Why do you release a product with known bugs?
I really don't understand the logic behind this.
1. Something works completely fine in Hardy
2. You release Intrepid betas and release candidates for testing
3. Testers find problems with the release candidates
4. You go ahead and release Intrepid anyway without fixing these problems and break everyone's computers
Bug 285323 was known since 2008-10-18, yet on 2008-10-30, you released Intrepid with this bug still present. My Nvidia card worked fine in Hardy, but is broken in Intrepid, and people seem to have known about that problem before release, too.
Why? Shouldn't a release not break things that previously worked?
Question information
- Language:
- English Edit question
- Status:
- Solved
- For:
- Ubuntu Edit question
- Assignee:
- No assignee Edit question
- Solved by:
- Koen
- Solved:
- Last query:
- Last reply:
Revision history for this message
|
#1 |
Unfortunately Endolith, I can only agree with you. One wonders if the Ubuntu community is so hung up about windows that it is now mimicking the Windows Vista fiasco. Losing all my personal files because of a "recommended" update leaves me very bitter and back on a reliable XP platform.
It would be great if this usually helpful forum could return us to sanity but it seems the problems are too extensive to be answered. Even getting back to the 8.04 system would be nice but I do not even know how to do that...
Revision history for this message
|
#2 |
ubuntu has to keep to a release schedule, we do the best we can to fix all the bugs we can but there is only so much man power, perhaps you should consider donating your time/money/effort in order to help to get bugs fixed?
please note this is a place to ask actual questions to get support rather than rant.
Revision history for this message
|
#3 |
"ubuntu has to keep to a release schedule"
No it doesn't. Releasing a stable, bug-free product should be a much higher priority than releasing on a certain date. Why isn't it?
"perhaps you should consider donating your time/money/effort in order to help to get bugs fixed?"
I do.
(Also, this is a pretty typical response to complaints about open source software, and I find it unhelpful and somewhat condescending.)
"please note this is a place to ask actual questions to get support rather than rant."
I'm asking a legitimate question about Ubuntu's release priorities, and this is a better place to ask such a question than on the unofficial Ubuntu Forums or in the bug reports themselves.
We become aware of bugs during the testing phase, but then officially release the product anyway without fixing the bugs, meaning that a user "upgrading" their system will break things that were not previously broken. This doesn't make any sense. Can someone explain why this is so?
Revision history for this message
|
#4 |
this is for support with ubuntu not for questions regarding its release quality, please contact canonical if you are not happy with it. this is community support and you do not have a direct line to canonical via this answer support tracker.
Revision history for this message
|
#5 |
As I understood it, Endolith's first post was primarily a plea for help concerning his desire to install Ubuntu 8.10 whilst using an Nvidia Graphics card. This question interests me because I am experiencing similar problems. Judging by other threads here and elsewhere it seems that we are not alone in this quest; far from it in fact! I too have contributed to the community and I am quite prepared to do so again provided that we are given some hope of a solution; either forward, with the Intrepid Ibex, or backward to the comfort of the Hardy Heron. For the moment I have no access to my computer and have great fear that my personal files and folders may all be lost.
If you can help us Gord, please be tolerant and do so...? Sure, we're not ecstatic about all this but then don't most people complain about the cold in winter and the heat in summer?
Revision history for this message
|
#6 |
The question is obviously aggressive, but there is a real point behind: are these scheduled versions really necessary? I mean, it is great to *plan* to release a new version twice a year, but it is also OK to delay them because of important bugs. I think people would prefer to have a 8.12 instead of a 8.10 if the quality is better.
So a correct answer would be: because Ubuntu's policy is to release versions at a precise date. Everyone expects the 8.10 to be available in October, and people would be disappointed if new versions are delayed. It is thus a balance between keeping promises and maintaining quality, both are important for the public image of Ubuntu. Unfortunately, we can't really do anything here, after all, Canonical decides and we can just agree/disagree/
In my opinion, the O.P. is right in a way: it is likely that Intrepid has been released too soon. But this is a matter of debate, and it is not necessary to turn the discussion into a sterile war between pro/anti Ubuntu/Canonical.
Revision history for this message
|
#7 |
"As I understood it, Endolith's first post was primarily a plea for help concerning his desire to install Ubuntu 8.10 whilst using an Nvidia Graphics card."
Yes, I've had several issues after the upgrade, but that's a separate question (49574).
The example I listed here is better (Bug 285323). It shows a functionality (changing LCD brightness) that worked fine in Hardy, stopped working in Intrepid testing releases, which was noticed and a bug filed before release, but was not fixed before the official release. So now hundreds? thousands? of laptop users can't change the brightness of their screens without making their computers completely unresponsive. It worked fine in the last release, but now it's broken for anyone who upgrades. I'm trying to understand what kind of priorities are in place that cause this to happen.
"we do the best we can to fix all the bugs "
I'm talking about *known* bugs. Why are products released with known bugs?
"The question is obviously aggressive"
How so?
"but it is also OK to delay them because of important bugs. I think people would prefer to have a 8.12 instead of a 8.10 if the quality is better."
Yeah. I would think that a definite release date for the release *candidate* is desirable, but then you don't actually push out upgrades to users until the bugs have all been ironed out of it.
"Unfortunately, we can't really do anything here, after all, Canonical decides and we can just agree/disagree/
Where is this set of priorities established, though? Where do we go to change it?
Revision history for this message
|
#8 |
concerning the following question : Why are products released with known bugs?
The answer is easy : if we would never release until all known bugs would be fixed, there would probably never be a release, this is the same for every complex system
If you favor stability above latest edge features (and related bugs) you should consider using a Long Term Support version of ubuntu and/or only upgrade to a new version a reasonable time (a few months) after it has been released - so that most bugs have been corrected
The ubuntu release schedule is slated on a 6-month regular release with pre-defined deadlines.
The ubuntu release schedule is an aggresive release schedule aiming to provide a rich linux experience with a lot of 'new' features and less stress on having the most stable/bug-free system possible.
All the volunteers and ubuntu people do whatever is possible in this time to produce the best release possible
A (critical) discussion on the ubuntu release cycle and others can be found here : http://
also have look at this explanation of the ubuntu release schedule http://
As, I mentioned above, if you prefer stability above latest edge and still want to use ubuntu you should probably consider upgrading a few months after the official release and/or use LTS releases (and wait some time after official release before installing it - for example wait for the 4th point release of an LTS if you value bug-free-ness very highly), which will have allowed the ubuntu community to correct most of the remaining bugs.
This release strategy allows you the choice between latest edge (and higher risk on bugs when upgrading to the latest release immediately) and a high degree of bug-free-ness (if you wait some months after the official release)
You can, and many others do argue that an official release should wait until all known important bugs are corrected, however, what is important can be discussable (how many people should it affect, is there a workaround, etc...)
Having a pre-determined release date with all its problems also have many advantages and this is the choice ubuntu has made.
I have personally not had any problems with the intrepid upgrade but have had some in the past and this can be very frustrating.
I'm not sure what the best channel would be to get Canonical to consider reviewing this release strategy but this forum almost certainly isn't it as it's supposed to answer questions about how using a correctly working system and not to discuss bugs nor something fundamental as a release strategy.
I'd suggest http://
You can suggest a new release strategy idea there and people who agree/disagree can then vote
Koen
Revision history for this message
|
#9 |
there is a suggestion here about a change in release schedule http://
is does not seem to be popular though
Revision history for this message
|
#10 |
> The answer is easy : if we would never release until all known bugs would be fixed, there would probably never be a release, this is the same for every complex system
I don't know if I agree with that. It's certainly impossible to release a product without *any* bugs, but if you already know about them, then you can certainly postpone upgrading those packages on users' machines until the bugs are fixed.
> only
> upgrade to a new version a reasonable time (a few months) after it has
> been released - so that most bugs have been corrected
I thought that's what the alpha, beta, and release candidates are for... The final release should be stable and relatively bug-free. There is no excuse for releasing "upgrades" that make users computers lock up or become unusable.
> As, I mentioned above, if you prefer stability above latest edge and
> still want to use ubuntu you should probably consider upgrading a few
> months after the official release and/or use LTS releases (and wait some
> time after official release before installing it - for example wait for
> the 4th point release of an LTS if you value bug-free-ness very highly),
Why call it a "release" if it has major problems? It's not any different from the testing candidates then.
It seems the distinction is purely arbitrary, made to fit a rigid schedule that serves no benefit to the typical end user.
This pattern of increasingly buggy releases is very discouraging, and I would definitely not recommend or promote Ubuntu to anyone I know. I was excited when I first used it in 2006, but now I'm losing faith.
And the LTS has plenty of major bugs, too:
http://
http://
http://
> I'd suggest http://
> You can suggest a new release strategy idea there and people who agree/disagree can then vote
You're right, thanks. It looks like a lot of people agree with me. Hopefully Canonical will take notice and rethink their "strategy".
http://
http://
http://
http://
http://
Revision history for this message
|
#11 |
Thanks Koen Beek, that solved my question.
Revision history for this message
|
#12 |
Another relevant Brainstorm Idea:
Avoiding feature regressions should be more important than (exact) time based releases ( http://
Revision history for this message
|
#13 |
"This pattern of increasingly buggy releases is very discouraging, and I would definitely not recommend or promote Ubuntu to anyone I know. I was excited when I first used it in 2006, but now I'm losing faith."
I share your opinion. I installed a fresh copy of 8.10 over 8.04 in order to avoid any possible bugs resulting from an "upgrade". I experienced that in going from 7.10-->8.04. Which I would not be worried about in Mac...(ie. upgrading)
However, a clean install was still a Huge disappointment on my dell inspiron because 8.10 would freeze up if I used my wireless internet everytime within first 30 minutes of use. It would not if I had wireless off.
I am so frustrated because wireless worked fine in 8.04. How can something that worked in a previous release stop working in a new and "much-improved" release. It is so wrong! I ask this question not from a technical point of view but from a user-experience stand point.
Revision history for this message
|
#14 |
If keeping release dates are so important, do keep them, but cut the features that are not ready for a release. Why getting the newest and greatest if it doesnt work?
You just need a backup plan for each of the new packages. I believe fedora has implemented a similar way of doing it. Each of their release features have a contingency plan. In general ive always been much more pleased with their reliability.
Who on earth ever heard of (knowingly) having to push broken new version when the old one works just fine. This is madness!
Revision history for this message
|
#15 |
I love a release cycle that keeps to its promise, and I wouldn't want it any other way. So far, my upgrades have almost all been without fail, and surely it requires a little insight now and then to get smaller issues fixed.
While the complaint has been written in good spirit, it is much more effective to care about and treat a specific problem instead of blowing it up and generalizing it.
Your complaint makes the impression to other users that there is something inherently wrong with Ubuntu's release schedule and treatment of bugs, when actually a small bug is pestering you, and your dispute is located to the work of one or two core maintainers.
Nevertheless, I think that it's good these kind of complaints come up. It means Ubuntu has passed the point where it's just another geekish Linux distro, and becomes something average consumers care about with a passion they show towards mainstream products.
We used to have quite different complaints a few years back, concerning interface attractiveness, ease of use and usability, mainly.
I have not forgotten where Ubuntu came from and what it essentially is. It is free on the one hand, but ambitious on the other.
Revision history for this message
|
#16 |
> Each of their release features have a contingency plan.
That might be a good alternative. Instead of delaying the release for a buggy new version, just keep the existing version?
> I love a release cycle that keeps to its promise, and I wouldn't want it any other way. So far, my upgrades have almost all been without fail, and surely it requires a little insight now and then to get smaller issues fixed.
I've had major problems like this with every release, and the problems always turn out to be known already, but were released anyway. There is nothing good that can be said about releasing something that you know is broken. Most other people have similar experiences: http://
This is what Ubuntu is known for: http://
> Your complaint makes the impression to other users that there is something inherently wrong with Ubuntu's release schedule and treatment of bugs
It seems that there is. Meeting an arbitrary deadline should not be more important than the user experience.