3:14 p.m.
Dimitris:
We just got a last-minute idea from Max Spevack et al, to put a hit
tracker on the splash page. A box in the upper corner/somewhere like
below[1].
I know this is going to be untranslated, but we can catch it in an
update to the package, and many people won't get a translated page,
anyway. The value this gives the whole project is very worth the
effort.
[1] Please add the following:
######### Begin Box #################################################
<div class="tracking-spot">
<p>
Hey, Fedora users! We are trying to figure out how many desktop users
we have. We are using a simple image to count the number of hits we get
from unique IP addresses. We are <b>not</b> tracking any identifiable
information.
</p>
<p>
This image right here:
</p>
<img src="http://fedoraproject.org/fc6-desktop-tracker.png" alt="blue
tracking dot" width="1" height="1"/>
<p>
...is being used for that purpose. If you see the image, that means
you are online, and you have been counted. If you do not see the image,
it is because you are offline, and the image cannot load.
</p>
<p>
Why is this important? Because we need to know how many users we have,
which helps help us fund innovations to Fedora.
</p>
</div>
######### End Box ####################################################
How does that sound? Any alteration suggestions?
Thanks - Karsten
--
Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project
Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject
quaid.108.redhat.com | gpg key: AD0E0C41
////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
![src="http://fedoraproject.org/fc6-desktop-tracker.png"](http://src="http://fedoraproject.org/fc6-desktop-tracker.png%22){kind=link}
3:19 p.m.
Karsten Wade wrote:
Dimitris:
We just got a last-minute idea from Max Spevack et al, to put a hit
tracker on the splash page. A box in the upper corner/somewhere like
below[1].
I know this is going to be untranslated, but we can catch it in an
update to the package, and many people won't get a translated page,
anyway. The value this gives the whole project is very worth the
effort.
[1] Please add the following:
######### Begin Box #################################################
<div class="tracking-spot">
<p>
Hey, Fedora users! We are trying to figure out how many desktop users
we have. We are using a simple image to count the number of hits we get
from unique IP addresses. We are <b>not</b> tracking any identifiable
information.
</p>
<p>
This image right here:
</p>
<img src="http://fedoraproject.org/fc6-desktop-tracker.png" alt="blue
tracking dot" width="1" height="1"/>
<p>
...is being used for that purpose. If you see the image, that means
you are online, and you have been counted. If you do not see the image,
it is because you are offline, and the image cannot load.
</p>
<p>
Why is this important? Because we need to know how many users we have,
which helps help us fund innovations to Fedora.
</p>
</div>
######### End Box ####################################################
How does that sound? Any alteration suggestions?
Max should probably announce it elsewhere like fedora-devel list and
discuss this for transparency. I dont see any harm in doing it but it's
important that we dont do this discretely. Moreover the note on
innovations in not very clear to me. It would be better to make our
intentions more explicit.
Rahul
4:21 p.m.
On Fri, 29 Sep 2006, Rahul Sundaram wrote:
> Why is this important? Because we need to know how many users we
have,
> which helps help us fund innovations to Fedora. </p>
> </div>
> ######### End Box ####################################################
change it to "Because we need to know more about what *types* of users we
have, so that we can have a better idea of what sorts of innovations will
be interesting to those users."
> Max should probably announce it elsewhere like fedora-devel list
and
> discuss this for transparency. I dont see any harm in doing it but it's
> important that we dont do this discretely. Moreover the note on
> innovations in not very clear to me. It would be better to make our
> intentions more explicit.
I'm not announcing anything until I know that it's actually in. Once it
is *in*, I'll be happy to send something out about it.
--Max
4:24 p.m.
Max Spevack wrote:
On Fri, 29 Sep 2006, Rahul Sundaram wrote:
>> Why is this important? Because we need to know how many users we
>> have, which helps help us fund innovations to Fedora. </p>
>> </div>
>> ######### End Box ####################################################
change it to "Because we need to know more about what *types* of users
we have, so that we can have a better idea of what sorts of innovations
will be interesting to those users."
How will a hit counter tell us what types of users they are?
Rahul
4:53 p.m.
On Fri, 2006-09-29 at 02:54 +0530, Rahul Sundaram wrote:
Max Spevack wrote:
> On Fri, 29 Sep 2006, Rahul Sundaram wrote:
>
>>> Why is this important? Because we need to know how many users we
>>> have, which helps help us fund innovations to Fedora. </p>
>>> </div>
>>> ######### End Box ####################################################
>
>
> change it to "Because we need to know more about what *types* of users
> we have, so that we can have a better idea of what sorts of innovations
> will be interesting to those users."
>
How will a hit counter tell us what types of users they are?
Right, that is now more ambiguous.
What is wrong with 'innovations'? It seems intentionally vague because
there is not a cabal deciding what we should innovate into. How about
just:
"Why is this important. Because we need to know how many users we have
to make better decisions for the project."
etc.
- Karsten
--
Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project
Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject
quaid.108.redhat.com | gpg key: AD0E0C41
////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
4:55 p.m.
Karsten Wade wrote:
Right, that is now more ambiguous.
What is wrong with 'innovations'? It seems intentionally vague because
there is not a cabal deciding what we should innovate into. How about
just:
"Why is this important. Because we need to know how many users we have
to make better decisions for the project."
Ok. That sounds more open. Do it. Announce it.
Rahul
5:15 p.m.
On Fri, 29 Sep 2006, Rahul Sundaram wrote:
How will a hit counter tell us what types of users they are?
It will tell us they're desktop users.
--g
-------------------------------------------------------------
Greg DeKoenigsberg || Fedora Project || fedoraproject.org
Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors
-------------------------------------------------------------
12:17 a.m.
Greg DeKoenigsberg wrote:
On Fri, 29 Sep 2006, Rahul Sundaram wrote:
> How will a hit counter tell us what types of users they are?
It will tell us they're desktop users.
--g
-------------------------------------------------------------
Greg DeKoenigsberg || Fedora Project || fedoraproject.org
Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors
-------------------------------------------------------------
I can not support this idea, IMO it is to late in the release process to
work out all the logistics and potential problems in the time we have. I
feel that the potential problems and community repercussions will far
out weigh the minor benefits from a "half assed" attempt at getting to
know our user base.
If we have a vote in this I say "NO"
I feel that if we would like to gather information about our user base
we should think about the bigger picture and create a survey for the FC7
release. This may allow us to get real information and opinions from the
users with out the risk of offending anyone.
Bob
P.S. My opinions are mine alone and do not reflect the opinions of any
group I may be a part of.
--
Robert 'Bob' Jensen * * http://fedoraproject.org/wiki/BobJensen
gpg fingerprint: F9F4 7243 4243 0043 2C45 97AF E8A4 C3AE 42EB 0BC6
Fedora Docs Projects FDSCo http://fedoraproject.org/wiki/DocsProject
10:05 a.m.
On Sat, 2006-09-30 at 00:17 -0500, Robert 'Bob' Jensen wrote:
Greg DeKoenigsberg wrote:
> On Fri, 29 Sep 2006, Rahul Sundaram wrote:
>
>> How will a hit counter tell us what types of users they are?
>
> It will tell us they're desktop users.
>
> --g
>
> -------------------------------------------------------------
> Greg DeKoenigsberg || Fedora Project || fedoraproject.org
> Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors
> -------------------------------------------------------------
I can not support this idea, IMO it is to late in the release process to
work out all the logistics and potential problems in the time we have. I
feel that the potential problems and community repercussions will far
out weigh the minor benefits from a "half assed" attempt at getting to
know our user base.
If we have a vote in this I say "NO"
I feel that if we would like to gather information about our user base
we should think about the bigger picture and create a survey for the FC7
release. This may allow us to get real information and opinions from the
users with out the risk of offending anyone.
+1 on this. I think this is too rushed and un-thought-out. I don't
really see how this could really be such a benefit, but I do see the
potential for confusion/offense among some users. Personally I wouldn't
care, but objectively speaking I could see how people would think it was
a little questionable.
So, I vote "no" as well. If we start working now, we can have something
much better for FC7, which really isn't far away.
Sam
Bob
P.S. My opinions are mine alone and do not reflect the opinions of any
group I may be a part of.
--
Robert 'Bob' Jensen * * http://fedoraproject.org/wiki/BobJensen
gpg fingerprint: F9F4 7243 4243 0043 2C45 97AF E8A4 C3AE 42EB 0BC6
Fedora Docs Projects FDSCo http://fedoraproject.org/wiki/DocsProject
12:01 p.m.
On Sat, 2006-09-30 at 11:05 -0400, Sam Folk-Williams wrote:
On Sat, 2006-09-30 at 00:17 -0500, Robert 'Bob' Jensen
wrote:
> Greg DeKoenigsberg wrote:
> > On Fri, 29 Sep 2006, Rahul Sundaram wrote:
> >
> >> How will a hit counter tell us what types of users they are?
> >
> > It will tell us they're desktop users.
>
> I can not support this idea, IMO it is to late in the release process to
> work out all the logistics and potential problems in the time we have. I
> feel that the potential problems and community repercussions will far
> out weigh the minor benefits from a "half assed" attempt at getting to
> know our user base.
>
> If we have a vote in this I say "NO"
>
> I feel that if we would like to gather information about our user base
> we should think about the bigger picture and create a survey for the FC7
> release. This may allow us to get real information and opinions from the
> users with out the risk of offending anyone.
>
+1 on this. I think this is too rushed and un-thought-out. I don't
really see how this could really be such a benefit, but I do see the
potential for confusion/offense among some users. Personally I wouldn't
care, but objectively speaking I could see how people would think it was
a little questionable.
So, I vote "no" as well. If we start working now, we can have something
much better for FC7, which really isn't far away.
And in fairness, we can turn out an update to fedora-release-notes
package with this kind of material during the FC6 cycle if no one wants
to wait until FC7.
--
Paul W. Frields, RHCE http://paul.frields.org/
gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717
Fedora Project Board: http://fedoraproject.org/wiki/Board
Fedora Docs Project: http://fedoraproject.org/wiki/DocsProject
9:33 p.m.
O/H Karsten Wade έγραψε:
Dimitris:
We just got a last-minute idea from Max Spevack et al, to put a hit
tracker on the splash page. A box in the upper corner/somewhere like
below[1].
Karsten, Max, et al,
I believe the addition of such a tracker (solution 1) will have a bad impact on
the image of the Fedora Project on our users, which will not be worth the
information gained from this addition.
I'll try to elaborate my thoughts.
Trackers
========
Such trackers were always faced with bad criticism. This takes place with every
equipment used for tracking: emails containing notices like "the sender asked to
be notified when you read this email", cameras in the streets and shops,
security checks in airports, digital IDs, RFIDs, even Credit Cards. People even
feel irritated with phrases like "this conversation will be recorded for
training purposes". They usually don't behave the same, even if they were
guaranteed that the data will be trashed right after the event.
I admit to be quite a bit irritated when I get messages from webpages saying
"Hi. You are 84.9.192, you are using Firefox 1.5 on Linux" even though *I know*
that my browser does expose this information. I know this is silly, but even
though I know it happens, I don't like hearing it. Notifying me that it does
happen makes things worse.
BTW this is not a matter of conspiracy theories: It's just that I like websites
that don't do this or remind me that it is done more than those which do.
The user's desktop is like his digital home. It's private, it's exposed to
dangers from intruders, the user is bound to it in a very special way.
Applications that somehow used the Internet for a reason the company saw fit
were always badly criticized from the users. I believe that's why, even today,
99% of the proprietary applications don't use the connection to the Internet to
check if the software is cracked.
Other solutions
===============
So, the question is: Is it worth it?
Well, my opinion is that it is *not* and the extra information will not be much
more worth than, say, the visits to our website.
But let's say it does.
One approach would be to reduce the fuss about it: Add a 1x1 image somewhere on
the page that comes from the net (solution 2). User won't be told so he won't
have the chance to think our motivations are bad (which of course aren't). This
"secret" approach will also have a bad impact on our image (this time in the
media shouting about user privacy). But: a) some people will think "big deal,
it's just a pixel, the guys want to count their users" and b) at least 95% of
the users won't find out about it, which probably is not bad because we are not
doing anything wrong.
Another line of thinking (better) is that we can count users in other ways, less
intrusive, just like other projects do. And, in addition, some of these ways
provide more information than the tracker one.
For example, we could ask the user to send us a ping during the first boot of
the system (solution 3). Package `firstboot` could say something like:
We would like to count the users of Fedora. You could send us a ping
which
will be more or less like saying to us "Hi guys!". Would you like to send us
a ping then? (Please do!)
["Hi guys!"]
If not, you can click "Next" to continue to the next page.
If we need to count more than systems that re/installed FC6, and, say, count
unique users, we could have that message on top of the splash page (solution 4).
Whatever the users clicks, a cookie stores her choice in a cookie so that she
won't be asked again.
Finally, the most useful and non-intrusive approach (solution 5) is to ask the
user fill a short questionnaire. More helpful data than any other approach, less
suspicious intentions.
Summary of solutions
--------------------
1. Original suggestion: image on splash page with description.
2. Tiny invisible image with no description.
3. Ping from Firstboot.
4. Ping from splash (once).
5. Questionnaire from firstboot or splash.
Accuracy
--------
None of the solutions give has any accuracy on the number of users using Fedora.
Analytically the solutions:
1,2: The image could be put in any webpage and the users with dynamic IPs would
be counted as multiple users.
3: Doesn't count upgrades to FC6 and also gives option of just clicking
"next".
4: Suffers like 1: multiple locations could point there and people who clear
their cookies often will see the message over and over.
5: Like 4, but a bit better because the bogus calls will be fewer.
Conclusion
==========
To summarize a very long email (sorry): I believe users would like it better if
we did not use their desktop to track anything or ask them any such question. I
wasn't convinced that the value it will give the project will be worth the cost
of the image of the project.
I'm sorry but I feel this is a very delicate matter that touches sensitive
chords and fragile nerves. Actions like these sometimes make people feel that
their privacy was disturbed or their personal normal lives intruded for
third-party benefits.
If I had to choose from the solutions, I would go for the most "human"/less
intrusive one, #5 (the questionnaire).
Sorry again for the length of the email.
-dim
I know this is going to be untranslated, but we can catch it in an
update to the package, and many people won't get a translated page,
anyway. The value this gives the whole project is very worth the
effort.
[1] Please add the following:
######### Begin Box #################################################
<div class="tracking-spot">
<p>
Hey, Fedora users! We are trying to figure out how many desktop users
we have. We are using a simple image to count the number of hits we get
from unique IP addresses. We are <b>not</b> tracking any identifiable
information.
</p>
<p>
This image right here:
</p>
<img src="http://fedoraproject.org/fc6-desktop-tracker.png" alt="blue
tracking dot" width="1" height="1"/>
<p>
...is being used for that purpose. If you see the image, that means
you are online, and you have been counted. If you do not see the image,
it is because you are offline, and the image cannot load.
</p>
<p>
Why is this important? Because we need to know how many users we have,
which helps help us fund innovations to Fedora.
</p>
</div>
######### End Box ####################################################
How does that sound? Any alteration suggestions?
Thanks - Karsten
--
Dimitris Glezos
Jabber ID: glezos(a)jabber.org, GPG: 0xA5A04C3B
http://dimitris.glezos.com/
"He who gives up functionality for ease of use
loses both and deserves neither." (Anonymous)
--
10:33 p.m.
since dimitri's email was too long for 6am :) , i've skiped most of it
but, imho:his idea to let the user choose it's great, and, it's very
important to let the user read any time after, what happened when he
clicked, hence, if we put the link(s) on the start page, we can let
him click whenever he wants to ...
On 9/29/06, Dimitris Glezos <dimitris(a)glezos.com> wrote:
O/H Karsten Wade έγραψε:
> Dimitris:
>
> We just got a last-minute idea from Max Spevack et al, to put a hit
> tracker on the splash page. A box in the upper corner/somewhere like
> below[1].
Karsten, Max, et al,
I believe the addition of such a tracker (solution 1) will have a bad impact on
the image of the Fedora Project on our users, which will not be worth the
information gained from this addition.
I'll try to elaborate my thoughts.
Trackers
========
Such trackers were always faced with bad criticism. This takes place with every
equipment used for tracking: emails containing notices like "the sender asked to
be notified when you read this email", cameras in the streets and shops,
security checks in airports, digital IDs, RFIDs, even Credit Cards. People even
feel irritated with phrases like "this conversation will be recorded for
training purposes". They usually don't behave the same, even if they were
guaranteed that the data will be trashed right after the event.
I admit to be quite a bit irritated when I get messages from webpages saying
"Hi. You are 84.9.192, you are using Firefox 1.5 on Linux" even though *I
know*
that my browser does expose this information. I know this is silly, but even
though I know it happens, I don't like hearing it. Notifying me that it does
happen makes things worse.
BTW this is not a matter of conspiracy theories: It's just that I like websites
that don't do this or remind me that it is done more than those which do.
The user's desktop is like his digital home. It's private, it's exposed to
dangers from intruders, the user is bound to it in a very special way.
Applications that somehow used the Internet for a reason the company saw fit
were always badly criticized from the users. I believe that's why, even today,
99% of the proprietary applications don't use the connection to the Internet to
check if the software is cracked.
Other solutions
===============
So, the question is: Is it worth it?
Well, my opinion is that it is *not* and the extra information will not be much
more worth than, say, the visits to our website.
But let's say it does.
One approach would be to reduce the fuss about it: Add a 1x1 image somewhere on
the page that comes from the net (solution 2). User won't be told so he won't
have the chance to think our motivations are bad (which of course aren't). This
"secret" approach will also have a bad impact on our image (this time in the
media shouting about user privacy). But: a) some people will think "big deal,
it's just a pixel, the guys want to count their users" and b) at least 95% of
the users won't find out about it, which probably is not bad because we are not
doing anything wrong.
Another line of thinking (better) is that we can count users in other ways, less
intrusive, just like other projects do. And, in addition, some of these ways
provide more information than the tracker one.
For example, we could ask the user to send us a ping during the first boot of
the system (solution 3). Package `firstboot` could say something like:
> We would like to count the users of Fedora. You could send us a ping which
> will be more or less like saying to us "Hi guys!". Would you like to send
us
> a ping then? (Please do!)
>
> ["Hi guys!"]
>
> If not, you can click "Next" to continue to the next page.
If we need to count more than systems that re/installed FC6, and, say, count
unique users, we could have that message on top of the splash page (solution 4).
Whatever the users clicks, a cookie stores her choice in a cookie so that she
won't be asked again.
Finally, the most useful and non-intrusive approach (solution 5) is to ask the
user fill a short questionnaire. More helpful data than any other approach, less
suspicious intentions.
Summary of solutions
--------------------
1. Original suggestion: image on splash page with description.
2. Tiny invisible image with no description.
3. Ping from Firstboot.
4. Ping from splash (once).
5. Questionnaire from firstboot or splash.
Accuracy
--------
None of the solutions give has any accuracy on the number of users using Fedora.
Analytically the solutions:
1,2: The image could be put in any webpage and the users with dynamic IPs would
be counted as multiple users.
3: Doesn't count upgrades to FC6 and also gives option of just clicking
"next".
4: Suffers like 1: multiple locations could point there and people who clear
their cookies often will see the message over and over.
5: Like 4, but a bit better because the bogus calls will be fewer.
Conclusion
==========
To summarize a very long email (sorry): I believe users would like it better if
we did not use their desktop to track anything or ask them any such question. I
wasn't convinced that the value it will give the project will be worth the cost
of the image of the project.
I'm sorry but I feel this is a very delicate matter that touches sensitive
chords and fragile nerves. Actions like these sometimes make people feel that
their privacy was disturbed or their personal normal lives intruded for
third-party benefits.
If I had to choose from the solutions, I would go for the most "human"/less
intrusive one, #5 (the questionnaire).
Sorry again for the length of the email.
-dim
> I know this is going to be untranslated, but we can catch it in an
> update to the package, and many people won't get a translated page,
> anyway. The value this gives the whole project is very worth the
> effort.
>
> [1] Please add the following:
>
> ######### Begin Box #################################################
> <div class="tracking-spot">
> <p>
> Hey, Fedora users! We are trying to figure out how many desktop users
> we have. We are using a simple image to count the number of hits we get
> from unique IP addresses. We are <b>not</b> tracking any identifiable
> information.
> </p>
> <p>
> This image right here:
> </p>
> <img src="http://fedoraproject.org/fc6-desktop-tracker.png"
alt="blue
> tracking dot" width="1" height="1"/>
> <p>
> ...is being used for that purpose. If you see the image, that means
> you are online, and you have been counted. If you do not see the image,
> it is because you are offline, and the image cannot load.
> </p>
> <p>
> Why is this important? Because we need to know how many users we have,
> which helps help us fund innovations to Fedora.
> </p>
> </div>
> ######### End Box ####################################################
>
> How does that sound? Any alteration suggestions?
>
> Thanks - Karsten
--
Dimitris Glezos
Jabber ID: glezos(a)jabber.org, GPG: 0xA5A04C3B
http://dimitris.glezos.com/
"He who gives up functionality for ease of use
loses both and deserves neither." (Anonymous)
--
--
fedora-docs-list mailing list
fedora-docs-list(a)redhat.com
To unsubscribe:
https://www.redhat.com/mailman/listinfo/fedora-docs-list
12:22 a.m.
Dimitris Glezos wrote:
Karsten, Max, et al,
I believe the addition of such a tracker (solution 1) will have a bad impact on
the image of the Fedora Project on our users, which will not be worth the
information gained from this addition.
I'll try to elaborate my thoughts.
Trackers
========
Such trackers were always faced with bad criticism. This takes place with every
equipment used for tracking: emails containing notices like "the sender asked to
be notified when you read this email", cameras in the streets and shops,
security checks in airports, digital IDs, RFIDs, even Credit Cards. People even
feel irritated with phrases like "this conversation will be recorded for
training purposes". They usually don't behave the same, even if they were
guaranteed that the data will be trashed right after the event.
I admit to be quite a bit irritated when I get messages from webpages saying
"Hi. You are 84.9.192, you are using Firefox 1.5 on Linux" even though *I
know*
that my browser does expose this information. I know this is silly, but even
though I know it happens, I don't like hearing it. Notifying me that it does
happen makes things worse.
BTW this is not a matter of conspiracy theories: It's just that I like websites
that don't do this or remind me that it is done more than those which do.
The user's desktop is like his digital home. It's private, it's exposed to
dangers from intruders, the user is bound to it in a very special way.
Applications that somehow used the Internet for a reason the company saw fit
were always badly criticized from the users. I believe that's why, even today,
99% of the proprietary applications don't use the connection to the Internet to
check if the software is cracked.
Other solutions
===============
So, the question is: Is it worth it?
Well, my opinion is that it is *not* and the extra information will not be much
more worth than, say, the visits to our website.
But let's say it does.
One approach would be to reduce the fuss about it: Add a 1x1 image somewhere on
the page that comes from the net (solution 2). User won't be told so he won't
have the chance to think our motivations are bad (which of course aren't). This
"secret" approach will also have a bad impact on our image (this time in the
media shouting about user privacy). But: a) some people will think "big deal,
it's just a pixel, the guys want to count their users" and b) at least 95% of
the users won't find out about it, which probably is not bad because we are not
doing anything wrong.
Another line of thinking (better) is that we can count users in other ways, less
intrusive, just like other projects do. And, in addition, some of these ways
provide more information than the tracker one.
For example, we could ask the user to send us a ping during the first boot of
the system (solution 3). Package `firstboot` could say something like:
If we need to count more than systems that re/installed FC6, and, say, count
unique users, we could have that message on top of the splash page (solution 4).
Whatever the users clicks, a cookie stores her choice in a cookie so that she
won't be asked again.
Finally, the most useful and non-intrusive approach (solution 5) is to ask the
user fill a short questionnaire. More helpful data than any other approach, less
suspicious intentions.
Summary of solutions
--------------------
1. Original suggestion: image on splash page with description.
2. Tiny invisible image with no description.
3. Ping from Firstboot.
4. Ping from splash (once).
5. Questionnaire from firstboot or splash.
Accuracy
--------
None of the solutions give has any accuracy on the number of users using Fedora.
Analytically the solutions:
1,2: The image could be put in any webpage and the users with dynamic IPs would
be counted as multiple users.
3: Doesn't count upgrades to FC6 and also gives option of just clicking
"next".
4: Suffers like 1: multiple locations could point there and people who clear
their cookies often will see the message over and over.
5: Like 4, but a bit better because the bogus calls will be fewer.
Conclusion
==========
To summarize a very long email (sorry): I believe users would like it better if
we did not use their desktop to track anything or ask them any such question. I
wasn't convinced that the value it will give the project will be worth the cost
of the image of the project.
I'm sorry but I feel this is a very delicate matter that touches sensitive
chords and fragile nerves. Actions like these sometimes make people feel that
their privacy was disturbed or their personal normal lives intruded for
third-party benefits.
If I had to choose from the solutions, I would go for the most "human"/less
intrusive one, #5 (the questionnaire).
Sorry again for the length of the email.
-dim
Note to all we do not even have a "final" Privacy Policy [1} on the Wiki
yet.
[1] http://fedoraproject.org/wiki/Legal/PrivacyPolicy
--
Robert 'Bob' Jensen * * http://fedoraproject.org/wiki/BobJensen
gpg fingerprint: F9F4 7243 4243 0043 2C45 97AF E8A4 C3AE 42EB 0BC6
Fedora Docs Projects FDSCo http://fedoraproject.org/wiki/DocsProject
2:25 p.m.
On Fri, 2006-09-29 at 03:33 +0100, Dimitris Glezos wrote:
O/H Karsten Wade έγραψε:
> Dimitris:
>
> We just got a last-minute idea from Max Spevack et al, to put a hit
> tracker on the splash page. A box in the upper corner/somewhere like
> below[1].
Karsten, Max, et al,
I believe the addition of such a tracker (solution 1) will have a bad impact on
the image of the Fedora Project on our users, which will not be worth the
information gained from this addition.
I'll try to elaborate my thoughts.
Dimitris, thanks for your well written response. I tend to agree with
your analysis. The intentions of getting a rough count of desktop users
are obviously the best, but the timing is too short to work out all the
ramifications (social, technological, etc.). In American football, this
is where we punt, meaning to kick the idea to the next release.
If we had just a bit more time, we could have a button to click to get
us some rough numbers, but timing is just too close.
- Karsten
--
Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project
Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject
quaid.108.redhat.com | gpg key: AD0E0C41
////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
1:04 p.m.
On Sat, 30 Sep 2006, Karsten Wade wrote:
Dimitris, thanks for your well written response. I tend to agree
with
your analysis. The intentions of getting a rough count of desktop users
are obviously the best, but the timing is too short to work out all the
ramifications (social, technological, etc.). In American football, this
is where we punt, meaning to kick the idea to the next release.
If we had just a bit more time, we could have a button to click to get
us some rough numbers, but timing is just too close.
Here's the thing:
We have no useful metris about Fedora *at all*. Just some bittorrent
download numbers, which are not very complete.
We have no clue about desktop usage. We have no clue about server usage.
We have no clue about fc5, fc4, fc6, etc. usage.
We basically don't know *anything*. And how on earth can we effectively
make a case for resources, money, engineers, etc. if we don't have any way
to *prove* the manner in which people are using our software.
It would be wonderful to do this "right" for FC7. But it would be a HUGE
mistake to allow another 6 months to go buy with no metrics because we're
afraid of having some people feel like they're being tracked.
Doing it secretly is not an option -- that's the sort of thing that will
get people up in arms.
Firstboot -- a great idea.
A survey -- a great idea.
A simple URL that lets us know there is a desktop user somewhere out the
on the internets? A decent idea that we can *just make happen*. And if
people want to flame it, they can flame me directly and I'll forward this
email to fedora-announce-list.
I honestly don't understand what the big deal is. It's no more or less
information than you give out *every time* you visit *any website*.
I'm tired of being embarassed when people ask me things like "so, how many
users ya got?" and I can't answer.
--Max
--
Max Spevack
+ http://fedoraproject.org/wiki/MaxSpevack
+ gpg key -- http://spevack.org/max.asc
+ fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21
1:32 p.m.
On Tue, 2006-10-03 at 14:04 -0400, Max Spevack wrote:
We have no useful metris about Fedora *at all*. Just some bittorrent
download numbers, which are not very complete.
We have no clue about desktop usage. We have no clue about server usage.
We have no clue about fc5, fc4, fc6, etc. usage.
We basically don't know *anything*. And how on earth can we effectively
make a case for resources, money, engineers, etc. if we don't have any way
to *prove* the manner in which people are using our software.
I guess my question is, how could this have happened? How could nothing
have been gathered in all this time? Also, why did this just now become
an urgent thing?
Given that the answers to those questions wont help with the present
problem...
What about getting things like number of people registered with the
wiki/on the mailing lists/registered to community sites like fedora
forum/attending fedora conferences, etc?
It would be wonderful to do this "right" for FC7. But it
would be a HUGE
mistake to allow another 6 months to go buy with no metrics because we're
afraid of having some people feel like they're being tracked.
Again -- why hasn't this been raised sooner?
Doing it secretly is not an option -- that's the sort of thing
that will
get people up in arms.
Firstboot -- a great idea.
A survey -- a great idea.
Right so this should be on track for FC7 - any time really.
A simple URL that lets us know there is a desktop user somewhere out
the
on the internets? A decent idea that we can *just make happen*. And if
people want to flame it, they can flame me directly and I'll forward this
email to fedora-announce-list.
I honestly don't understand what the big deal is. It's no more or less
information than you give out *every time* you visit *any website*.
I don't personally think it's a big deal. I think my resistance is
mostly speculation about how people *might* react, which is nothing
anything should be based on.
But, can you really count on this data for anything useful? I mean, I
think a lot of people install several distributions just to check them
out when they are first released, for example, but may not have any
intention of actually /using/ it long term. Does that count as a user?
I'm tired of being embarassed when people ask me things like
"so, how many
users ya got?" and I can't answer.
I agree - it's silly that we can't answer that question.
Why don't we start to prepare some kind of survey of people who are
registered at various fedora-related sites now? I don't know that it has
to correspond with a release.
On the other hand, if you really think this idea will give you data you
can trust and urgently need, it might well be worth it to go ahead. I
just personally think this feels a bit cobbled together.
Just a couple cents here for discussion.
--Max
--
Max Spevack
+ http://fedoraproject.org/wiki/MaxSpevack
+ gpg key -- http://spevack.org/max.asc
+ fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21
--
Sam Folk-Williams, RHCE
Global Support Services
1:52 p.m.
On Tue, 3 Oct 2006, Sam Folk-Williams wrote:
I guess my question is, how could this have happened? How could
nothing
have been gathered in all this time? Also, why did this just now become
an urgent thing?
Just because it's a new problem to you doesn't mean that it's a new
problem. It just means that it's a new problem to you.
Right so this should be on track for FC7 - any time really.
So great. We've got a great idea for FC7. I think it'll be great to
write a firstboot extension that allows us to collect all kinds of spiffy
data. Maybe we can get some community resources to help make it happen --
because we can't necessarily count on internal dedicated resources.
It's worth pointing out at this point that we also had a great idea for
FC4 *and* FC5 -- redirecting yum requests through bouncer and collecting
stats that way. But we didn't execute. Why? Because it wasn't judged to
be important enough to expend resources to *make* it happen.
It's all about execution. We do what we have with what we've got. Ideas
and promises are great, but they don't feed the bulldog. And when we ask
for MORE FUNDING TO MAKE FEDORA BETTER, and the Powers That Be ask, "well,
how many Fedora users do you have?" and we have NO ANSWER, that is BAD FOR
FEDORA. BAAAAAAAD.
AND I'M SORRY FOR SHOUTING... but I guess I just can't help it. It's
frustrating. I'll go stand in the corner now and suck my thumb until I
feel better.
But, can you really count on this data for anything useful? I mean,
I
think a lot of people install several distributions just to check them
out when they are first released, for example, but may not have any
intention of actually /using/ it long term. Does that count as a user?
It's certainly much better than what we have right now, which is
*literally nothing at all*. How many times do we have to reiterate this
point?
Why don't we start to prepare some kind of survey of people who
are
registered at various fedora-related sites now? I don't know that it has
to correspond with a release.
I guess I don't understand what makes you think that survey data will be
anywhere near as accurate.
I mean, here's some simple back-of-the-envelope calculations:
How many people who install Fedora actually join a Fedora mailing list, or
go to a Fedora website? 20 percent? 5 percent? On the other hand: how
many Fedora Desktop users open a browser? 75 percent? 90 percent?
On the other hand, if you really think this idea will give you data
you
can trust and urgently need, it might well be worth it to go ahead. I
just personally think this feels a bit cobbled together.
We had great, large plans, as early as FC4. Grand plans. Beautiful
plans. What happened? We did not execute.
Therefore, we dropped back to the *simplest possible plan* that would give
us *any data at all*.
--g
-------------------------------------------------------------
Greg DeKoenigsberg || Fedora Project || fedoraproject.org
Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors
-------------------------------------------------------------
3:16 p.m.
On Tue, 2006-10-03 at 14:52 -0400, Greg DeKoenigsberg wrote:
On Tue, 3 Oct 2006, Sam Folk-Williams wrote:
> I guess my question is, how could this have happened? How could nothing
> have been gathered in all this time? Also, why did this just now become
> an urgent thing?
Just because it's a new problem to you doesn't mean that it's a new
problem. It just means that it's a new problem to you.
Point taken. The reason for my question is that it came to the list as a
last minute thing.
It's worth pointing out at this point that we also had a great idea for
FC4 *and* FC5 -- redirecting yum requests through bouncer and collecting
stats that way. But we didn't execute. Why? Because it wasn't judged to
be important enough to expend resources to *make* it happen.
This is definitely worth pointing out...
It's all about execution. We do what we have with what we've
got. Ideas
and promises are great, but they don't feed the bulldog. And when we ask
for MORE FUNDING TO MAKE FEDORA BETTER, and the Powers That Be ask, "well,
how many Fedora users do you have?" and we have NO ANSWER, that is BAD FOR
FEDORA. BAAAAAAAD.
Understood.
It's certainly much better than what we have right now, which is
*literally nothing at all*. How many times do we have to reiterate this
point?
I get it. I think there's value to looking at things like traffic to
fp.o, registered users on fedoraforums, etc. But I should have taken
into account that you have already looked into all that.
> Why don't we start to prepare some kind of survey of people
who are
> registered at various fedora-related sites now? I don't know that it has
> to correspond with a release.
I guess I don't understand what makes you think that survey data will be
anywhere near as accurate.
I think it's generally accepted that surveys are a good and useful tool.
I mean, here's some simple back-of-the-envelope calculations:
How many people who install Fedora actually join a Fedora mailing list, or
go to a Fedora website? 20 percent? 5 percent? On the other hand: how
many Fedora Desktop users open a browser? 75 percent? 90 percent?
> On the other hand, if you really think this idea will give you data you
> can trust and urgently need, it might well be worth it to go ahead. I
> just personally think this feels a bit cobbled together.
We had great, large plans, as early as FC4. Grand plans. Beautiful
plans. What happened? We did not execute.
Ok, with that in mind...
Therefore, we dropped back to the *simplest possible plan* that would
give
us *any data at all*.
I'd say go for it. You have definitely showed me that this was not a
last minute thing, which is the impression that I had. Given everything
that has happened it sounds to me like if you don't do this nothing will
ever happen. Which means, DO IT!
Thanks for elaborating.
Sam
--g
-------------------------------------------------------------
Greg DeKoenigsberg || Fedora Project || fedoraproject.org
Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors
-------------------------------------------------------------
--
Sam Folk-Williams, RHCE
Global Support Services
3:02 p.m.
On Tue, 2006-10-03 at 14:52 -0400, Greg DeKoenigsberg wrote:
On Tue, 3 Oct 2006, Sam Folk-Williams wrote:
> I guess my question is, how could this have happened? How could nothing
> have been gathered in all this time? Also, why did this just now become
> an urgent thing?
Just because it's a new problem to you doesn't mean that it's a new
problem. It just means that it's a new problem to you.
Where you == 95% of the Fedora contributors.
> Right so this should be on track for FC7 - any time really.
So great. We've got a great idea for FC7. I think it'll be great to
write a firstboot extension that allows us to collect all kinds of spiffy
data. Maybe we can get some community resources to help make it happen --
because we can't necessarily count on internal dedicated resources.
It's worth pointing out at this point that we also had a great idea for
FC4 *and* FC5 -- redirecting yum requests through bouncer and collecting
stats that way. But we didn't execute. Why? Because it wasn't judged to
be important enough to expend resources to *make* it happen.
It's all about execution. We do what we have with what we've got. Ideas
and promises are great, but they don't feed the bulldog. And when we ask
for MORE FUNDING TO MAKE FEDORA BETTER, and the Powers That Be ask, "well,
how many Fedora users do you have?" and we have NO ANSWER, that is BAD FOR
FEDORA. BAAAAAAAD.
AND I'M SORRY FOR SHOUTING... but I guess I just can't help it. It's
frustrating. I'll go stand in the corner now and suck my thumb until I
feel better.
Chicken and egg situation, as usual. We need the resources to do the
work to show that we need the resources.
When all the incorrect privacy concerns are out of the way, this is how
the whole situation reads to me. I suspect others agree;
<how it looks from outside of the supposed-cabal>
"There are some number crunchers somewhere who think that the value
that Fedora is getting out of Red Hat is *greater* than the value that
Red Hat is getting out of Fedora. So, they are making it incumbent
upon Fedora to prove its worth before they will approve more funds for
Fedora."
</>
Folks who contribute to and use Fedora have the impression that Fedora
is the unloved stepchild of RHEL, right? I think this action is easily
perceived by Fedorans that we need to forget our concerns and once again
do for Red Hat what Red Hat should have been doing for itself.
Bottom line -- if it is so important for Red Hat to know what the Fedora
metrics are, Red Hat should be investing some FTE time into generating
these metrics. Instead, Red Hat is telling Fedora to get off its butt
and prove its own worth.
Bottom line ^2 -- Max, it is a sad state of things that you do have to
turn to the community to accept and do this. Where is your army of
flying monkeys to do your bidding?
I say all this as a person in a position to know better, to have a real
idea how important and central Fedora is to Red Hat. It is painfully
obvious to all that when *anyone* at Red Hat who works on Fedora gets
the RHEL siren call, they respond. We say and know that Fedora is
important, but when it comes time to prove that, we ask Fedora to prove
it for us.
Even for Red Hat, denial is not a river in Egypt.
> Why don't we start to prepare some kind of survey of people
who are
> registered at various fedora-related sites now? I don't know that it has
> to correspond with a release.
I guess I don't understand what makes you think that survey data will be
anywhere near as accurate.
Survey == choice
Click here == choice
Tracker PNG == no choice
That is the difference in methods here. The quality of the data is
secondary.
> On the other hand, if you really think this idea will give you
data you
> can trust and urgently need, it might well be worth it to go ahead. I
> just personally think this feels a bit cobbled together.
We had great, large plans, as early as FC4. Grand plans. Beautiful
plans. What happened? We did not execute.
Therefore, we dropped back to the *simplest possible plan* that would give
us *any data at all*.
Sorry we didn't think of this simple solution early enough to gain some
consensus.
I have other concerns we do not have time to resolve:
* The text is going to be untranslated. We actually had another chance
to get this translated, if this thread had been resolved on Saturday.
Thus, our warning is going to be unreadable by a % of our userbase.
* The text and image is going to stay there forever, each time someone
loads the page. Forget the polluted data quality, it's going to signal
"Fedora are asses" every time someone loads that page.
* Our privacy policy is incomplete (separate email to f-a-b about that
forthcoming)
* Fedora leadership is going to take some heat, but in the end, all of
the hard work and hard hassle is going to be felt by non-Red Hat and Red
Hat contributors on the ground.
All righty, I think I have gotten all my concerns and feelings off my
chest.
It is obvious to me now that an order has been given, so we're going to
go ahead and proceed with this plan:
1. Include text as proposed in this thread in
the /usr/share/doc/HTML/index.html
2. Finish the privacy policy and link that in
3. Figure out at feedback mechanism and link that in
Since the folks who worked on these parts have voiced their opposition,
and I have already honored that in this thread, I'll personally do all
the work to make this happen before Jesse spins the ISO.
- Karsten
--
Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project
Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject
quaid.108.redhat.com | gpg key: AD0E0C41
////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
1:52 p.m.
On Tue, 2006-10-03 at 14:04 -0400, Max Spevack wrote:
On Sat, 30 Sep 2006, Karsten Wade wrote:
> Dimitris, thanks for your well written response. I tend to agree with
> your analysis. The intentions of getting a rough count of desktop users
> are obviously the best, but the timing is too short to work out all the
> ramifications (social, technological, etc.). In American football, this
> is where we punt, meaning to kick the idea to the next release.
>
> If we had just a bit more time, we could have a button to click to get
> us some rough numbers, but timing is just too close.
Here's the thing:
[snip useful truths that are not relevant to why people rejected the
idea]
No disputing the need for more metrics.
I think the point you didn't expect ... and still haven't addressed ...
is that community members on this list who have to deal with #fedora and
fedora-list and the like are concerned about the impression. The
arrival of the request on top of all others and "at the last minute"
feels hasty and the result of inaction. I have the same concerns I did
last week before I pushed the request here, and nearly everyone here has
echoed those same concerns without any prompting from me. Right or
wrong, people are going to complain that "Fedora is tracking me!" Don't
underestimate the size or annoyingly loud voices of our newbie
population.
That all said, I'm a good soldier when treated like one. If you have a
direct order, and sorry if I missed that happen, go ahead and make it.
As commander you'll take some heat. Fedora may take some heat as well,
but if you are convinced that the value of this data is greater than the
devalue of the heat, then I'll keep the faith and go ahead.
- Karsten
--
Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project
Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject
quaid.108.redhat.com | gpg key: AD0E0C41
////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
10:22 a.m.
New subject: (read this one) Re: Firefox splash page tracker
On Wed, 4 Oct 2006, Karsten Wade wrote:
That all said, I'm a good soldier when treated like one. If you
have a
direct order, and sorry if I missed that happen, go ahead and make it.
As commander you'll take some heat. Fedora may take some heat as well,
but if you are convinced that the value of this data is greater than the
devalue of the heat, then I'll keep the faith and go ahead.
No, that's not what I want to do. See my latest on f-a-b in which I am
leaving the decision to a subset of folks (the only one of whom who works
for Red Hat is Karsten).
https://www.redhat.com/archives/fedora-advisory-board/2006-October/msg000...
The only "direct order" that I'm comfortable with issuing is:
Even though we're at the last minute, we need *some kind of solution* for
FC6.
I do not want to "order" folks to do something that so many of them are
opposed to. Especially folks who are working on Fedora out of the
kindness of their own hearts, folks for whom I am not their "manager",
etc.
--
Max Spevack
+ http://fedoraproject.org/wiki/MaxSpevack
+ gpg key -- http://spevack.org/max.asc
+ fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21
6412
days inactive
6419
days old
20 comments
9 participants
participants (9)
-
cornel panceac -
Dimitris Glezos -
Greg DeKoenigsberg -
Karsten Wade -
Max Spevack -
Paul W. Frields -
Rahul Sundaram -
Robert 'Bob' Jensen -
Sam Folk-Williams