With the advent of Fedora 15 we have lost features that have been the bedrock of a reliable system - all the name of "progress". Here are my pleas to the Fedora development team:
Give us back a single entry of LUKS passphrase for all filesystems.
Give us back an intelligible display of boot progress.
Give us back the fsck progress crawl so we know boot isn't hung.
Give us back a simple way to see what services are running, started, and failed in a concise format.
Give us back system configuration tools that work in a root window.
Give us back the mount command - that shows only the actual physical filesystems that are mounted - in a readable format.
Give us back a humanly comprehensible device naming system.
Give us back the reboot command that doesn't hang the system, along with informative progress status reports.
Give us back nfs and autofs that work, and which can shutdown if network connectivity has failed.
Give us back USB wireless interfaces that work reliably.
Give us back a filesystem that obeys the access rules for root so that ~/.gvfs is readable/searchable for backups.
Give us back kernels that don't abort instead of an abrt reporting system.
Give us back the assurance of never having to reboot.
Give us back the legendary reliability that was the hallmark of Linux and Fedora.
On 07/31/2011 03:59 PM, David A. De Graaf wrote:
Give us back the assurance of never having to reboot.
Running F 14 and only rebooting for kernel updates:
[joe@khorlia Desktop]$ uptime 16:09:36 up 46 days, 23:29, 2 users, load average: 1.02, 0.84, 0.83
Can anybody running F 15 match that?
On Sun, Jul 31, 2011 at 04:10:39PM -0700, Joe Zeff wrote:
On 07/31/2011 03:59 PM, David A. De Graaf wrote:
Give us back the assurance of never having to reboot.
Running F 14 and only rebooting for kernel updates:
[joe@khorlia Desktop]$ uptime 16:09:36 up 46 days, 23:29, 2 users, load average: 1.02, 0.84, 0.83
Can anybody running F 15 match that?
(mcpierce@mcpierce-worker:~)$ rpm -q kernel kernel-2.6.38.7-30.fc15.x86_64 kernel-2.6.38.8-32.fc15.x86_64 kernel-2.6.38.8-35.fc15.x86_64 (mcpierce@mcpierce-worker:~)$ uptime 09:56:54 up 18 days, 23:05, 3 users, load average: 0.06, 0.12, 0.14
Rebooted due to a kernel release.
On 08/01/2011 12:59 AM, David A. De Graaf wrote:
With the advent of Fedora 15 we have lost features that have been the bedrock of a reliable system - all the name of "progress". Here are my pleas to the Fedora development team:
Give us back a single entry of LUKS passphrase for all filesystems.
I have a single passphrase for the encrypted root fs and encrypted swap on a laptop with F15.
Give us back an intelligible display of boot progress.
Boot with norhgb noquiet nomodeset or something like that. Google is your friend.
Give us back kernels that don't abort instead of an abrt reporting system.
Maybe you should use RHEL or CentOS. Fedora has pretty cutting edge kernels that are bound to have bugs. Be realistic in your expectations. And the abrt system helps developers learn about the bugs in the kernel so they can be fixed. Would you rather have a kernel with more bugs or less bugs?
Give us back the assurance of never having to reboot.
Kernel updates that fix exploits & bugs require a reboot to activate that new kernel. Alternatively you could get Oracle's ksplice service while you can. Even RHEL/CentOS occasionally get new kernels which require a reboot. I have never come across an Internet facing box that did not require a reboot at some point.
Give us back the legendary reliability that was the hallmark of Linux and Fedora.
Fedora & legendary reliability? As much as I like Fedora and use it daily, for anything that requires legendary reliability I use RHEL/CentOS.
Regards, Patrick
Am 01.08.2011 01:34, schrieb Patrick Lists:
Give us back an intelligible display of boot progress.
Boot with norhgb noquiet nomodeset or something like that. Google is your friend.
this does not change the missing OK / FAILED in red/green this time you have a big text-desert at boot
the same for "systemctl start/stop/restart" no feedback, nothing -> usability dehraded
the missing fsck-prigress on boot hurts much if you are on lights out managment or VMware-Console where you can not image if the systm hangs of does fsck
On Sun, Jul 31, 2011 at 6:59 PM, David A. De Graaf dad@datix.us wrote:
Give us back an intelligible display of boot progress.
What's wrong with removing "quiet" and "rhgb" from the "kernel" line?
Give us back a simple way to see what services are running, started, and failed in a concise format.
File an RFE bug against systemctl for whatever functionality you think is missing.
Give us back the mount command - that shows only the actual physical filesystems that are mounted - in a readable format.
"lsblk" and "findmnt" are your friends.
Give us back a humanly comprehensible device naming system.
If you mean device names of NICs, you can uninstall biosdevname and have them named ethX again.
Give us back the reboot command that doesn't hang the system, along with informative progress status reports.
Give us back nfs and autofs that work, and which can shutdown if network connectivity has failed.
If you've encountered bugs in the latter two points, have you filed bug reports?
On Sun, Jul 31, 2011 at 18:59:29 -0400, "David A. De Graaf" dad@datix.us wrote:
Give us back a single entry of LUKS passphrase for all filesystems.
Having to sometimes enter your password twice is a bug, not an intentional design.
On Sun, 2011-07-31 at 18:59 -0400, David A. De Graaf wrote:
Give us... Give us... Give us...
I wonder, do you really assume that sending a list of demands will actually accomplish anything or were you simply trolling?
... Because if you were actually trying to be constructive, oh man, you chose the wrong way to do it.
- Gilboa
Gilboa Davara <gilboad <at> gmail.com> writes:
On Sun, 2011-07-31 at 18:59 -0400, David A. De Graaf wrote:
Give us... Give us... Give us...
...
That's very customary to say "Give us ...". He is not demanding anything at gunpoint :-)
I wonder, do you really assume that sending a list of demands will actually accomplish anything or were you simply trolling?
... Because if you were actually trying to be constructive, oh man, you chose the wrong way to do it.
- Gilboa
He is highlighting a larger problem with Fedora ... These, and others (GNOME, systemd, etc) ... creeping in (more or less openly). Linux Windows (Fedora, Red Hat, ...) ?!
JB
On Mon, 2011-08-01 at 09:04 +0000, JB wrote:
Gilboa Davara <gilboad <at> gmail.com> writes:
That's very customary to say "Give us ...". He is not demanding anything at gunpoint :-)
Oh, OK.
He is highlighting a larger problem with Fedora ... These, and others (GNOME, systemd, etc) ... creeping in (more or less openly). Linux Windows (Fedora, Red Hat, ...) ?!
The OP's "give us" list can be divided into 4 different groups: Bugs: Luks-password-for-all-fs [WORKSFORME], reboot-hang [WORKSFORME], autofs-that-works [WORKSFORME], USB-wireless. RFE: fs_check_display, gvfs? Google-is-your-friend: services_are_running and devices_names. Weird: assurance_never_to_reboot and legendary_fedora_reliability (...).
75% of the items in group 1 WORKFORME - making the *US* in "give us" somewhat questionable. plus, posting bugs in fedora-users rarely help fix problems. (Especially when there's little information about the setup / environment).
Group 2 should have also found its rightful place in bugzilla. Group 3 should have been ... a question. Group 4... well, you be the judge.
- Gilboa
On Mon, Aug 1, 2011 at 04:14, Gilboa Davara gilboad@gmail.com wrote:
... Because if you were actually trying to be constructive, oh man, you chose the wrong way to do it.
I always applaud civilized and well-argumented rants. They are good lists of annoyances (real or perceived) from end users, and people on the list end up learning one thing or another as part of the exchange of opinions.
FC
On 08/01/2011 03:31 AM, Fernando Cassia wrote:
On Mon, Aug 1, 2011 at 04:14, Gilboa Davara <gilboad@gmail.com mailto:gilboad@gmail.com> wrote:
... Because if you were actually trying to be constructive, oh man, you chose the wrong way to do it.I always applaud civilized and well-argumented rants. They are good lists of annoyances (real or perceived) from end users, and people on the list end up learning one thing or another as part of the exchange of opinions.
I agree. I always thought it sad and counter-productive that user complaints are not more welcomed as indicators of where developers' view users' needs and users' view of same, differ. But I suppose that is ignoring the reality of human nature and the desire to do what one wants, free of critisism.
If complaints were less suppressed here, maybe the situation could even be turned into a positive...Fedora could perhaps raise some funds by selling "I survived Fedora-15 T-shirts".
On 08/01/2011 09:52 AM, Stuart McGraw wrote:
If complaints were less suppressed here, maybe the situation could even be turned into a positive...Fedora could perhaps raise some funds by selling "I survived Fedora-15 T-shirts".
<snicker!>
I think it says quite a bit about F 15 that such a shirt would be appropriate, although I don't plan to qualify for one. And, I might add, this will be the first iteration since the infamous F 10 that I've skipped over.
On 8/1/2011 10:05 AM, Joe Zeff wrote:
On 08/01/2011 09:52 AM, Stuart McGraw wrote:
If complaints were less suppressed here, maybe the situation could even be turned into a positive...Fedora could perhaps raise some funds by selling "I survived Fedora-15 T-shirts".
<snicker!>
I think it says quite a bit about F 15 that such a shirt would be appropriate, although I don't plan to qualify for one. And, I might add, this will be the first iteration since the infamous F 10 that I've skipped over.
I for one tried F-15 on one workstation and after two frustrating days reverted to F-14. I am not going to allow F-15 anywhere near my servers.
Mike D
On 08/02/2011 04:44 AM, Mike Dwiggins wrote:
On 8/1/2011 10:05 AM, Joe Zeff wrote:
On 08/01/2011 09:52 AM, Stuart McGraw wrote:
If complaints were less suppressed here, maybe the situation could even be turned into a positive...Fedora could perhaps raise some funds by selling "I survived Fedora-15 T-shirts".
<snicker!>
I think it says quite a bit about F 15 that such a shirt would be appropriate, although I don't plan to qualify for one. And, I might add, this will be the first iteration since the infamous F 10 that I've skipped over.
I for one tried F-15 on one workstation and after two frustrating days reverted to F-14. I am not going to allow F-15 anywhere near my servers.
Wow. The T-Shirt idea got me.
Nice to know that I'm not weird for hopping over F10 and F15.
It could be fun to maintain a historical [Awesome <---- OK ----> Sucks] [ 5 4 3 2 1 ] meter of Fedora releases. Certain types of changes come in waves, and people familiar with the project and developers could look back at version numbers and the subsystem/interface default changes associated with them and chart their own desired development course accordingly.
-Iwao
On 1 August 2011 17:52, Stuart McGraw smcg4191@frii.com wrote:
On 08/01/2011 03:31 AM, Fernando Cassia wrote:
On Mon, Aug 1, 2011 at 04:14, Gilboa Davara <gilboad@gmail.com mailto:gilboad@gmail.com> wrote:
... Because if you were actually trying to be constructive, oh man, you chose the wrong way to do it.
I always applaud civilized and well-argumented rants. They are good lists of annoyances (real or perceived) from end users, and people on the list end up learning one thing or another as part of the exchange of opinions.
I agree. I always thought it sad and counter-productive that user complaints are not more welcomed as indicators of where developers' view users' needs and users' view of same, differ. But I suppose that is ignoring the reality of human nature and the desire to do what one wants, free of critisism.
Possibly the key phrase there is "well argued". I didn't see any arguments, just a list of wishes. None of which I particularly empathise with in relation to F15. (I've have my problems with F15, but none of those.) As others have said the last two: * Give us back the assurance of never having to reboot.
* Give us back the legendary reliability that was the hallmark of Linux and Fedora.
Are simply misleading anyway.
If complaints were less suppressed here, maybe the situation could even be turned into a positive...Fedora could perhaps
Constructive criticism and suggestions are far more useful than just enumerating grievances. Many of the items originally listed were bugs, but I don't see bugzilla ids next to them. In any case, this complaint is hardly suppressed, it's been posted to everyone on the users list, is in the archive, will sit in my email account till I finally run out of server space and has generated about a dozen responses.
raise some funds by selling "I survived Fedora-15 T-shirts".
Will they be available in the UK? ;)
On 8/1/11 11:46 PM, Ian Malone wrote:
On 1 August 2011 17:52, Stuart McGrawsmcg4191@frii.com wrote:
On 08/01/2011 03:31 AM, Fernando Cassia wrote:
On Mon, Aug 1, 2011 at 04:14, Gilboa Davara<gilboad@gmail.commailto:gilboad@gmail.com> wrote:
... Because if you were actually trying to be constructive, oh man, you chose the wrong way to do it.I always applaud civilized and well-argumented rants. They are good lists of annoyances (real or perceived) from end users, and people on the list end up learning one thing or another as part of the exchange of opinions.
I agree. I always thought it sad and counter-productive that user complaints are not more welcomed as indicators of where developers' view users' needs and users' view of same, differ. But I suppose that is ignoring the reality of human nature and the desire to do what one wants, free of critisism.
Possibly the key phrase there is "well argued". I didn't see any arguments, just a list of wishes. None of which I particularly empathise with in relation to F15. (I've have my problems with F15, but none of those.) As others have said the last two:
- Give us back the assurance of never having to reboot.
This is a silly requirement. There are always reasons to reboot. What you mean, from what I gather, is don't make me reboot unless it is absolutely necessary (major kernel upgrade type stuff...)
- Give us back the legendary reliability that was the hallmark of
Linux and Fedora.
Who are you kidding here. Fedora != reliability. RedHat === reliability (I work with RH 5.6/RH 6 servers and they HAVE to be reliable.)
Are simply misleading anyway.
If complaints were less suppressed here, maybe the situation could even be turned into a positive...Fedora could perhaps
Constructive criticism and suggestions are far more useful than just enumerating grievances.
Amen said the choir. Bellyaching about something you cannot change is not going to change it and might just dig the heels in of the developers more. I've been of that ilk for YEARS. If you have something to say, say it, but coat it with HONEY. Remember, the song in Mary Poppins "A Spoon Full of Sugar Make the Medicine Go Down"? That is what we should be doing. Something like this: 'Hey Gnome devs: I know that you are trying to attract new Linux users by giving them an easy to use, hard to mess up interface in Gnome3. Good work and I think you've made great strides in that direction. However, in the process, you have made things more difficult for me, the experienced Gnome user. Can you help me out by keeping Gnome2 alive until I learn all the tweaks and tricks for Gnome3? That would be really nice on your part and it would give me a fall-back position in case something goes horribly wrong with my use of Gnome3."
Many of the items originally listed were bugs, but I don't see bugzilla ids next to them. In any case, this complaint is hardly suppressed, it's been posted to everyone on the users list, is in the archive, will sit in my email account till I finally run out of server space and has generated about a dozen responses.
I'll agree here as well. Bugzilla is YOUR FRIEND. If they do nothing else than mark the bug 'Invalid' at least you told them in their forum that something is not right. Do expect to see the following: Worksasdesigned. Yep, that's the direction the developers are moving and you are not going to change their minds. INVALID. Same as above, or there is no bug. WORKSFORME. You are just going to have to figure out what the developers wanted to do here or ask for help from them (or someone else who figured it out.)
I like the idea of a 'Survived Fedora 15' tee shirt. Add that to the collection of miscues and broken software that I've worked with over thirty years. What is happening here, happens to most folks who fiddle with 'bleeding edge' software. Sometimes the edge is a little too 'sharp' and we get 'cut'. However, what we find will help the overall project and RedHat and the Linux community in general.
James
James McKenzie wrote:
On 8/1/11 11:46 PM, Ian Malone wrote:
On 1 August 2011 17:52, Stuart McGrawsmcg4191@frii.com wrote: .. but none of those.) As others have said the last two:
- Give us back the assurance of never having to reboot.
This is a silly requirement. There are always reasons to reboot. What you mean, from what I gather, is don't make me reboot unless it is absolutely necessary (major kernel upgrade type stuff...)
I agree, reboot after kernel upgrade is perhaps tolerable.
- Give us back the legendary reliability that was the hallmark of
Linux and Fedora.
Who are you kidding here. Fedora != reliability. RedHat === reliability (I work with RH 5.6/RH 6 servers and they HAVE to be reliable.)
From RH 4.x up to F12-F13 I wasn't afraid put it to servers (either
productions) even several weeks before final distro release. And all works well, although sometimes with particular, quickly repaired problems. F14 and F15 are from my experience troubled, at least for me. There are unstable pieces as sssd, replace openssl with nss, systemd, and others. Plus desktop stuff, with from bad to worse, always problematic vga drivers (not server, but desktop problem, and most likely affected by accidental HW/political things in this area). For years I sneered to mswindows users for their machine freezes - and now it is reality on Fedora Linux desktops, X on my Intel 82G35 based PC hung at least once a week (and as can see, I'm not single afflicted).
Uptime from one my unattended server/desktop mix (CIFS server for small firm with cca 6 win PCs + LTSP server for 3 diskless K12LTSP stations (implied NFS stuff) + itself acting as desktop for 7 days x 12 hours dispatching service + some other network services /ntp,dns,dhcp,dovecot, squid,xfs,sendmail+clamav/regex/greylist milters,squirrelmail/apache, saslauthd,cups,...):
$ cat /proc/version Linux version 2.6.23.17-88.fc7 (mockbuild@xenbuilder2.fedora.redhat.com) (gcc version 4.1.2 20070925 (Red Hat 4.1.2-27)) #1 SMP Thu May 15 00:35:10 EDT 2008 $ cat /proc/uptime 90926960.50 211065.67 $ bc -l <<<"90926960.50/(24*3600)" 1052.39537615740740740740
(it is from old gold FedoraUn*x times, not today's FedoraWinDOS era ;)
Franta Hanzlik
On 3 August 2011 03:15, James McKenzie jjmckenzie51@gmail.com wrote:
On 8/1/11 11:46 PM, Ian Malone wrote:
As others have said the last two:
- Give us back the assurance of never having to reboot.
This is a silly requirement. There are always reasons to reboot. What you mean, from what I gather, is don't make me reboot unless it is absolutely necessary (major kernel upgrade type stuff...)
- Give us back the legendary reliability that was the hallmark of
Linux and Fedora.
Who are you kidding here. Fedora != reliability. RedHat === reliability (I work with RH 5.6/RH 6 servers and they HAVE to be reliable.)
Are simply misleading anyway.
To clarify, these weren't my points, they were from the original post, but since there was nothing else from it I decided to leave out a preceeding >>>>>>>>>>>>>>.
On 08/02/2011 08:15 PM, James McKenzie wrote:
On 8/1/11 11:46 PM, Ian Malone wrote: [...]
Constructive criticism and suggestions are far more useful than just enumerating grievances.
Amen said the choir. Bellyaching about something you cannot change is not going to change it and might just dig the heels in of the developers more. I've been of that ilk for YEARS. If you have something to say, say it, but coat it with HONEY. Remember, the song in Mary Poppins "A Spoon Full of Sugar Make the Medicine Go Down"? That is what we should be doing. [...]
I certainly agree "Constructive criticism and suggestions are far more useful than just enumerating grievances" but don't think that means the latter are useless.
As to the honey part, I disagree. Developers and users constitute a partnership and I don't think it is good to set the smaller number of developers on a pedestal, approachable only with circumspection and deference reminiscent of a French commoner approaching the court of Louis XVI.
Both partners get pluses and minuses from the partnership. Developers get bug reporting, bug fixes, and the ego strokes from being developer of a widely used piece of software. They may also get financial rewards in the form of consulting or book contracts, or job offers. Users of course get to use useful software for free.
If a developer needs deference as well, then they should distribute their software under license terms that require that. Or restrict its distribution to people they know will not be critical.
But if they want the benefits of widespread use, I think they need to understand that those benefits are going to come with some criticism as well as admiration and gratitude.
None of this should be taken as encouraging the dissing of developers or their work (and certainly not blatantly insulting or personal attacks), but rather discouraging the 10 times as many "don't say that" responses that each of the former seem to generate.
On 8/3/11 12:35 PM, Stuart McGraw wrote:
On 08/02/2011 08:15 PM, James McKenzie wrote:
On 8/1/11 11:46 PM, Ian Malone wrote: [...]
Constructive criticism and suggestions are far more useful than just enumerating grievances.
Amen said the choir. Bellyaching about something you cannot change is not going to change it and might just dig the heels in of the developers more. I've been of that ilk for YEARS. If you have something to say, say it, but coat it with HONEY. Remember, the song in Mary Poppins "A Spoon Full of Sugar Make the Medicine Go Down"? That is what we should be doing. [...]
I certainly agree "Constructive criticism and suggestions are far more useful than just enumerating grievances" but don't think that means the latter are useless.
They may not be useless, but as a developer I would find them offensive and not pay much attention if the management has told me to 'go another way'.
As to the honey part, I disagree. Developers and users constitute a partnership and I don't think it is good to set the smaller number of developers on a pedestal, approachable only with circumspection and deference reminiscent of a French commoner approaching the court of Louis XVI.
We already have done this. We basically have alienated the developers to the point that they don't even reply to legitimate bug reports. That is a BAD place to be.
Both partners get pluses and minuses from the partnership. Developers get bug reporting, bug fixes, and the ego strokes from being developer of a widely used piece of software. They may also get financial rewards in the form of consulting or book contracts, or job offers. Users of course get to use useful software for free.
I agree here. But we also have to remember the phrase "Don't bite the hand that feeds you." If the developers are doing this for 'free' then we should be as gracious as we can when we find something offensive. This has driven more free-lance developers to 'pay jobs' where they are isolated from the 'common chaff' and don't have to listen to nor react to the grumblings of the 'masses'. And much of Linux development is done by ordinary people on their 'free time'. And many have found 'much better' things to do with it.
If a developer needs deference as well, then they should distribute their software under license terms that require that. Or restrict its distribution to people they know will not be critical.
No, we need to be critical, but not destructive. There is a fine line between the two. Complaining that feature 'A' does not work as designed/intended is just fine. Keep it civil. However, complaining that feature 'A' is a Piece of Foul-Smelling Dung that is not worthy of being on Planet Earth is neither and does not win friends in the development world. That is most of what is being said about Gnome3. It works, that we all can agree on. It is NOT efficient, and most of us can agree on that too. However, it does meet the needs of the target audience and that is what is the crux of the matter. We, the 'all knowing' want our Gnome2. Well we have to come up with a really good business case as to why. Otherwise, the developers are going to put a massive amount of energy into Gnome3 and Gnome2 will go by the wayside. This is called 'Progress'. And it stinkith madly.
But if they want the benefits of widespread use, I think they need to understand that those benefits are going to come with some criticism as well as admiration and gratitude.
Again, criticism yes. Slobbering at the mouth fanaticism no. Remember, developers are PEOPLE first and foremost. I go back to my 'Honey' statement. Read through it. This is how to give criticism per the 'Leadership Guide'. You point out what is good and right about the product and then you state "But it really does not meet the requirements of group XXX" and state what will meet their needs. We the knowing don't need the features of Gnome3, but if you sit there and bad-mouth it all day, the developers are just going to say 'They are Grumpy and THEY WILL ADAPT to what we give them." Same was said for Windows95. Do we still have, to this day, a command line interface? Yep, sure do and I use it everyday. Why? Because a group of folks approached Redmond with a 'really good reason to keep it'. And it stuck.
None of this should be taken as encouraging the dissing of developers or their work (and certainly not blatantly insulting or personal attacks), but rather discouraging the 10 times as many "don't say that" responses that each of the former seem to generate.
Again, I agree. Constructive criticism, delivered honestly, is a good thing. Dissing and bad-mouthing a product does nothing for you (well you'll feel better until you realize that Fedora 17 is out and Gnome2 is nowhere to be found) and definitely does not paint a good picture of you in the development team's vision.
Again, like Mary Poppins said "A spoon full of sugar makes the medicine go down." In other words, if we give the good as well as the bad, we have a possible way of making things better. If we point out only the bad, then things can only get worse. (And I've been there with this so I have the Practical Experience badge to go with this.)
James
-
Bruno Wolff III -
Darryl L. Pierce -
David A. De Graaf -
Fernando Cassia -
Franta Hanzlík -
Gilboa Davara -
Ian Malone -
James McKenzie -
JB -
Joe Zeff -
mike -
Patrick Lists -
Reindl Harald -
Stuart McGraw -
Tom H -
夜神 岩男