Glyphobet writes “I’ve largely stopped reporting bugs to Ubuntu because of the condescending and dismissive attitude from their developers. Until Ubuntu’s bug management culture starts to change, people like me, who can actually help make Ubuntu better, will be less and less likely to contribute.”
‘Why I’ve Stopped Reporting Bugs to Ubuntu’
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
So a guy has a couple of bug reporting issues. Well, him and his friend. The way he states it, he makes it out to be the way the handle all issues – but I’ve had the opposite experience with fairly quick and helpful responses.
I do think he has a point that there was a lack of communication regarding the one f-spot bug, but he should have brought it up on a mailing list where there is a larger number of ears in order to address it. Sometimes just creating a bug entry isn’t enough – they get 100s of these day, if not 1000s. Groups of people read mailing lists. Individuals check bugs.
Edited 2008-01-18 10:55 UTC
hm, i also reported a few bugs to ubuntu, not one has been fixed, a few are present since a few releases.
But i think this is simply a lack of developers doing bugfixing, and not a negative attitude.
“But i think this is simply a lack of developers doing bugfixing, and not a negative attitude.”
Doesn’t that show a negative attitude? That’s their job. Even if they are volunteer, they took the responsibility to do work on Ubuntu, so they should do it. It’s that simple. If they don’t want to do it, they should move aside so somebody else can do it.
It is not their job. As volunteers they offer something freely and that part is the part they like working on. They have no moral obligation whatsoever to actually work on bugs.. as volunteers. None.
In that case I volunteer to sit around and do nothing in particular?
They have no moral obligation to volunteer in the first place; once they join the team, they do have to do their part and iron out bugs. Or leave any time they want.
Edited 2008-01-18 13:08 UTC
“It is not their job. As volunteers they offer something freely and that part is the part they like working on. They have no moral obligation whatsoever to actually work on bugs.. as volunteers. None.”
Uh, when you volunteer to do something, you are saying “I will do this task” so when you take responisbility for something, even if it is volunteer, you do it, or you get out of the way.
Exactly. A promise is a promise, and a real man (or woman) does not break his promises. If you volunteer [promise] to take care of bugs, you take care of bugs – barring any unforeseen events, of course.
And no, boring bugs are not unforeseen events.
When I volunteered for OSNews (yes, we are a voluntary effort too) I promised David, OSNews’ owners, I would take on the role of primary editor – including all the shite that comes with it. Does me being a volunteer mean I can ignore the more tedious tasks of OSNews? Of course not – they are part of the job and I knew that when I accepted the invitation to work here.
Edited 2008-01-18 13:25 UTC
So you don’t get paid?
Exactly, we don’t get paid.
Blue,
One thing that I think of is that these volunteers are likely already doing a boat-load of work… Ubuntu is certainly getting better with each release.
I doubt they could tackle all outstanding bugs, so there will certainly be bugs that hang around release after release after release as higher priority items are tackled.
Bug reporting/fixing is not a closed loop. Unless of course you guys know developers that are literally just sitting around doing nothing
“I doubt they could tackle all outstanding bugs, so there will certainly be bugs that hang around release after release after release as higher priority items are tackled.”
Could you be more of an apologist? The bug tracking system is designed to help them FIX BUGS. By blaming the user or closing bugs prematurely, they are not helping anybody.
When MS acts like this, people scream, kick , bite and generally make a lot of noise. Canonical is a COMPANY, trying to make money by supporting Ubuntu. They cannot do this if they do not fix bugs. If they don’t have enough developers, they should hire more, or retask the ones they have.
Quick note… I wasn’t rude in my post and don’t see the need for you to be in yours… I’m just making a point, hear me out, then reply.
Interestingly enough bug reporting and bug fixing is not a closed-loop system. It never has been and never will be. As your users scale up and potential reports for bugs increase, the gap between fixes and reports grows larger and larger, meaning that only “high profile” items will ever bubble to the top of the conscious of the developers. The way for a bug to become “high profile” is either it’s severity or popularity.
There are varying thoughts on how bug trackers should be managed and the issue is only relevant for huge projects. The Eclipse Foundation (www.eclipse.org) has a dialog that took more than a year just to decide on how to use the “REMINDER” flag for bugs, because of the volume of bug reports it was effectively removing bugs out of the lime-light and burying them due to lack of profile. That was a big hairy mess… in that particular case, if they had closed bugs as something they wouldn’t fix *right then*, it would have given an opportunity for other folks to re-report the issue and bring it back into the spotlight. But instead the “REMINDER” version of the bug would sit in limbo, sinking follow-up reports of the same bug.
Then there are other beliefs that a bug should never be touched unless it’s addressed… and in those cases as reports come in, in increasing scale, they simply sit at the “NEW” stage forever and never get touched.
Then there are the bug-tracker nazis that feel *Every* bug should be processed within days of being filed… in those cases bugs that are not immediately deemed to be show-stoppers are closed out as WONTFIX until their profile is inreased by user demand for example.
With a project as large as Ubuntu there is an actual limitation on what people can physically do… I think it’s absolutely rediculous to think that all these bugs *should* stay open or possibly *could* be fixed for a small collection of developers. And there is a real financial and psychological cost to growing a team beyond a point… so simply throwing more developers at the problem like they are squirrels *does not work* in the real world.
Your comments suggest to me that you have never shipped a real commercial product before or done development professionally for a commercial company… you’ll probably reply to the contrary, but it seems you haven’t experienced the real ins and outs of these settings… just theoretical versions of them.
This is a nonsensical argument… Microsoft and Canonical are structured and run in very different manners. You cannot compare the two (one glaring difference would be the lack of an open bug tracker and communication with the dev team at MS).
You have made a leap of faith, equating “fixing bugs” and “fixing all bugs” some how… I’m not sure, but again, it’s a nonsense assumption… your logic here suggests that there are companies that have products that are free of bugs… or some other absolute along those lines, like Canonical is unique in it’s inability to tackle every bug that is filed.
Again, big leap of faith here in your logic, making all sorts of assumptions about developers, team size, user base, team dynamics, etc. The rest of my posts explains in more detail why this is a narrow-minded and incorrect assumption.
“Quick note… I wasn’t rude in my post and don’t see the need for you to be in yours… I’m just making a point, hear me out, then reply.”
I don’t think I was rude. If honest and rude are the same thing, then I guess I will continue to be rude.
“You have made a leap of faith, equating “fixing bugs” and “fixing all bugs” some how… I’m not sure, but again, it’s a nonsense assumption”
Uh, some of the bugs in the system span 2 or 3 different releases, that is unacceptable. it is not a leap of faith, if I was paying for Ubuntu support, I would expect that bugs that affect me are fixed. period. Not closed with no explanation, and don’t blame me either. If it not going to be fixed, TELL ME. Don’t just close the ticket.
There is a difference between prioritizing bugs and just not dealing with bugs. One of the bugs listed in the article linked here:
https://bugs.launchpad.net/ubuntu/+source/procps/+bug/131188
was not even looked at, if you look at the comments, two assumptions were made, and then it was closed, and misreporting swap is a pretty big bug. This is the sort of behavior that is unacceptable.
The problem with that attitude is– using the aumix bug as an example –many of the bugs reported can be fixed rather simply and easily. The aumix bug was simply a matter of improper compilation according the the blogger. How hard is it for a developer to simply recompile a package?
I’ve been using Ubuntu and testing it on my various machines since it came out, but unfortunately while it does get better in every release, it also gets buggier in every other release or so.
Much of the complaints people I’ve seen posting (on the Ubuntu forums, news sites, etc) seem to deal with the fact the previous release worked. Then Gutsy comes along, being shoved out the door prematurely in order to get as many new features into the LTS release as possible and breaks much of what people liked about Feisty. There’s a reason why the forums were packed with people on the advent of Gutsy’s release and for weeks afterwards with people complaining they were regressing to the previous release.
The biggest issue I’ve seen has to do with the fact things worked in one release and don’t in another. There isn’t any consistency across releases. And that has to change or a lot of people will start distro hopping again…
–bornagainpenguin
Exactemento. Issues with Windows are basically not a problem, since 99% of the people will stick to Windows anyway. This is not the case for Ubuntu – there are ten billion million alternative distributions out there, waiting with open arms. The Ubuntu guys cannot afford things like this.
Sorry, Thom. A statement like this one is an open invitation to flamebait especially the claim of non-issues on Windows system. You should know better.
I don’t think you quite understood what I was saying. Nor do I understand what you are saying, so we’re pretty much even.
Nice to see we can agree on something these days…
–bornagainpenguin
bornagainpenguin,
That is certainly a valid and important data point (that atrophy in the quality is setting in).
I’m sorry to hear that… it’s unfortunate that teams over-extend themselves horizontally across a product at the cost of the depth or overall quality of the product.
I haven’t used Gutsy full time yet so I wasn’t aware of this.
Let’s hope things tighten up in 8.10.
Stop being petty. When you volunteer, you typically volunteer /some/ of your time, not all of it. Even if you do dedicate /all/ of your time, you may still not have enough time to get to everything so you have to prioritize and pick your battles.
Please complete the following sentence:
If you’re not part of the solution, then you’re a part of the _________________
This isn’t about prioritizing or picking one’s battles–both very important to any project or endeavor… This is about Glyphobet (and people like him, including myself) making the decision to stop reporting bugs due to the condescending and dismissive attitude from [Ubuntu] developers. This is about sounding an alarm and warning the people in charge of the project that until Ubuntu’s bug management culture starts to change, people like me, who can actually help make Ubuntu better, will be less and less likely to contribute.
I can’t think of anything more relevant to an OS News site than that, particularly given Ubuntu’s current status as the Linux Desktop for the non-geek Everyman looking for a Operating System alternative that won’t cost them an arm and a leg.
If you dispute the previous statement, please feel free to explain the run away popularity of the Walmart gOS PC which sold out in a day in my area, which is hardly the tech capital of the world…
Thanks!
–bornagainpenguin
I am not being petty, I am being honest. If I don’t have time to do something, I tell people.
I don’t blame the user for bugs.
I don’t close tickets without trying to solve the problem.
It’s not a problem if they said “We’re not going to get that fixed anytime soon, we’re short handed”, it is a problem when they ignore bugs, close bugs prematurely, or blame the user. It’s that simple.
This is not new. Most open-source projects are know for discarting enhancement suggestions, and minor bugs. People complain open-source software is not polished, is more difficult to use, is designed a way that doesn’t always make sense.
I have taken much of my time using open-source software and reporting suggestions to revamp the applications. All of these suggestions have been either:
1. Left as “NEW” even after years
2. Marked as “WONTFIX” with a rude comment underneath
When you have reported more than a hundred, if not hundreds of feature requests, enhancements, and minor bugs and none of them have been adressed, you just stop involving yourself in testing. This is not just Ubuntu, but also software like GIMP, KDE, Seamonkey, OpenOffice.org which are also notorious for this kind of negative attitude.
Oddly the aforementioned applications aren’t the most polished.
Edited 2008-01-18 12:30 UTC
To be fair, at least with Open Source projects you *can* leave bug reports and access the bug databases. I’m still wondering where I can file in a bug report for MS Word’s horrible handling of image positions in a document…
hahh… you must be Mike Cox in disguise.
The last I checked, Windows had 44 critical security vulnerabilities that have not been resolved.
I’ve complained about this, and also made feature requests to Microsoft regarding removal of DRM from Vista.
Microsoft hasn’t fixed these bugs, or removed the DRM from Vista.
Using your logic – Windows sucks.
I couldn’t agree more, but why compare with the worst?!
I’m pretty interested in this list of 44 vulns… can you point them out?
About the DRM… I’m sorry that it bothers you. If you really care, just don’t run DRMed media, and nothing will happen. The alternative to the current state of affairs is that folks will start writing their own DRM software and destabilize your machine by doing unsafe operations (like the well-known Sony product, or another funny program that was using the Debug Registers for DRM purposes).
One could just make DRM illegal… that’ll solve everything.
Try using the bucket tool inside a lasso selection. Fireworks did this, Photoshop does this…GIMP can’t. It fills outside the selection.
When I tracked down the appropriate bug list and filed it, the answer I got from the maintainer was that the bucket fill code was “too optimized” (read: indecipherable) to adapt to anything except its current behavior (using the document’s own bounding box to restrict fill). The prospect of sitting down with the code to understand its workings or rewrite it altogether was off the table.
Hell, GIMP’s own native file format is only documented inside the application’s source code the last time I checked.
Which doesn’t change the fact that he was directly told that the bug WILL NOT get fixed in 7.10 (regarding aumix).
Also why the hell should he go on a mailing list? There’s this bug reporting function in Launchpad. Why use two different approaches for the same thing? Shouldn’t they be equally important?
(Yeah, I know that I’ll get modded down by a specific group of fanboys for this…)
I guess you never were modded down that much.
I completely agree with you, though. As someone posted above you either do what you said you will or don’t hinder other people in doing “your” part. And apparently kinda ignoring people that even make patches is a very bad approach.
Regarding some bugs I had the feeling Ubuntu ships disposable-distributions. That is maybe a disadvantage of so many releases.
Well I don’t think it’s right but at work as a box tester (where I get paid and everyone around me also does), I have to remind people over email about submitted bugs and if I don’t they complain about it because I didn’t constantly remind them about the bug.
It’s like no one reads the bug report, we have board meetings to discuss them and every week I have to remind the developers what the bug is about and during that meeting they’d tell me to email them about it (what I do is copy paste the bug description, they seem to be happy with that).
So I have been told one too many times that it is my job as a tester to make sure people work on the bug.
Perhaps it’s not ideal, but that’s what needs to be done to get the job done…
Well he details more than a couple of issues, and frankly my experience has been very similar. The refusal to even look into certain common issues until users install debugging packages is a bit over the top in my opinion, especially when it comes to upgrade bugs where as the article says, reproducing the bug would require wiping one’s PC and re-installing. If several people confirm the same bug then the developers should look into it whether or not the reporter can provide debugging information.
My personal issue with Ubuntu is how willing they are to ship a new release with major bugs. They seem to be far more concerned about their 6 month release schedule than shipping a solid product. IMHO they really need a shorter development period and a longer bug fixing period for each release.
I like Ubuntu, I use it for all my desktops and laptops, and for my home server (Debian for production servers), but I’m finding it increasingly frustrating when release after release ships with showstopper bugs. I almost always have a botched upgrade on at least one of my machines each time I upgrade to a new version.
That attitude is huge in the OSS community. There is an unwillingness to make products better they way users want it. They expect the users to change all study 4 years of computer science before they even take their idea into consideration… Anything less they are too stupid to bring up anything (although they themselves my not have Computer Science Degrees). As a programmer myself I know it is easy to get lost in your code and making it do what you envisioned and when someone else brings up the fact they hate it, it hurts, but you need to bite the bullet and rethink the process. OSS systems are often very configurable. Having a user change their configuration to get a seemingly unrelated application to work is a bug on the software side not on the user side. The software should be able to handle the different configurations.
yes an No, reporting the Bug in bugzilla an on the mailing list should be all it takes to get the Bug addressed but the Devs should still always look in bugzilla no matter what IMO
Wow, more whining on a blog. What a surprise. Whine whine whine. Me me me. Why don’t *I* get any attention with *my* problems.
Seriously, these are bugs that affect ONE person. One person out of all the Ubuntu users have had these problems. I wouldn’t lose any sleep over that either if I was at Ubuntu.
aumix isn’t even officially supported. It’s like complaining to MS that Quicktime doesn’t work.
Next up; News at Eleven: Not everyone get along.
Just because someone mentioned Ubuntu (on good light or bad) doesn’t mean it’s interesting news.
The arrogance of open source and why *buntu isn’t more than just a copycat of Windows or Microsoft per se.
Read the bug reports. These issues don’t just affect One person.
Given that most people don’t report bugs in the first place, and most of the bugreports had multiple iindependent confirmations, it should be obvious that these are widely experienced bugs.
I did. Most of them only had trouble reports from the blog guy.
At any rate, quibbling over what is essentially anecdotal evidence is of limited value. Most distros use some sort of formal bug tracking system, so it should be possible for someone who wanted to wade through it all to publish a meta-analysis of how well each distro does in resolving issues. Yes, it would be exceedingly messy, and the results would be fuzzy. (I’m not volunteering!) But surely anything is better than following random people’s blog postings.
Actually, we’re both wrong.
As far as I can see, there are 6 launchpad bugs linked by the article. 3 bugs are only reported by the author, the other three have multiple independant confirmations.
So, it’s a 50/50 situation.
Of course, only 1/10 of available packages are officially supported. That takes a lot out of the Ubuntu magic, doesn’t it?
Of the distros I have used, I’ve had the best “customer” experience regarding bug fixing with Debian. Prompt and efficient. These people are experts and there are about 1000 of them. Ubuntu has much less people and those that handle universe are learning on the job.
This doesn’t mean Ubuntu is very bad at bug fixing, it means Debian is really good at it.
Btw I am not using Debian or Ubuntu any more (too overengineered for my taste). I’m using and like Arch which ironically is worse at bug fixing than either 😛
Edited 2008-01-18 12:56 UTC
Debian was responsible for OpenOffice crashing every time you tried to open a digitally signed document because they used a deprecated encryption option in the compile. This went on for something like two years with a myriad of complaints (in both Ubuntu and Debian mailing lists, Launchpad, etc.). The Ubuntu maintainer essentially pointed his finger at Debian instead of recompiling it with the correct (working) option. Before anyone asks, the change in compile options referred to key storage, so it would not have affected file encryption compatibility.
“If I don’t use it, it isn’t broken” is not a viable perspective.
Uh…there are several replies of the “I also have this problem” nature to the aumix bug.
And it’s a rather ‘standard’ application on Linux. As the bug reporter notes, it’s the most commonly used and straightforward way to script changes to mixer values, and several apps actually assume its presence (and functioning state) for that purpose.
A problem with open source is that a lot of developers only cares about functionality that they use themselves. With unpaid contributers, this is understandable since they donate their own time. And we can’t complain about how people use their own time 🙂
Perhaps they could introduce some sort of karma system, where each bug you fix adds to your score. The number of points could be a combination of severity, age and complexity.
A trivial bug doesn’t give many points when it is reported, but will accumulate extra points for each month it is not fixed.
This might give them added incentive to also do the boring jobs. When the bugs are not interesting, perhaps the extra recognition will help?
Edited 2008-01-18 11:16 UTC
Come on, this has NOTHING to do with open source. Same thing happen with closed source software. No company is going to spend resources fixing obscure bugs that affect a tiny fraction of their users.
What is this? Kindergarten?
“Hmmm..sit at home working on a boring bug I dont care about so I can get same “karma” points that is about as usefull in RL as Beenz or do something I really like?”
I seriously doubt the people with the required skills would care about some abstract karma points.
Edited 2008-01-18 11:22 UTC
Yeah pretty much. Sadly, enough most people, especially some of the basement dwelling youth that are starting to get into open sourced programming, do have that kindergarten mentality.
Actually there are a surprising amount of people who work on open source programs with the single-minded purpose of getting that “Look at what I did.” feeling when they fix a bug or write and release a useful program to the world. That’s kind of how the whole thing started, bragging rights on who could program better.
Edited 2008-01-18 14:05 UTC
You ever been on experts-exchange? That’s a technical karma-driven system that consicely disproves your statement.
No it doesn’t. It proves that it works for experts-exchange which is a tech advise site.
That’s very different from a bug tracker and getting people to spend serious amount of time solving problems they dont like or care about.
The amount of time you spend giving the same advise for the unmpteenth time is considerably less than the time you’ll have to spend researching and fixing bugs.
Read up on Windows’ patch database before making glib comments like that. There are patches inside XP to guarantee that DOS versions of SimCity still run because customers complained about it.
When did I mention Windows or even Microsoft. They’re not the only closed-source companies, you know. No need to get all defensive.
If you think commercial entities spend time and resources on fixing tiny bugs that affects virtually noone I have a bridge to sell you.
http://ianmurdock.com/2007/01/14/on-the-importance-of-backward-comp…
You’re funny when you’re dead wrong.
Again, who is talking about Windows? There are thousands of closed-source companies out there and most do not care about tiny bugs that affects virtually no one.
Have you been to Ubuntu’s bug-tracker, launchpad.net recently?
It has karma just like that, and some of us do gain points from doing the boring jobs, such as translations
yeah, so with this karma system, develoeprs will be waiting with the fix, untill it provides suffecient number of points?
sweet system… lol
Ouch… that is a factor that needs to be considered in any system like this.
Any ideas on how to work around that factor?
–bornagainpenguin
Don’t implement it in the first place. Rewarding people for fixing bugs is open to too many social engineering hacks. Not enough bugs to fix? Make them yourself.
a lot do get paid for there work, but look at rpm5.org Jeff doesnt get paid for his work on rpm5
I’ve reported several bugs to Launchpad, some which were gneuine, some not. In my experience, the staff were very polite and helpful regardless of whether the bug was genuine. If this guy is upset with the Ubuntu people he should try reporting a bug in PHP….
I have also given up reporting ubuntu bugs. I have reported quite a few, but not only have they not been fixed, but they have been ignored completly. Maybe I have just been unlucky, but I have given up.
PPC users of Ubuntu are unbelievably aware of this problem. Vector math acceleration libraries that used to take into account G3s without AltiVec now default to only supporting AltiVec. That means the Gnash plugin and other multimedia apps which use it no longer work since Edgy.
Feisty PPC’s kernel can’t competently recognize IDE hardware because the maintainers basically forgot to continue supporting certain modules that were present in Edgy.
Both of these cases have been bombarded with Launchpad requests for fixes and the maintainers have ignored them straight through testing releases all the way to final releases. If you’re lucky, someone will reply that if it bothers you so much, go fix the code yourself.
Annoying yes, but what choice do the Ubuntu devs have in this case?
With the LTS release, they’re trying to make a distro that’s considered reliable and stable enough to be used in corporate situations. Linux gets a considerable amount of criticism for it’s constant-flux nature, and is a significant reason it’s not used in a lot of corporations.
Consider Windows XP: there are bugs and annoyances that have been there since it was first released, but admins, users and developers have grown to know the bugs and therefore avoid the situations that trigger them, or in some cases rely on them. The last thing a company wants is a standard bugfix update released to smash any software they might have that is depending on buggy behavior. Occasionally it’s necessary for critical bugs and security bugs, or severe regressions, but such updates are dealt with caution.
Canonical have decided in that for LTS releases, only critical bugs should be fixed and have set out guidelines for what constitutes a critical bug, and the process developers should go through when fixing them. The Ubuntu developers are following the guideline. This is a good thing. It is possible that in this particular case, aumix should be fixed, but rather than breaking the guideline, perhaps people should try to consider how these cases should be dealt with in a consistent manner and change the guideline as required.
One of the posters to the bug report said
I know it sounds absurd, but you have to see where they’re coming from. Perhaps they should include an optional “unstable-fixes” repository or something.
My experience is exactly like the one described. You report a bug or confirm an already reported one. Then nothing happens while more and more people confirms the bug. And eventually the bug is confirmed but is never fixed. Then, six or eight months later, you install the new version and it might have been fixed or not.
I’m starting to be tired of that “Don’t touch absolutely nothing between releases” attitude of Ubuntu and actually I’m exploring other distributions. I’ve installed OpenSuse on my desktop PC and if all goes well, I’ll replace Ubuntu on my laptop as well.
Judging from my own, very limited experience the Ubuntu folks are friendly.
I reported a bug of Totem (at least that’s what I thought at the time).
For about one month nothing happend and I thought “Allright, my bug report was probably too stupid to even bother with…”
Then I got a whole bunch of emails asking me about specifics. Eventually I could reduce the problem to “Totem freezes when I use it to play a directory that contains one or more mp3s and a text file”.
Then I was asked to check if the problem went away when I used xine instead of gstreamer. It did. So the bug was transferred to gstreamer.
Next they wanted a stack trace and were very helpful in explaining what exactly I needed to do to obtain one.
I should mention here that nobody could reproduce the bug although it happened on my system every time I tried. Nevertheless, they seemed to care about my report.
After I had done the stack trace they occasionally asked me to update the package or install a new version of Ubuntu. The bug didn’t go away and after some time I was tired of doing the same test over and over and stopped responding. Eventually the bug was closed.
That’s not what I had hoped for but still, everybody was friendly and helpful. Maybe I just fscked up my install somehow although I don’t recall doing anything unusual…
The team that puts Ubuntu releases together is quite small and usually very busy with the next planned release. Therefore some bugs get fixed, some don’t. Hell, you cannot even expect that some major bugs like the broken nautilus-python-bindings get fixed.
If you want something more stable, get Debian.
One time I reported a bug… year or two ago maybe…
I had installed ubuntu alongside FreeBSD… which seemed like something that should be simple and without problem… except grub was installed overtop of my previous bootloader and ignored the fact that I had FreeBSD installed…
“Not the end of the world” I thought, as I spend 30 mins reading about grub config and trying to change it to give myself a choice.. All was well until some months later I did an ubuntu update which must have resulted in a grub update — which nuked my changes and made my FreeBSD partition unavailable to boot again.
I reported this as a bug due to the anti-social nature of what was happening…. eventually I got a reply along the lines of “you did your config wrong you should do this” bla bla bla, which is fine… appreciated.
Although I was disappointed that it seemed this was being written off as my fault and no attempt was going to be made to fix it… it’s pretty lame if you install an OS thinking it’s dual-boot safe and it mangles stuff up so you can’t access your other OS. This in my opinion is a bug, and I shouldn’t have to muck around with grub config to fix something like this.
I’m not going to fight with stuff like this.. I just quit using ubuntu in a dual boot environment…
You can’t expect an installation to leave everything intact, but I’ve experienced this more than a few times with the *buntus after an update. The evms thing is a classic thing that tends to destroy systems, even though you’re not actually using evms!
I don’t know how exactly you had your grub configured but I know all my recent Ubuntu installs have menu.lst which is configured to be automatically updated with new kernel installs and the like. This menu.lst has a bunch of “commented” variables actual end up being used for regenerating a configuration and there are also specific places marked as safe and unsafe for editing. If you didn’t try to adapt your configuration around this then it isn’t really unfair to say that it destroyed your changes. They provide a way for you to work around the auto-configuration updater (which is needed if they want to provide a seemless kernel update which doesn’t require manual grub reconfiguration by the end user).
My dual boot Ubuntu installs have usually detected other installs and put those into the correct places in the customized menu.lst configuration, I would say it’s a valid complaint if the installer did not detect your other installs but the updating scripts overwriting your completely custom menu.lst I would not say is a valid bug.
Developers are very often too lazy to explain why a bug was closed. Very often it’s a problem that the upstream is aware about and have plans to fix it – or can’t do anything about it ATM. In other cases, when it’s fixed, a developer.
Sometimes it’s a part of larger changes and mantainers don’t have patience to write what really was the issue.
I am on the mailing list of a severe bug that prevents Ubuntu from updating to 7.10. It was reported the night 7.10 came out and still hasn’t been addressed. It took two months and tens of confirmations (or maybe it is up to 100+ now) just for someone to be assigned to the bug.
Why I’ve stopped reading why anyone has stopped reporting bugs to anything.
Edited 2008-01-18 13:08 UTC
If I’ve said it once, I’ve said it a million times: Ubuntu devs are too busy worrying about their precious next release and all the snazzy new features it’ll contain to worry about the current or (heaven forbid) past releases. They simply seem not to care. Or, the ones that do are too mired down in red tape (or their own egos) to get around to fixing the problem. Oh, occasionally they’ll throw us a bone and kick a few bug fixes into backports or the current release, but otherwise your only hope is to wait for the next big one and pray that one has the bug fix you need (and not too many new bugs.) Or, you could compile your own packages but then why not just run pure Debian?
On related topic, the Ubuntu community is much like the proverbial little girl who had the little curl. When it’s good, it’s very good but when it’s bad it’s horrid. Ubuntu folks need to stop ripping to shreds everyone who criticizes their precious OS. Stop attacking people in forums, blogs and tech site comment boards for constructive criticism. If that energy were funneled into fixing problems, Ubuntu would be a damn perfect OS.
Edited 2008-01-18 13:34 UTC
Yep. If only they chose to support their LTS better I’d be still using dapper, pity that.
But remember the precious is always calling them, wanting them, my…..precious.
It seems to me that Ubuntu devs are more interested in providing the latest version of everything in sacrifice of a proper QA or bug tracking. The existence of the back-ported repositories confirms this trend of the Ubuntu orientation.
Here is a severe bug reported since end of Nov 07
https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.22/+bug/1…
At end, this bug was also related to the fact that nfsv4 + kerberos is broken in the current kernel version and yet no official patch has been provided. This bug is presented in the two kernel versions that have appeared in Gutsy, so using a back version of the kernel did not work.
Broking nfsv4 + kerberos = unable to deploy Ubuntu in our secure network = IT guys not supporting Ubuntu 7.10 = Ubuntu 7.10 banned from list of recommended distributions.
I have an Gutsy 7.10 box that has to access the network servers via Samba shares (instead of nfsv4 mount points), you can laugh about the irony.
Open Suse 10.2, Fedora 8 and previous versions of Ubuntu do not have this problem.
I agree with previous comments of people saying that in general the Ubuntu devs are mostly polite. It just seems that they’re overwhelmed by the amount of work and it translates in a very slow response.
Either Ubuntu starts telling the people the risks of using their distribution (“you never know what is going to be broken next time you do a [sudo apt-get update; sudo apt-get upgrade] “), either they start to put an extra effort in bug tracking.
If people would report bugs for ‘unsupported’ packages in Ubuntu to the Debian maintainers (which are usually the same people, and more than likely the very same packages) maybe it’d be fixed more than through Launchpad. I’ve sent in several bug reports, both through Debian and through Launchpad. I think some of the ones (like the nvidia-settings fiasco) are still not fixed. In Debian, nvidia-settings is a separate package. Under Ubuntu it is, but it conflicts with nvidia-glx. Ubuntu Gutsy and Debian Sid both have the same version of nvidia-settings as well (20070502).
I reported this bug over a year ago. They basically said it wasn’t an issue.
Anymore, I just use Debian Sid. Granted there are a few things that are touch and go occasionally, but for the most part, it runs solid. I would suggest for people who want the latest software, can set up their system themselves, prefer more control and choice, but still want to use the great apt/dpkg system, just use Debian Testing. It’s generally only 10 days behind Unstable, so that any critical bugs can be sorted. It’s a moving target, and is pretty rock solid.
Really at this point the only three things that Ubuntu has over Debian are;
1) The latest Gnome. Every release (why it’s every 6 months) is coincided to provide the latest version of Gnome. This is what drew me (and I’m sure many others) initially to it from Debian.
2) Fantastic forum support. Debian now has several user forums, but still the official way to communicate with developers (much like Ubuntu) is the mailing lists. Mailing lists for Debian have been around forever, but for the average (or new) user, a forum is a much nicer way to talk to people and get help. The Debian forums are no where near as huge as the Ubuntu ones.
3) Quick and easy set up. Ubuntu is very good at a quick, full desktop right out of the box. Of course some see this as a disadvantage that they can’t really optimize or choose what packages are installed. Debian has this just cornered, since you can literally start with just the bare minimum, and build your environment anyway that you’d like. (of course to be fair, you can do the same with the Alternative install of Ubuntu, but in my opinion, if you’re going to do that, you may as well just use Debian anyhow. Ubuntu is great for desktops.)
Debian Testing even usually has newer software than the latest release of Ubuntu. Since Ubuntu usually takes a Sid dump after it makes a release, then synchronizes the packages for a few months, then does a freeze. That leaves a few more months for Testing to have more up to date software.
Ubuntu was literally created out of the frustration that Mark Shuttleworth had with Debian’s release schedule. I don’t really blame the Debian project for releasing the way they do. When Debian makes a release it’s like the earth has quaked and good things have spewed forth into the universe. When Ubuntu makes a release these days, it’s more like “Yay, let’s play with the new GNOME. Well ok, that was kind of boring, with little new excitement. Oops, that bug wasn’t fixed, and here are a few new ones.”
I’m not saying Ubuntu is bad. But I’m just not the type of person to stand still as far as trying out new software. I just love to every day run Synaptic to see what is new in the repository and to see if I have any updates. Yeah, I know, I’m a sick man. But I’ve tried out and found out a more useful programs that way than I could count.
Of all the bug reports that I filled, none got fixed of even officially answered. Some have huge number of posts (like the one about extreme slow Gnome startup in Gutsy) but all you can find there is exotic ways to get around the problem, submitted by the users, that often don’t work anyway.
I also am tempted to stop reporting bugs to Ubuntu.
Good luck with that one. In all fairness though, how do you define slow and how can anyone really give a definitive answer?
The bug thread is very detailed, 30s to 1 minute from GDM to Gnome desktop on many recent machines, machines that would load Gnome in Feisty in around 10 second or less… machine (mine) that load Dropline Gnome in Slackware in 3 seconds…
voila
Well, I personally want this request to be fixed or even decided:
https://bugs.launchpad.net/dapper-backports/+bug/123753
Well, it’s been with “undecided” status for almost a year now, so I don’t except any miracles to happen. And I’m pretty sure that this isn’t the only one… and nowhere near the worst example. It’s just what I would like to have fixed (or please mark it with won’t fix).
While Canonical does have paid employees, it’s not that many. Most of the people working on Ubuntu are volunteers who deserve your praise for doing anything at all, not your derision for failing to work 80 hours/week to fix your particular bug. Also, keep in mind that many bugs have to be escalated to some other project — whenever I submit a KDE bug, they have to then submit it to the KDE people (so what I have done since I figured that out is submit the bug in both places and post cross links so that everyone knows what’s going on).
But wait… why would they have to spend 80 hours/week to work on your bug? They wouldn’t. The problem is that they have hundreds of other bugs that have a higher priority than yours. If they’re smart, they get to these bugs in some kind of priority ordering. The major and common bugs tend to get fixed rather quickly.
Now, yes, there are problems with the slow rate at which some of the more obscure bugs are fixed. By the time someone gets around to investigating it, I’m no longer using that software. For instance, I found a bug in K3b. For files > 2GiB, an extension is required, and a file cannot be larger than 4GiB. Well, it turns out that if you have a file that is exactly 2GiB, it fails to select the extension, so you get a coaster. If you have a file that’s exactly 4GiB, it doesn’t recognize that the file is too big, and you get a coaster. Both cases fail at verify, and the file is zero bytes on the disc. Well, by the time they asked if the problem still happened, I was doing all of my DVD burning on a Mac, and the only PC here with Ubuntu on it didn’t have a DVD burner in it. Oh well. This bug is more amusing than it is a major problem, because you can always check the file size yourself and act appropriately.
Here’s another one. First off, their partitioner has all sorts of bugs in it that have required me to stick the drive in another machine and zero it completely before I would set up partitions the way I wanted. This interacts with their support for software RAID on boot volumes, which is miserable. You have to set up everything manually and work around some of their assumptions. However, 99% of Ubuntu users do not use software RAID for anything at all. (Those that do are using Red Hat or Debian or something else.) So there’s no major urgency to fix this problem since there are countless issues that affect many more people that they really should get to first.
This guy also complains about their apparent “attitude.” Keep in mind that these are geeks, and geeks aren’t known for their social skills. This isn’t a jab at geeks. Intellect is a tradeoff in everyone. We need these geeks for their technical skills, not their ability to make you feel good when they don’t have time to address your bug.
I lead an open source project. No one can be forced to contribute. And those who do contribute should be thanked for their personal time.
Okay, let’s give you that open source doesn’t require this from developers. But for open source to succeed in a marketplace that is increasingly demanding quality and stability, this kind of behavior ain’t gonna fly. Somebody in open source is going to have to pick up the slack, or it will fail miserably in the marketplace within a few years.
Quality and stability? What marketplace is that? The one with 5 gazillion buggy and incomplete Notepad clones? The one with buggy OS releases? The one where security issues goes ignored and unfixed for years?
Edited 2008-01-21 04:39 UTC
ubuntu support is horrible
This was already blogged about before in the same rant-ish way, and the author apologized for it:
http://www.oreillynet.com/linux/blog/2007/12/ubuntu_linux_users_bug…
Last spring I’ve reported a bug that actually got fixed. Around ten of others are left untouched, even though they are confirmed in the comments by others.
Really wish Canonical invested more effort into classifying/fixing the stuff reported on launchpad. At some point they might not be able to stop this snowball and catch really important reports among tons of useless ones.
Anyways, imagine what would happen if MS launched a similar project for reports on it’s products :-D.
I agree with the author. As another example, let’s talk about the wireless (wpa_supplicant ?) bug preventing you from connecting to any acces point with hidden SSID.
This one seems to be here from a long time.
see
https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/39707
or
http://ubuntuforums.org/showthread.php?p=2265168
Hey, I can’t connect at all. Bug reported in Ubuntu 5.x something and confirmed. In 7.10 it still is not fixed. Many people reported and confirmed it for several versions. On some releases just activating the card crashed Ubuntu.
The old reports seems to have “magically disappeared”, but a new report confirms that rt2400 cards still does not work out of the box:
https://bugs.launchpad.net/ubuntu/+source/linux-ubuntu-modules-2.6.2…
If you read Ubuntu’s wiki, you find that cards with rt2400 chipsets are actually recommended. Several years have passed with this recommendation, and with new Ubuntu releases shipping defect drivers for it.
After reporting enough bugs that did not even get a reply, I thought that maybe if I simply did the work myself it would be quicker. Over 3 months after submitting a patch in gnome bugzilla, it still has status “unreviewed” and is not commented. I will of course not bother with writing more patches.
Edited 2008-01-18 17:27 UTC
I agree with other posters on why this is on OSNEWS. The blogger’s just had some negative experiences reporting bugs. With Ubuntu’s release schedule, they would have to prioritise the bugs. In an ideal world it would be nice to have prompt responses to bugs and get an up to date current desktop but that can’t happen. Something has to give.
And the complaint given here :
<quote>The focus seems to be on reducing the number of bug reports and getting their users to do more of their work for them”</quote>
I agree with the developer in trying to push as much work onto the user as possible. Imagine ( being a developer) YOU had to take the whole day to find a specific bug (when the user could do it) when you have other higher priority obligations.
Look we understand the frustration, but ease up on the guys, Man power in the open source universe isn’t as easy to come by as it used to be ( look at KDE4 and other projects)
IMHO this attitude is quickly becoming a part of the problem. Especially with regards to Ubuntu. Look you’re either a hobby organization playing at making a release or you’re a serious company bringing the best Linux has to offer–it doesn’t work both ways. How long does Canonical think ican get away with this attitude of ‘Oh well, we’ll fix it in the next release’ before people give up on them?
You could even say that’s already begun…
That something is giving and the something that is beginning to give is called users.
Rightly or wrongly–and consider the community propaganda before saying its wrongly–the common belief is that Open Source and GPL software is packaged and distributed by developers who CARE about the code and want the application, the APIs, the Operating System environment to grow and to thrive.
It’s even the main selling point in some cases! People hear that in OSS you can talk to the developers yourself and get help or fixes for the program directly. People are given the impression that by submitting bug reports they are helping not only themselves but the entire community of users!
So when they report bugs only to be belittled, see the report vanish or be ignored despite confirmation from various others it belies the belief people have in OSS being any better than their previous alternative.
This is especially aggravating if you have a case such as the one cited in the article where the ‘bug’ is a package being compiled incorrectly and the developers would rather argue or ignore the situation rather than fix the problem by recompiling the dang package…
Either Canonical wants to play with the big boys or they don’t, but they’d better make the decision soon or their users will make it for them–by walking out the door.
–bornagainpenguin (proud previous Feisty Fawn User not so happy dealing with the Gutsy Gibbon flinging poo at me these days…)
Isn’t this reality?
Ubuntu is an open source free application.
People aren’t being paid to fix bugs. People join out of interest.
I’m a software engineer, and yeah there’s a huge gap between things I work on. Somethings my mind is absolutely intrigued by and I’ll work on it with passion.
Others, well…I do them because it’s my job and the product manager tells me its important.
So I can imagine a developer getting all these bug requests, and not being interested in fixing them…after all its volunteering and the product is normally ‘good enough’
I recently moved completely to ubuntu from Windows after my laptop decided NTDLL is corrupt and I couldn’t find my HP windows CD. I came into knowing reality.
I know it’s going to be unpolished.
I know I’m going to have to tweak and what not to get it working.
I know there are going to be loads of products with many bugs (and there have been)
Overall, I’m more than happy with it. The forums have been helpful and rarely has my question gone answered.
Maybe I am a realist so my expectations are lower. But if I want a polished product…I’m pretty sure I’ll have to pay for it. Most of what they develop is ‘good’ enough.
That said, one the examples in the blog disturbs me…when his friend actually went out and fixed the bug with a patch and then ran into trouble submitting it. Now that is something that disturbs me. I don’t have the details on what bug it was, on which project it was, or what the risk level was for the fix…
Allowing people to effectively volunteer is something you have to have in your organization. If that is what is broken…fix it.
I’ve liked Ubuntu – it’s a great Gnome distro, looks good, has great ease of use features, and a huge repository.
However, I’ve noticed that it has grown increasingly bloated and buggy with every release. It seems like Ubuntu is most focused on latest features and it’s 6 month release cycle. Thus, bugs can take a long time to fix, and the distro has become really resource hungry (enabling lot’s of services and background utilities by default).
Thus, I’ve moved on to other distros that I have found to be just as easy, but much more efficient and stable and bug free.
Namely, SimplyMepis is looking really good these days. First, it’s based on the Debian stable branch, with it’s own backports of selected popular packages. This is a really smart strategy, IMHO – it’s based on ultra stable, efficient, and relatively bug free Debian stable, but still has some latest goodies. Also, Mepis ships with media codecs, Flash, Java, and other conveniences that make it more automatic out of the box than Ubuntu. The Mepis installer is second to none, and the resulting install is not overly resource hungry and is quite efficient.
Anyway, this isn’t meant to be a “my distro is better than your distro” rant. It’s just that Ubuntu just isn’t cutting it for me anymore.
Correct me if I’m wrong, but I was under the impression that Mepis was based on Ubuntu…
–bornagainpenguin
You blinked! 😉
Warren Woodford, with much fanfare, moved to Ubuntu 6.04 LTS as the base for Mepis, discovered to his shock and horror that LTS doesn’t mean “Debian Unstable”, and switched, with many face-saving words, back to Debian as his base, leaving the rest of us wondering what he was thinking in the first place. When he originally decided to base Mepis on Ubuntu, I assumed he meant to base each release upon the current development version of Ubuntu, and not just on 6.04 LTS. But he acted surprised that LTS was not continuously updated with all the latest packages. Go figure…
Made me seriously wonder about Mepis’ technical leadership.
Edited 2008-01-20 02:00 UTC
Whoa! Really??
I wish I’d had known that when I was updating my parents computer last week to Linux –only to discover Ubuntu (yes I tried the alt installer, what do you take me for?) had decided to shit a brick on a PCI video card being installed in place of the onboard card. Strangely enough, neither Fedora Core or any version of Windows had had any issues with this…
Yeah well again correct me if I’m wrong but didn’t the OS formerly owned by Michael Robertson also make the jump to Ubuntu as the base distro? I seem to recall reading the Mepis and the *spire announcements within days or weeks of each other. This really conspired to ruin my day, because in the past I’d have just downloaded a different Debian based distro and tried again. This time I had to go with Fedora because its installer was the only one of the various distros I had on hand or could get quickly that worked.
Turned out Fedora had become nicer than I remembered from the last time I played with it around Core 4 or 5. I’m actually tempted to give it a whirl on my Desktop, which is currently tethered to Windows XP due to a refusual for any distro I’ve triied recently to notice and activate my Verizon Westell DSL modem…but that’s a headache for another time…
–bornagainpenguin (not as Ubuntu chauvinist as he once was..)
Well, that’s more or less the way of things in distroland. I use Fedora/CentOS/Ubuntu in various places depending upon my needs, and overall I think Ubuntu has done a little better than the other two on hardware support. But I can point to pieces of hardware which have failed to work in any one of them, but that item worked fine on the other two. Ubuntu is the only distro I’ve tried that get’s the infernal broadcom wireless card in my laptop working, and suspends and hibernates it properly.
Yes, I believe so. The thing that made me wonder about Warren was not that he chose to base off of Ubuntu, but that he chose to base off of 6.04 LTS forever. And his surprise that what was clearly stated to be a stable, frozen, release supported for 5 years was not managed like Fedora Rawhide. Presumably Linspire/Freespire weren’t so foolish. But I haven’t heard much about them since they signed their pact with the devil. 😉
In many ways this is what the stock market calls a painful readjustment. Ubuntu shares have been trading at unrealistically high prices because of such high expectations among its users. Those expectations have been high enough to cause users to overlook some shortcomings – until now. It’s not as if some of this bug trouble started yesterday. In fact it’s been going on for a while but users chose to overlook a lot of it.
A less rose-tinted view will help everyone. Users may come to see that no distro could deliver what they were hoping for. And Ubuntu may come to see that it is trying to do too much with pretty limited resources. It may also come to see that stuffing each release with oodles of the latest and greatest is questionable unless you really do have the resources to cover things. Yes, you may attract the cutting-edge compiz fanboy crowd. But you may also put off a lot of other people, perhaps more people, who want the basics done really well before anything else. People forget that a blown install due to an avoidable bug may be curtains for a teacher up-country in, say, Africa or Indonesia whereas it’s little or nothing to someone in the West with a Core Duo and fast broadband.
Having been nipped by Ubuntu showstoppers more than once I don’t use it, even though I can see that it’s very good. Debian testing here. This to me combines the best of everything. It’s not cutting edge but very nearly, it’s still darn reliable, it doesn’t need dist-upgrades as upgrading is incremental, and it delivers the Debian Way, the system goodness that is also Ubuntu’s core strength, imho.
I don’t know if Ubuntu’s attitude to bug reports in much different from others? Maybe they do lack resources in fixing bugs to some extent, however? Ubuntu has become a huge project with huge amounts of software and users reporting huge amount of bugs all the time. Naturally there will be some lack of resources occasionally in fixing some bugs. A little more good will and understanding from both sides, both bug reporters and developers, would help communication a great deal.
Tony Mobily has written a nice little article about bug reports from the point of view of a free software developer reading and trying to fix those dozens of bug reports.
Love your bugs: a zen guide to keeping your sanity while managing a free software project
http://www.freesoftwaremagazine.com/blogs/love_your_bugs
A few quotes:
“I started getting a little restless about bugs. I would go to bed thinking: my goodness, I really hope there won’t be any nasties tomorrow morning. Sure enough though, I logged in to my Drupal dashboard the following morning, and there they were, the happy bunch waiting for me: one nasty bug, a couple of vague reports, and a feature request. ”
“Eventually, I realised that bug reports are there just to bug you – but in a good way. I realised that if bugs weren’t reported, they wouldn’t get fixed either. And I realised that they, in the end, only matter a little bit: at the end of the day, it’s better to have a program with a couple of bugs than not having a program at all… I also learned that very, very often behind vague, apparently worthless reports there are real bugs waiting to be found.”
“A program is like a plant. To stay alive, it needs considerable amounts of work, wisdom, and love. A watchful eye needs to make absolutely sure that bugs don’t end up ruining this plant, and that any problem is dealt with timely. You have to eliminate the bugs, squash them mercilessly, and yet love the bug reports which are so important to find out where those bugs are in your plant.”
Edited 2008-01-18 20:19 UTC
Debian stable is very stable and relatively bug free. But it’s packages are old, and it doesn’t have the latest goodies/features that everyone wants.
By contrast, cutting edge distros, which Ubuntu is (being based on a frozen version of Debian Sid and Experimental), you get lot’s of new goodies/features, but you also get bugs.
Then look at Mac OSX Leopard. Riddled with bugs at the moment, and there are a lot of unhappy Mac users. But the bugs will be worked out over time. Leopard’s bugs won’t be fully worked out until the next Mac “cat” is about to be released.
And Windows … need I go on? XP, for Windows, is fairly stable and lot’s of bugs have been worked out. But after 7 years, it still has plenty of bugs. And then there’s Vista – again, riddled with bugs. SP1 will probably take care a lot of them, but certainly not all.
Thus, I don’t think Ubuntu has any more bugs than anything else that’s relatively up to date.
I find it amazing one would think the world revolved around Ubuntu…
I tried it out and it was hard to use, Red Hat and Fedora are so much easier to configure (just my opinion)…
Now I will get slammed for not saying Ubuntu is better than sliced bread.
Fedora/CentOS/Ubuntu are my distros of choice. I prefer Fedora or CentOS for my multiuser XDMCP servers at my clients’ sites. Ubuntu is definitely best for my laptop. On my desktop box, I’m comfortable with whichever one I have installed, and feel little pressure to move to any one of the others.
This is quite an accomplishment for a Debian-based distro, as my background was RedHat and I had been rather biased against Debian-based offerings for years because… hmmm… hahhh… we tend to prefer what we are used to. 😉
Edited 2008-01-20 02:22 UTC
It was the same story about Con Kolivas. Dismissive attitude. When bugs are reported to Open Solaris, SUN sure take them serously. Hope this incident changes the Linux community changes to the better and we all learn from this.