating network shares it is sort-of tolerable. 
>  But what good is a file manager that makes it too inconvenient to use 
>  for real work? 
I have "real" filesystem stuff, I just make proper use of when to double
click and when to shift-double click.  Often I'm moving content from one
dir to another, and doing that by having multiple windows open to views
of those folders (and some of their parents) makes this a lot easier
than a single browse like window.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Les Mikesell]

Jesse Keating wrote:
>  On Thu, 2008-03-27 at 08:11 -0500, Les Mikesell wrote:
> > If you have very shallow directory trees and don't attempt system 
> > administration or navigating network shares it is sort-of tolerable. 
> > But what good is a file manager that makes it too inconvenient to use 
> > for real work?
>  
>  I have "real" filesystem stuff, I just make proper use of when to double
>  click and when to shift-double click.  Often I'm moving content from one
>  dir to another, and doing that by having multiple windows open to views
>  of those folders (and some of their parents) makes this a lot easier
>  than a single browse like window. 
But, but, but... You've already got a perfectly fine multitasking, 
multiwindowing system where you can open as many instances of something 
as you want.  If you want several file manager window open, just open 
them and park them where you want separately.  Why do applications need 
to have anything to do with this on their own?

-- 
   Les Mikesell
    lesmikesesll(a)gmail.com

[Jesse Keating]

On Thu, 2008-03-27 at 08:51 -0500, Les Mikesell wrote:
>  But, but, but... You've already got a perfectly fine
> multitasking, 
>  multiwindowing system where you can open as many instances of something 
>  as you want.  If you want several file manager window open, just open 
>  them and park them where you want separately.  Why do applications need 
>  to have anything to do with this on their own? 
It's still a multiwindowing multitasking system, it's just showing me a
more convenient view of things.  Add to that remembering where those
windows go means I can work even more efficiently on commonly used
directory sets since everything will be where my muscles remember them.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Les Mikesell]

Jesse Keating wrote:
>  On Thu, 2008-03-27 at 08:51 -0500, Les Mikesell wrote:
> > But, but, but... You've already got a perfectly fine multitasking, 
> > multiwindowing system where you can open as many instances of something 
> > as you want.  If you want several file manager window open, just open 
> > them and park them where you want separately.  Why do applications need 
> > to have anything to do with this on their own?
>  
>  It's still a multiwindowing multitasking system, it's just showing me a
>  more convenient view of things.  
But you can arrange the same view even more conveniently yourself by 
explicitly opening the few locations you want and positioning the 
windows where you want them instead of having to do oddball commands to 
not get all the intermediate windows you don't want if they stay around 
by default.

>  Add to that remembering where those
>  windows go means I can work even more efficiently on commonly used
>  directory sets since everything will be where my muscles remember them. 
I rarely do repetitious work in GUI's since it is so much easier to 
script it or recall/re-execute command lines.  But what I do want is 
consistency in the way apps act regarding opening new windows instead of 
having to remember different actions depending on which app the mouse 
happens to be positioned over. And it doesn't make any more sense to me 
for a file browser to default to open a new window on every click than 
it would for a web browser. What I'd like in both, and for everything 
else where it would make sense, is to have a right-mouse menu with 'open 
in new window' as an alternative to 'open'.  I'd like the choice, but in 
my experience I'll want to use it much less often than using the current 
window.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Dimi Paun]

On Thu, 2008-03-27 at 09:17 -0400, Jesse Keating wrote:
>  I have "real" filesystem stuff, I just make proper use of
> when to
>  double click and when to shift-double click. 
And you do that because the spacial concept (while interesting), is
broken as a UI metaphor for folders.

Spacial makes a lot of sense for "objects": computer, Network, a disk.
Those have real-world corespondents, and people relate to that. However,
for most people, folders are seen as a _path_ to the object that you
want to work with. (duh, we even call it a "path" ...)

Any sane engineering/UI design/consistency principle that was listed in
this discussion suggests that going with a non-spacial default is the
only reasonable thing to do. Not only would we please a lot more users,
we wouldn't be violating the principle of least surprise, but from the
use cases listed, it seems that even the spacial-proponents would be
just as happy.

Folks, this is not a corner case -- it's the first thing new users bang
their heads on. And it creates a bad impression. It's an interesting
idea, it had 4 years to prove itself, it failed. Listing all sorts of
theoretical reasons why it _should_ be better is just mental
masturbation. 

If you think people like, let's run a poll on Slashdot...

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Jonathan Underwood]

On 27/03/2008, Dimi Paun <dimi(a)lattica.com&gt; wrote:
>   On Thu, 2008-03-27 at 09:17 -0400, Jesse Keating wrote:
>   > I have "real" filesystem stuff, I just make proper use of when to
>   > double click and when to shift-double click.
>
>   And you do that because the spacial concept (while interesting), is
>   broken as a UI metaphor for folders.
>
>   Spacial makes a lot of sense for "objects": computer, Network, a disk.
>   Those have real-world corespondents, and people relate to that. However,
>   for most people, folders are seen as a _path_ to the object that you
>   want to work with. (duh, we even call it a "path" ...)
>
>   Any sane engineering/UI design/consistency principle that was listed in
>   this discussion suggests that going with a non-spacial default is the
>   only reasonable thing to do. Not only would we please a lot more users,
>   we wouldn't be violating the principle of least surprise, but from the
>   use cases listed, it seems that even the spacial-proponents would be
>   just as happy.
>
>   Folks, this is not a corner case -- it's the first thing new users bang
>   their heads on. And it creates a bad impression. It's an interesting
>   idea, it had 4 years to prove itself, it failed. Listing all sorts of
>   theoretical reasons why it _should_ be better is just mental
>   masturbation. 
Personally speaking, the whole spatial nautilus thing became a lot
more usable when I realized what the middle mouse button click does.

[Dimi Paun]

On Thu, 2008-03-27 at 15:00 +0000, Jonathan Underwood wrote:
>  Personally speaking, the whole spatial nautilus thing became a lot
>  more usable when I realized what the middle mouse button click does. 
I understand. Most people will not "discover" it however (I've been
using Linux for 10+ years and I didn't know), and will be frustrated.

We're just arguing the painfully obvious: let the people in-the-know
that want to use spatial discover the magic incantation. Give the
regular users what they expect by default.

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Brian Pepple]

On Thu, 2008-03-27 at 11:16 -0400, Dimi Paun wrote:
>  On Thu, 2008-03-27 at 15:00 +0000, Jonathan Underwood wrote:
>  > Personally speaking, the whole spatial nautilus thing became a lot
>  > more usable when I realized what the middle mouse button click does.
>  
>  I understand. Most people will not "discover" it however (I've been
>  using Linux for 10+ years and I didn't know), and will be frustrated. 
Umm, the default method *is* in the help file.

later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Dimi Paun]

On Thu, 2008-03-27 at 11:25 -0400, Brian Pepple wrote:
>  Umm, the default method *is* in the help file. 
If a user needs to read the help just to get to a file there's
something wrong. People normally don't read in-your-face-alerts,
(as people have discovered around here), let alone obscure help 
files for trivial things.

Blaming it on the user that they didn't read the help is just
counter-productive. If they had to do it in the first place,
we've done something wrong.

Why is that so hard to face?

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Les Mikesell]

Brian Pepple wrote:
>  On Thu, 2008-03-27 at 11:16 -0400, Dimi Paun wrote:
> > On Thu, 2008-03-27 at 15:00 +0000, Jonathan Underwood wrote:
> >> Personally speaking, the whole spatial nautilus thing became a lot
> >> more usable when I realized what the middle mouse button click does.
> > I understand. Most people will not "discover" it however (I've been
> > using Linux for 10+ years and I didn't know), and will be frustrated.
>  
>  Umm, the default method *is* in the help file.
>   
Where do you find the way to swap the left/middle actions?  And what 
works best with a trackpad?

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Jesse Keating]

On Thu, 2008-03-27 at 10:35 -0500, Les Mikesell wrote:
>  
>  Where do you find the way to swap the left/middle actions?  And what 
>  works best with a trackpad? 
Are you being purposefully blunt?  Using the shift key to modify your
double click is the default way, middle click is the alternative.

To close the current folder while opening the new one, hold down Shift
when double-clicking, or press Shift+Alt+down arrow.

-- 
Jesse Keating
Fedora -- All my bits are free, are yours?

[Les Mikesell]

Jesse Keating wrote:
>  On Thu, 2008-03-27 at 10:35 -0500, Les Mikesell wrote:
> > Where do you find the way to swap the left/middle actions?  And what 
> > works best with a trackpad?
>  
>  Are you being purposefully blunt?  Using the shift key to modify your
>  double click is the default way, middle click is the alternative. 
No, I want the application to do things my way, not be forced to use 2 
hands for my most common action.

>  To close the current folder while opening the new one, hold down
> Shift
>  when double-clicking, or press Shift+Alt+down arrow. 
That's pure insanity.  I don't want to do that in a file browser any 
more than I would in a web browser.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Stephen John Smoogen]

2008/3/27 Jesse Keating <jkeating(a)redhat.com&gt;:
>  On Thu, 2008-03-27 at 10:35 -0500, Les Mikesell wrote:
>   >
>   > Where do you find the way to swap the left/middle actions?  And what
>   > works best with a trackpad?
>
>   Are you being purposefully blunt?  Using the shift key to modify your
>   double click is the default way, middle click is the alternative.
>
>   To close the current folder while opening the new one, hold down Shift
>   when double-clicking, or press Shift+Alt+down arrow.
>  
Hehehe.. could we add in a Meta key action so I can do Alt-X and get a
menu of all the low level commands too? Oh wait you are being serious,
I thought we were talking about emacs.. [And god, I am agreeing with
Mr Mikesell of all people.]

I want something that a 7 year old can deal with. My kid plays on his
OS-X and doesn't have any problems with Windows.. but I am having to
come in and close windows that keep popping up because Fedora doesn't
act in a way that he is used to.



-- 
Stephen J Smoogen. -- CSIRT/Linux System Administrator
How far that little candle throws his beams! So shines a good deed
in a naughty world. = Shakespeare. "The Merchant of Venice"

[Les Mikesell]

Stephen John Smoogen wrote:
>  [And god, I am agreeing with
>  Mr Mikesell of all people.]
>  
Hey, just because I'm not a rabid GPL fanatic doesn't mean I don't know

what is usable.

-- 
   Les Mikesell
     lesmikesell(a)gmail.com

[Tomasz Torcz]

Dnia 27-03-2008, czw o godzinie 10:35 -0500, Les Mikesell pisze:
>  Brian Pepple wrote:
>  > On Thu, 2008-03-27 at 11:16 -0400, Dimi Paun wrote:
>  >> On Thu, 2008-03-27 at 15:00 +0000, Jonathan Underwood wrote:
>  >>> Personally speaking, the whole spatial nautilus thing became a lot
>  >>> more usable when I realized what the middle mouse button click does.
>  >> I understand. Most people will not "discover" it however (I've
> been
>  >> using Linux for 10+ years and I didn't know), and will be frustrated.
>  > 
>  > Umm, the default method *is* in the help file.
>  > 
>  
>  Where do you find the way to swap the left/middle actions?  And what 
>  works best with a trackpad? 
 Middle click works with middle trackpad button. If you want
tap-to-click, you need to have synaptic touchpad with multi finger tap
enabled. Then you can tap with two fingers to get middle click (and with
three fingers to get right click).

-- 
Tomasz Torcz

[Stephen John Smoogen]

2008/3/27 Brian Pepple <bpepple(a)fedoraproject.org&gt;:
>  On Thu, 2008-03-27 at 11:16 -0400, Dimi Paun wrote:
>   > On Thu, 2008-03-27 at 15:00 +0000, Jonathan Underwood wrote:
>   > > Personally speaking, the whole spatial nautilus thing became a lot
>   > > more usable when I realized what the middle mouse button click does.
>   >
>   > I understand. Most people will not "discover" it however (I've been
>   > using Linux for 10+ years and I didn't know), and will be frustrated.
>
>   Umm, the default method *is* in the help file.
>  
Hmmm middle button doesnt seem to do anything for me... but it could
be due to some other setting while I tried to learn the latest Gnome
decisions. Personally I can't get my brain to grok spatial.. it always
seems to do what I don't want to happen when I am looking for
something. I think the issue is that like most GUI issues.. some
people's brains like it and some people don't. At this point, I am
going to try switching to KDE for 9 because it can't be any worse.


-- 
Stephen J Smoogen. -- CSIRT/Linux System Administrator
How far that little candle throws his beams! So shines a good deed
in a naughty world. = Shakespeare. "The Merchant of Venice"

[Bill Crawford]

On 27/03/2008, Stephen John Smoogen <smooge(a)gmail.com&gt; wrote:

>  Hmmm middle button doesnt seem to do anything for me... but it could
>   be due to some other setting while I tried to learn the latest Gnome
>   decisions. Personally I can't get my brain to grok spatial.. it always
>   seems to do what I don't want to happen when I am looking for
>   something. I think the issue is that like most GUI issues.. some
>   people's brains like it and some people don't. At this point, I am
>   going to try switching to KDE for 9 because it can't be any worse. 
Wrong time to do so (in the middle of the switch to KDE4) but it is
better for me, not sure about others. Main thing for me is the "single
click" option (I hate hate hate having to double-click everything in
GNOME) to activate / open / whatever things. Although it's not perfect
(you then have to drag to select things. Oh well :o)

[Michal Schmidt]

On Thu, 27 Mar 2008 17:07:26 +0000
"Bill Crawford" <billcrawford1970(a)gmail.com&gt; wrote:

>  On 27/03/2008, Stephen John Smoogen <smooge(a)gmail.com&gt; wrote:
>  
>  > Hmmm middle button doesnt seem to do anything for me... but it could
>  >  be due to some other setting while I tried to learn the latest
>  > Gnome decisions. Personally I can't get my brain to grok spatial..
>  > it always seems to do what I don't want to happen when I am looking
>  > for something. I think the issue is that like most GUI issues.. some
>  >  people's brains like it and some people don't. At this point, I am
>  >  going to try switching to KDE for 9 because it can't be any worse.
>  
>  Wrong time to do so (in the middle of the switch to KDE4) but it is
>  better for me, not sure about others. Main thing for me is the "single
>  click" option (I hate hate hate having to double-click everything in
>  GNOME) to activate / open / whatever things. Although it's not perfect
>  (you then have to drag to select things. Oh well :o) 
You can have single-click behaviour in Gnome too.
In GConf set /apps/nautilus/preferences/click_policy to "single".

Michal

[Brian Pepple]

On Thu, 2008-03-27 at 18:16 +0100, Michal Schmidt wrote:
>  On Thu, 27 Mar 2008 17:07:26 +0000
>  "Bill Crawford" <billcrawford1970(a)gmail.com&gt; wrote:
>  > 
>  > Wrong time to do so (in the middle of the switch to KDE4) but it is
>  > better for me, not sure about others. Main thing for me is the "single
>  > click" option (I hate hate hate having to double-click everything in
>  > GNOME) to activate / open / whatever things. Although it's not perfect
>  > (you then have to drag to select things. Oh well :o)
>  
>  You can have single-click behaviour in Gnome too.
>  In GConf set /apps/nautilus/preferences/click_policy to "single". 
Or select single-click from the preferences dialog.

Later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Jakub 'Livio' Rusinek]

Dnia 27-03-2008, czw o godzinie 15:00 +0000, Jonathan Underwood pisze:
>  On 27/03/2008, Dimi Paun <dimi(a)lattica.com&gt; wrote:
>  >
>  >  On Thu, 2008-03-27 at 09:17 -0400, Jesse Keating wrote:
>  >  > I have "real" filesystem stuff, I just make proper use of when to
>  >  > double click and when to shift-double click.
>  >
>  >  And you do that because the spacial concept (while interesting), is
>  >  broken as a UI metaphor for folders.
>  >
>  >  Spacial makes a lot of sense for "objects": computer, Network, a disk.
>  >  Those have real-world corespondents, and people relate to that. However,
>  >  for most people, folders are seen as a _path_ to the object that you
>  >  want to work with. (duh, we even call it a "path" ...)
>  >
>  >  Any sane engineering/UI design/consistency principle that was listed in
>  >  this discussion suggests that going with a non-spacial default is the
>  >  only reasonable thing to do. Not only would we please a lot more users,
>  >  we wouldn't be violating the principle of least surprise, but from the
>  >  use cases listed, it seems that even the spacial-proponents would be
>  >  just as happy.
>  >
>  >  Folks, this is not a corner case -- it's the first thing new users bang
>  >  their heads on. And it creates a bad impression. It's an interesting
>  >  idea, it had 4 years to prove itself, it failed. Listing all sorts of
>  >  theoretical reasons why it _should_ be better is just mental
>  >  masturbation.
>  
>  Personally speaking, the whole spatial nautilus thing became a lot
>  more usable when I realized what the middle mouse button click does. 
O! Nice!

Thanks for the tip :) .

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Matthew Saltzman]

On Thu, 2008-03-27 at 10:45 -0400, Dimi Paun wrote:

>  If you think people like, let's run a poll on Slashdot... 
Oh, yeah.  *That*'s scientific!


-- 
                Matthew Saltzman

Clemson University Math Sciences
mjs AT clemson DOT edu
http://www.math.clemson.edu/~mjs

[James Antill]

On Thu, 2008-03-27 at 10:45 -0400, Dimi Paun wrote:

>  Any sane engineering/UI design/consistency principle that was listed
> in
>  this discussion suggests that going with a non-spacial default is the
>  only reasonable thing to do. 
 Please stop trying to confuse everyone, and asking for the same thing
again and again and again while adding no new information to the
thread...
 This thread has had at least 4 different tracks:

1. Fedora localization spins shouldn't change preferences and pass it
off as a localization change.

2. Without overwhelming support, Fedora doesn't diverge from upstream
esp. when it comes to configuration options that have had lots of
discussion upstream.

3. "Discussion" about various people's (who happen to be on f-d-l)
preferences of spacial vs. browser mode in nautilus.

4. Discussion about trying a "close old window, when opening new window
by default" option in spacial nautilus.


...from what I can see you keep arguing that your POV on #3 is the "one
true way" and so we should just ignore all the other points, and
everyone else's opinion, and do what you want.

-- 
James Antill <james(a)fedoraproject.com&gt;
Fedora

[Rodd Clarkson]

On Thu, 2008-03-27 at 08:11 -0500, Les Mikesell wrote:
>  Rodd Clarkson wrote:
>  > 
>  >>>> Thats because Apple has chosen not to make it default behavior, not
> because
>  >>>> it is not included.
>  >>> Interesting... "Apple does it that way" was one of the reasons
> given for the 
>  >>> default in GNOME. It looks like Apple realized it's a bad idea. M$ went
> through 
>  >>> that long ago, Window$ 95 worked that way, they added browser mode and
> changed 
>  >>> the default to it in Window$ 98. Why does GNOME have to copy ancient design
>
>  >>> flaws which both M$ and Apple recognized as such?
>  >> Don't forget to mention users' opinion. They do not like spatial mode
>  >> because of many windows, but it's noticeably faster than browser mode.
>  > 
>  > I like spacial browsing, so maybe I'm strange. ;-]
>  
>  If you have very shallow directory trees and don't attempt system 
>  administration or navigating network shares it is sort-of tolerable. 
>  But what good is a file manager that makes it too inconvenient to use 
>  for real work? 
I guess it depends on how you approach it.

I do a lot of sysadmin using the terminal.

But when I need some graphical representation (and you do from time to
time) then a little CTRL-L magic helps skip all the windows.


R.
-- 
"It's a fine line between denial and faith.
 It's much better on my side"

[Dimi Paun]

On Sun, 2008-03-23 at 17:36 -0800, Jeff Spaleta wrote:
>  And in the case of the functionality which started this gripefest,
> you
>  have that very solution at hand... middle click is exposed via the
>  shift key as previously explained.  
It's not a issue for me, I can get around it. The discussion was about
flipping the *default* setting to make Fedora more useful for newbies.

If we want Fedora be useful to more people out of the box, flipping the
default seems like an obvious choice to me. I don't even understand why
we go about arguing against it, and suggesting that people should change
their mice... :)

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Les Mikesell]

Jeff Spaleta wrote:
>  On Sun, Mar 23, 2008 at 7:13 AM, Les Mikesell
> <lesmikesell(a)gmail.com&gt; wrote:
>  
> > I don't see how anyone with directories more than 2 levels deep or uses
> > it for system file access can tolerate the useless trail of intermediate
> > windows it leaves behind.  Have I missed something basic about how to
> > get rid of them?
> >
>  
>  double-middle-click  will make the parent window close when you open the new
>  folder window.
>  I use double-middle-click religiously. 
I take it you don't regularly bounce among different platforms...

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Dimi Paun]

On Sun, 2008-03-23 at 07:59 -0400, Jesse Keating wrote:
>  I rather like the spatial mode and I keep it that way. 
I wonder if you are not in the (vocal) minority. Of all the 
people that I know directly, none like this option, and ask 
for help to turn it off.

IIRC this interface change was pretty much forced onto users
under the theory that once they get used to it, they'll become
true believers. It seemed reasonable; I tried it for quite a
bit, couldn't get used to it.

I think it's a failed experiment. It had enough time to prove
itself, we should run a poll to figure out if people really
like it enough to be worth keeping it as the default mode.

-- 
Dimi Paun <dimi(a)lattica.com&gt;
Lattica, Inc.

[Henrique Junior]

+1 Dimi
Em Dom, 2008-03-23 às 12:36 -0400, Dimi Paun escreveu:
>  On Sun, 2008-03-23 at 07:59 -0400, Jesse Keating wrote:
>  > I rather like the spatial mode and I keep it that way.
>  
>  I wonder if you are not in the (vocal) minority. Of all the 
>  people that I know directly, none like this option, and ask 
>  for help to turn it off.
>  
>  IIRC this interface change was pretty much forced onto users
>  under the theory that once they get used to it, they'll become
>  true believers. It seemed reasonable; I tried it for quite a
>  bit, couldn't get used to it.
>  
>  I think it's a failed experiment. It had enough time to prove
>  itself, we should run a poll to figure out if people really
>  like it enough to be worth keeping it as the default mode.
>  
>  -- 
>  Dimi Paun <dimi(a)lattica.com&gt;
>  Lattica, Inc.
>   
-- 
Henrique "LonelySpooky" Junior <henriquecsj(a)gmail.com&gt;
Projeto Fedora Brasil

[Igor Pires Soares]

Em Dom, 2008-03-23 às 12:36 -0400, Dimi Paun escreveu:
>  On Sun, 2008-03-23 at 07:59 -0400, Jesse Keating wrote:
>  > I rather like the spatial mode and I keep it that way.
>  
>  I wonder if you are not in the (vocal) minority. Of all the 
>  people that I know directly, none like this option, and ask 
>  for help to turn it off.
>  
>  IIRC this interface change was pretty much forced onto users
>  under the theory that once they get used to it, they'll become
>  true believers. It seemed reasonable; I tried it for quite a
>  bit, couldn't get used to it.
>  
>  I think it's a failed experiment. It had enough time to prove
>  itself, we should run a poll to figure out if people really
>  like it enough to be worth keeping it as the default mode. 
+1

Democracy is the base of freedom.

[Matej Cepl]

On 2008-03-23, 18:58 GMT, Igor Pires Soares wrote:
> > I think it's a failed experiment. It had enough time to
> prove
> > itself, we should run a poll to figure out if people really
> > like it enough to be worth keeping it as the default mode.
>
>  +1
>
>  Democracy is the base of freedom. 
-1

Democracy is root of all tyranny ;-). (BTW, it is totally 
off-topic, but I mean it).

Matej

[Ignacio Vazquez-Abrams]

On Sun, 2008-03-23 at 01:55 -0300, Henrique Junior wrote:
>  who really the uses the spatial mode? 
I do. I like knowing where the windows will pop up next.

-- 
Ignacio Vazquez-Abrams <ivazqueznet(a)gmail.com&gt;

PLEASE don't CC me; I'm already subscribed

[Brian Pepple]

On Sun, 2008-03-23 at 01:55 -0300, Henrique Junior wrote:
>  In all these years of Fedora I (really) never seen an user keeping
> the
>  spatial mode and I want to ask to everyone here: who really the uses
>  the spatial mode? 
I do, and have for years.

Later,
/B
-- 
Brian Pepple <bpepple(a)fedoraproject.org&gt;

http://fedoraproject.org/wiki/BrianPepple
gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

[Igor Pires Soares]

Em Sáb, 2008-03-22 às 21:21 -0700, Andrew Farris escreveu:
>  Igor Pires Soares wrote:
>  > I really understand and agree why we should not include proprietary and
>  > patented stuff in our spins, but I don't think changing the nautilus
>  > window behavior is bad and I think some people is making a big deal of
>  > this.
>  
>  I'm curious whether you've had a real discussion/vote/request from a large group
>
>  of potential users to do this, or whether it is the personal preference of the 
>  few who are planning to do the spin?  While its a small change to make, and you 
>  could say others are making a big deal out of it... it is only one checkbox 
>  easily accessible to every user in mere seconds after they login.  I don't see 
>  the point of anyone (on either side of the preference) making a big deal out of it.
Yes, we had a discussion about that in our local list and nobody was
against changing nautilus behavior. All the users I told about this were
glad to have this thing done by default.

I really don't think it is a big deal. If it is necessary to drop it, we
will. But I don't see the point to have a local spin if we cannot
customize it to fit the local culture. Ambassadors work close to the
users and they know what users are asking for. If many people is asking
for that on our lectures, there might be something we can do for them.

[Kevin Kofler]

Diego Escobar <dezone24 <at> gmail.com> writes:
>  By other side, I dont know why, but Yumex works faster than Pirut
> and
>  provides a face similar to Synaptic. 
If Synaptic is what you want, then that's what you should ship. Are you aware 
that Synaptic is available in the Fedora repository? I use it all the time.

        Kevin Kofler

[Seth Vidal]

On Sat, 2008-03-22 at 13:42 -0500, Diego Escobar wrote:

>   
>       1. Yum is working good... but if it could be faster...(Especially
>          solving dependencies) nobody will unhappy.  

How about 2x-3x faster for depsolving and 1/3rd as much memory used? Is
that better? If so, it's already done in rawhide.


-sv

[Paul W. Frields]

On Sun, 2008-03-23 at 11:41 -0400, seth vidal wrote:
>  On Sat, 2008-03-22 at 13:42 -0500, Diego Escobar wrote:
>  
>  >  
>  >      1. Yum is working good... but if it could be faster...(Especially
>  >         solving dependencies) nobody will unhappy. 
>  
>  
>  How about 2x-3x faster for depsolving and 1/3rd as much memory used? Is
>  that better? If so, it's already done in rawhide. 
And very noticeably so.  Just wanted to say "great work" to the yum
crew. :-)

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
  irc.freenode.net: stickster @ #fedora-docs, #fedora-devel, #fredlug

[Denis Leroy]

Paul W. Frields wrote:
>  On Sun, 2008-03-23 at 11:41 -0400, seth vidal wrote:
> > On Sat, 2008-03-22 at 13:42 -0500, Diego Escobar wrote:
> >
> >>  
> >>      1. Yum is working good... but if it could be faster...(Especially
> >>         solving dependencies) nobody will unhappy. 
> >
> > How about 2x-3x faster for depsolving and 1/3rd as much memory used? Is
> > that better? If so, it's already done in rawhide.
>  
>  And very noticeably so.  Just wanted to say "great work" to the yum
>  crew. :-) 
+1, noticed it also.

[Daniel P. Berrange]

On Sun, Mar 23, 2008 at 11:41:59AM -0400, seth vidal wrote:
>  On Sat, 2008-03-22 at 13:42 -0500, Diego Escobar wrote:
>  
>  >  
>  >      1. Yum is working good... but if it could be faster...(Especially
>  >         solving dependencies) nobody will unhappy. 
>  
>  
>  How about 2x-3x faster for depsolving and 1/3rd as much memory used? Is
>  that better? If so, it's already done in rawhide. 
Of course there's a bit of chicken & egg problem here. Users won't notice
the speed/memusage improvements during the big update from F8 to F9 (if 
doing a live in-place upgrade) because that will be done using the F8 
version of YUM. One of the interesting features of the RHN client was
that the first thing it did was look to see if there was an update of 
itself, apply that and then re-exec itself to do the rest of the upgrade.
So the bulk of work for any package upgrade was always done with latest
update client.

Dan.
-- 
|: Red Hat, Engineering, Boston   -o-   http://people.redhat.com/berrange/ :|
|: http://libvirt.org  -o-  http://virt-manager.org  -o-  http://ovirt.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505  -o-  F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

[Leszek Matok]

Dnia 2008-03-23, o godz. 16:21:24 "Daniel P. Berrange"
<berrange(a)redhat.com&gt;
napisał(a):

>  One of the interesting features of the RHN client was
>  that the first thing it did was look to see if there was an update of 
>  itself, apply that and then re-exec itself to do the rest of the upgrade. 
On the
other hand,
1. I always update rpm, yum and apt first, then the rest of the distro, so if
someone is doing this with two passes, the second pass will benefit from the
changes,
2. first pass always pulls in lots of bits (like libgcc, new Python and
anything that brings with itself).

So:
1. I guess most people upgrade yum itself first, anyhow, no need to do it
automatically,
2. it still doesn't help, because the new version has to be downloaded with
some deps using the old version :)

Lam

[Les Mikesell]

Daniel P. Berrange wrote:
>  On Sun, Mar 23, 2008 at 11:41:59AM -0400, seth vidal wrote:
> > On Sat, 2008-03-22 at 13:42 -0500, Diego Escobar wrote:
> >
> >>  
> >>      1. Yum is working good... but if it could be faster...(Especially
> >>         solving dependencies) nobody will unhappy. 
> >
> > How about 2x-3x faster for depsolving and 1/3rd as much memory used? Is
> > that better? If so, it's already done in rawhide.
>  
>  Of course there's a bit of chicken & egg problem here. Users won't notice
>  the speed/memusage improvements during the big update from F8 to F9 (if 
>  doing a live in-place upgrade) because that will be done using the F8 
>  version of YUM. One of the interesting features of the RHN client was
>  that the first thing it did was look to see if there was an update of 
>  itself, apply that and then re-exec itself to do the rest of the upgrade.
>  So the bulk of work for any package upgrade was always done with latest
>  update client. 
I usually try to update rpm and yum specifically before any other large 
update - mostly from painful past experiences where things did not go 
well when they updated themselves or their libraries in the wrong order. 
  It would be nice if they were smart enough to do this automatically.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

[Seth Vidal]

On Sun, 2008-03-23 at 16:21 +0000, Daniel P. Berrange wrote:
>  On Sun, Mar 23, 2008 at 11:41:59AM -0400, seth vidal wrote:
>  Of course there's a bit of chicken & egg problem here. Users won't notice
>  the speed/memusage improvements during the big update from F8 to F9 (if 
>  doing a live in-place upgrade) because that will be done using the F8 
>  version of YUM. One of the interesting features of the RHN client was
>  that the first thing it did was look to see if there was an update of 
>  itself, apply that and then re-exec itself to do the rest of the upgrade.
>  So the bulk of work for any package upgrade was always done with latest
>  update client. 
So you think we should push yum 3.2.13 (or so) back to F8?

-sv

[Daniel P. Berrange]

On Mon, Mar 24, 2008 at 11:07:24AM -0400, seth vidal wrote:
>  On Sun, 2008-03-23 at 16:21 +0000, Daniel P. Berrange wrote:
>  > On Sun, Mar 23, 2008 at 11:41:59AM -0400, seth vidal wrote:
>  > Of course there's a bit of chicken & egg problem here. Users won't
> notice
>  > the speed/memusage improvements during the big update from F8 to F9 (if 
>  > doing a live in-place upgrade) because that will be done using the F8 
>  > version of YUM. One of the interesting features of the RHN client was
>  > that the first thing it did was look to see if there was an update of 
>  > itself, apply that and then re-exec itself to do the rest of the upgrade.
>  > So the bulk of work for any package upgrade was always done with latest
>  > update client.
>  
>  So you think we should push yum 3.2.13 (or so) back to F8? 
I'm not really in a position to make that judgement call. Depends how stable
you think the new code is - if the risk of regression is small enough, then
I'm sure the speed improvements would be welcomed.

Dan.
-- 
|: Red Hat, Engineering, Boston   -o-   http://people.redhat.com/berrange/ :|
|: http://libvirt.org  -o-  http://virt-manager.org  -o-  http://ovirt.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505  -o-  F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

[Seth Vidal]

On Mon, 2008-03-24 at 15:15 +0000, Daniel P. Berrange wrote:
>  On Mon, Mar 24, 2008 at 11:07:24AM -0400, seth vidal wrote:
>  > On Sun, 2008-03-23 at 16:21 +0000, Daniel P. Berrange wrote:
>  > > On Sun, Mar 23, 2008 at 11:41:59AM -0400, seth vidal wrote:
>  > > Of course there's a bit of chicken & egg problem here. Users won't
> notice
>  > > the speed/memusage improvements during the big update from F8 to F9 (if 
>  > > doing a live in-place upgrade) because that will be done using the F8 
>  > > version of YUM. One of the interesting features of the RHN client was
>  > > that the first thing it did was look to see if there was an update of 
>  > > itself, apply that and then re-exec itself to do the rest of the upgrade.
>  > > So the bulk of work for any package upgrade was always done with latest
>  > > update client.
>  > 
>  > So you think we should push yum 3.2.13 (or so) back to F8?
>  
>  I'm not really in a position to make that judgement call. Depends how stable
>  you think the new code is - if the risk of regression is small enough, then
>  I'm sure the speed improvements would be welcomed. 
the biggest hang up is the obsoletes handling. We've fixed all the
unversioned obsoletes and provides in rawhide that I know of - but we
may not have fixed them in F8.

-sv

[Andrew Farris]

seth vidal wrote:
>  the biggest hang up is the obsoletes handling. We've fixed all
> the
>  unversioned obsoletes and provides in rawhide that I know of - but we
>  may not have fixed them in F8. 
I think one big reason to push it back to f8 would be the improved skip-broken 
handling for multiple arches.. specifically where one of the packages gets 
skipped but its match in the other architecture is not skipped.  People can run 
into some major problems updating with x64 systems that only have some of the 
matching i386 packages on them.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

[Richard Hughes]

On Mon, 2008-03-24 at 11:21 -0700, Andrew Farris wrote:
>  I think one big reason to push it back to f8 would be the improved
>  skip-broken handling for multiple arches.. specifically where one of
>  the packages gets skipped but its match in the other architecture is
>  not skipped.  People can run into some major problems updating with
>  x64 systems that only have some of the matching i386 packages on them. 
I've been using yum from git head on F8 for a few weeks now without
problems.

Richard.

[James Antill]

On Mon, 2008-03-24 at 18:33 +0000, Richard Hughes wrote:
>  On Mon, 2008-03-24 at 11:21 -0700, Andrew Farris wrote:
>  > I think one big reason to push it back to f8 would be the improved
>  > skip-broken handling for multiple arches.. specifically where one of
>  > the packages gets skipped but its match in the other architecture is
>  > not skipped.  People can run into some major problems updating with
>  > x64 systems that only have some of the matching i386 packages on them.
>  
>  I've been using yum from git head on F8 for a few weeks now without
>  problems. 
 Yeh, well I've been running git HEAD on Fed-8 for like 6 months ... and
for the vast majority of the time it rarely shows problems in normal
usage.

 Not that I'm against putting 3.2.13 into Fed-8 at some point, but it
might be nice to have it in rawhide for a few weeks first :)
 I mean the diff stats from 3.2.8 to 3.2.13 are:

 63 files changed, 26159 insertions(+), 3731 deletions(-)

...now admittedly about 20,000 lines of additions there are translations
for the new i18n support ... but we are also seeing bugs on that
support. Fed-8 _is_ supposed to be a stable version.

 Plus it's _very_ easy to upgrade to the yum version in rawhide on
Fed-8[1], and the two things I see people asking for in this thread are:

1. goes faster.
2. skip broken handling.

...neither of which seem worth any significant risk of putting a newer
yum into Fed-8 quickly (in fact the later problem should just not be
happening in Fed-8, IMNSHO).


[1] yum install pygpgme && yum --enablerepo=development update yum\*

    Anyone should feel free to promote this for the people who want to
experience the future now, with all the caveats that implies.

-- 
James Antill <james(a)fedoraproject.com&gt;
Fedora

[Andrew Farris]

James Antill wrote:
>      Anyone should feel free to promote this for the people who want
> to
>  experience the future now, with all the caveats that implies. 
I was mostly thinking about the upgrade path if people don't do yum prior to 
running full upgrade transactions to get from F8->F9, and for that the update 
getting into F8 yum wouldn't be needed for awhile still.  There are still some 
oddities with skip-broken for x86_64 and i386 packages its just doesn't show up 
in the repo very often to keep testing it.

-- 
Andrew Farris <lordmorgul(a)gmail.com&gt; www.lordmorgul.net
  gpg 0xC99B1DF3 fingerprint CDEC 6FAD BA27 40DF 707E A2E0 F0F6 E622 C99B 1DF3
No one now has, and no one will ever again get, the big picture. - Daniel Geer
----                                                                       ----

yum & slow internet

[Development discussions related to Fedora]

Diego Escobar wrote:
>  First: In Latin America most part of people does´t have a fast
> Internet
>  conection, and for this reason is difficult to install new software by YUM. 
Could
you detail "difficult" ?
- does not work at all ?
- getting repodata is ok, but downloads fail ?
- how long would say a single kernel update take to do ?

>     - Yumex 
Do you find yumex faster ?
At what part
- listing the available software / updates ?
- downloading the udpates ?
- installing the updates ?

By the way it sounds like a job for the live spin tool. Perhaps you have 
seen some announcements from Rahul about official spins; I understand 
they are quite easy to do...

DaveT.

yum & slow internet

[Jakub 'Livio' Rusinek]

Dnia 22-03-2008, sob o godzinie 16:56 +1100, David Timms pisze:
>  Diego Escobar wrote:
>  > First: In Latin America most part of people does´t have a fast Internet
>  > conection, and for this reason is difficult to install new software by YUM.
>  Could you detail "difficult" ?
>  - does not work at all ?
>  - getting repodata is ok, but downloads fail ?
>  - how long would say a single kernel update take to do ?
>  
>  >    - Yumex
>  Do you find yumex faster ?
>  At what part
>  - listing the available software / updates ?
>  - downloading the udpates ?
>  - installing the updates ?
>  
>  By the way it sounds like a job for the live spin tool. Perhaps you have 
>  seen some announcements from Rahul about official spins; I understand 
>  they are quite easy to do... 
And why Fedora should include Yumex, if

* Fedora had nice interface - Pirut,
* Fedora 9 will include PackageKit?

Diego may be dissatisfied with PackageKit as it doesn't download info
about all updates (when you click an update, info is downloaded, which
takes for me about minute) etc.

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

yum & slow internet

[Arthur Pemberton]

2008/3/22 Jakub 'Livio' Rusinek <jakub.rusinek(a)gmail.com&gt;:

>   * Fedora had nice interface - Pirut, 
Just to go off on a tangent a bit...

I have no proof that I could do any better, but the Pirut interface
isn't what I would call nice. I loathe using it. I have seen
screenshots of Ubuntu's version of Piruit, and at least outside of say
Anaconda, it is much nicer than Piruit UI wise.

-- 
Fedora 7 : sipping some of that moonshine
( www.pembo13.com )

yum & slow internet

[Richard Hughes]

On Sat, 2008-03-22 at 10:29 +0100, Jakub 'Livio' Rusinek wrote:
>  Diego may be dissatisfied with PackageKit as it doesn't download
> info
>  about all updates (when you click an update, info is downloaded, which
>  takes for me about minute) etc. 
Could you open a bug? Thanks.

Richard.

yum & slow internet

[Diego Escobar]

Hi

Yum in Fedora 8 is working slowly getting repodata and solving dependencies.
But Paul W. Frields tolds that we will have a better yum... 2x 3x faster and
less used memory.

Also, Fedora 9 will include package kit, (replacing pirut)... there are no
way... I will hope for Fedora 9.
Finally a comment: I was looking for a option as "download updates
automaticaly and notify to install", somebody told me about a plugin for yum
but I never found it. I don´t know if it will go on Fedora 9 but is a util
option.

Regards

-- 
Diego Escobar
Fedora Ambassador for Colombia
Fedora Release 8 (Warewolf): Linux Kernel 2.6.24 + Gnome 2.20

yum & slow internet

[Seth Vidal]

On Mon, 2008-03-24 at 10:01 -0500, Diego Escobar wrote:
>    
>  Hi
>   
>  Yum in Fedora 8 is working slowly getting repodata and solving
>  dependencies. But Paul W. Frields tolds that we will have a better
>  yum... 2x 3x faster and less used memory. 
>   
>  Also, Fedora 9 will include package kit, (replacing pirut)...
>  there are no way... I will hope for Fedora 9.
>  Finally a comment: I was looking for a option as "download updates
>  automaticaly and notify to install", somebody told me about a plugin
>  for yum but I never found it. I don´t know if it will go on Fedora 9
>  but is a util option. 
>    
yum install yum-downloadonly

that'll install the module you want.

-sv

yum & slow internet

[Les Mikesell]

seth vidal wrote:
> >  
> > Yum in Fedora 8 is working slowly getting repodata and solving
> > dependencies. But Paul W. Frields tolds that we will have a better
> > yum... 2x 3x faster and less used memory. 
> >  
> > Also, Fedora 9 will include package kit, (replacing pirut)...
> > there are no way... I will hope for Fedora 9.
> > Finally a comment: I was looking for a option as "download updates
> > automaticaly and notify to install", somebody told me about a plugin
> > for yum but I never found it. I don´t know if it will go on Fedora 9
> > but is a util option. 
> >  
>  
>  yum install yum-downloadonly
>  
>  that'll install the module you want. 
Personally I think people would be happier if the downloadonly plugin 
would be included in a base install and documentation/examples/help 
showed it as the typical update method.  That is, do the:
yum --downloadonly update
more or less unattended, repeating as necessary if it doesn't complete 
and knowing it won't hurt anything, then babysit the real
yum update
which will happen quickly and may need a reboot and/or some application 
testing afterwards.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

yum & slow internet

[Diego Escobar]

>  Personally I think people would be happier if the downloadonly plugin
>  would be included in a base install and documentation/examples/help
>  showed it as the typical update method.  That is, do the:
>  yum --downloadonly update
>  more or less unattended, repeating as necessary if it doesn't complete
>  and knowing it won't hurt anything, then babysit the real
>  yum update
>  which will happen quickly and may need a reboot and/or some application
>  testing afterwards.
>
>  
+1

Thats the way. Simplify the things for the final user. (Newbie or not).
Maybe this so clear for people who are in the know, but most part of people
not.
All my suggestions here, are for that reason.

I would like a simple option "Download and notification when updates are
ready to install" Should I know that I have download another package? Should
I know yum --downloadonly update? By this way no. Clear for the user.

Thanks.

-- 
Diego Escobar
Fedora Ambassador for Colombia
Fedora Release 8 (Warewolf): Linux Kernel 2.6.24 + Gnome 2.20

yum & slow internet

[James Antill]

On Mon, 2008-03-24 at 10:01 -0500, Diego Escobar wrote:
>    
>  Hi
>   
>  Yum in Fedora 8 is working slowly getting repodata and solving
>  dependencies. But Paul W. Frields tolds that we will have a better
>  yum... 2x 3x faster and less used memory.  
 Note that yum can't make the network faster[1], although it will do
it's local operations faster in newer versions (yum also has no control
over how long rpm takes).

[...]

>  Finally a comment: I was looking for a option as "download
> updates
>  automaticaly and notify to install", somebody told me about a plugin
>  for yum but I never found it. I don´t know if it will go on Fedora 9
>  but is a util option.  
 You probably want "do_update = no" in /etc/yum/yum-updatesd.conf but
"do_download = yes" and "do_download_deps = yes".


[1] But you can change your metadata_expire to be higher (although then
you'll probably want to change your mdpolicy to group:all, in newer
versions).

-- 
James Antill <james(a)fedoraproject.com&gt;
Fedora

yum & slow internet

[Sunil Ghai]

On Mon, Mar 24, 2008 at 11:44 AM, James Antill <james(a)fedoraproject.com&gt;
wrote:

>  On Mon, 2008-03-24 at 10:01 -0500, Diego Escobar wrote:
>  >
>  > Hi
>  >
>  > Yum in Fedora 8 is working slowly getting repodata and solving
>  > dependencies. But Paul W. Frields tolds that we will have a better
>  > yum... 2x 3x faster and less used memory.
>
>   Note that yum can't make the network faster[1], although it will do 


As Diego Escobar has mentioned, many people around the world have slow
internet connection.  How about having a feature (possibly in kernel itself)
to make updates using _idle_ bandwidth only? By idle I mean which is
currently not being used and is free. This would also keep responsiveness in
other network applications. We can use current traffic control tools like tc
to implement this.

This might take some extra time and in no way making the network faster but
user would be able to keep the system updated in best possible way on slow
internet connection.

-- 
Regards,
Sunil Ghai

yum & slow internet

[Les Mikesell]

Sunil Ghai wrote:
>  
>  
>  As Diego Escobar has mentioned, many people around the world have slow 
>  internet connection.  How about having a feature (possibly in kernel 
>  itself) to make updates using _idle_ bandwidth only? By idle I mean 
>  which is currently not being used and is free. 
If you have more than one system on your network, how can your kernel 
possibly know what is free?

>  This would also keep 
>  responsiveness in other network applications. We can use current traffic 
>  control tools like tc to implement this. 
The traffic is mostly coming the other direction.  You can deliberately 
slow down the ack's to slow it down but you can't really tell how much 
is appropriate.

>  This might take some extra time and in no way making the network
> faster 
>  but user would be able to keep the system updated in best possible way 
>  on slow internet connection. 
You don't need to update continuously. I think a better approach is to 
explicitly download at a time you are taking a break from other 
activity.  Go get some coffee, have lunch, etc. and come back with the 
download part completed.

-- 
   Les Mikesell
    lesmikesell(a)gmail.com

yum & slow internet

[Benny Amorsen]

Les Mikesell <lesmikesell(a)gmail.com&gt; writes:

>  If you have more than one system on your network, how can your
> kernel
>  possibly know what is free? 
The kernel can back off more aggressively than TCP specifies. It
doesn't take much backing off before the other flows pretty much steal
all the bandwidth. TCP can be quite unfair if routers don't use RED
and/or some kind of fair queueing.


/Benny

[Kevin Kofler]

Diego Escobar <dezone24 <at> gmail.com> writes:
>  With this ideas my suggestion is to include many software by
> default on
>  Fedora (Especially in Fedora LiveCD) 
I'm afraid the live CD spins in particular are already very space-constrained, 
there's simply no room to add more software onto them! Otherwise, we'd have to 
require a DVD or a special overlong CD, and unfortunately that tends to be a 
problem in the exact same countries where Internet connectivity is a problem.

>  OpenOffice 
Sorry, there's no way this is going to fit onto any of our live CD spins, ever. 
OO.o needs a CD on its own!

>  Linux Terminal Server (tsclient) 
>  Vino
>  Yumex 
While those (and amsn which you added in your followup) aren't size hogs by 
themselves, small packages add up, so our live CDs have to make a choice 
somewhere. There's simply no room left for those, and some of them are 
redundant with tools which are on the live CDs (Yumex does the same thing as 
Pirut/Pup in F8 and PackageKit in F9, amsn's functionality is covered by Pidgin 
on the GNOME Live spin or Kopete on the KDE Live spin, which both also handle 
several other protocols).

Note that the KDE Live spin does include a VNC and RDP client (Krdc, which is 
part of kdenetwork).

>  Fedora Frog (Easy front-end) 
This one can't be included for legal reasons.

>  Nautilus (always use browser by default) 
This is a default settings issue, it has nothing to do with download sizes.

>  Livna
>  Freshrpm 
Same remark as for Fedora Frog.

And the livna-release package isn't that much to download anyway! The packages 
are, but they all aren't included in Fedora for a reason, so they can't be 
added.

        Kevin Kofler

[Jakub 'Livio' Rusinek]

> > Nautilus (always use browser by default)
>  
>  This is a default settings issue, it has nothing to do with download sizes. 
We can file bug upstream and vote, but GNOME doesn't seem to like
changes like that.

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Jakub 'Livio' Rusinek]

Dnia 21-03-2008, pią o godzinie 23:05 -0500, Diego Escobar pisze:
>   
>  Hi:
>   
>  I have a suggestion about the default software for Fedora:
>   
>  First: In Latin America most part of people does´t have a fast
>  Internet conection, and for this reason is difficult to install new
>  software by YUM. 
>  Second: An important point that experts evaluate on a S.O. is "ready
>  out of the box" and you supose how many changes you have to do to
>  leave Fedora ready.
>   
>  With this ideas my suggestion is to include many software by
>  default on Fedora (Especially in Fedora LiveCD)
>   
>        * OpenOffice
>        * Linux Terminal Server (tsclient) 
>        * Vino
>        * Yumex
>        * Fedora Frog (Easy front-end)
>        * Nautilus (always use browser by default) 
* I disagree with OO.o inclusion. It's too big and ugly. Fedora provides
AbiWord and it is nice. Also, Gnumeric is by default.

* tsclient? There's no reason to put it, until majority of cumminity
will want it.

* Vino? Is it the video editor? Or VNC client? Also, I see no reason.

* Fedora Frog? It is unofficial.

* Browser mode by default? Go upstream, if this is very hard to change
one option

>  And why not? we can do easy including by default the most common
>  repositories for Fedora:
>   
>        * Livna
>        * Freshrpm
>  I hope my suggestion will be taken into account. Remember, this idea
>  is designed to do Fedora more competitive and more easy for the final
>  user. 
Livna, FreshRPMs and Dribble joined RPM Fusion, but still they provide
packages with non-free licences.

They simply cannot be just like that available. Livna is not official.

I only think that Livna could e approved and then used by Codeina, but
this is only in my imagine.

How it is currently, is good.

-- 
Jakub 'Livio' Rusinek
http://liviopl.jogger.pl/

[Yuan Yijun]

2008/3/22, Jakub 'Livio' Rusinek <jakub.rusinek(a)gmail.com&gt;:
>   * tsclient? There's no reason to put it, until majority of cumminity
>   will want it.
>
>   * Vino? Is it the video editor? Or VNC client? Also, I see no reason.
>  

Having tsclient is valid point, as I want putty on every windows
installation. If you do not know Vino then you may not use tsclient or
mstsc either.

I havn't check the RPM list of liveCD, maybe rdesktop and vnc client
are already included?



-- 
bbbush ^_^