SoylentNews is people
SoylentNews
posted by
LaminatorX
on Saturday October 25 2014, @02:41PM
from the whining-is-not-efficacious dept.
from the whining-is-not-efficacious dept.
A grave bug has been introduced into the "wine" package of Debian Jessie, just days before the November 5th freeze deadline. The /usr/bin/wine launch script fails with an "error: unable to find wine executable. this shouldn't happen." message.
Debian has already suffered much unrest lately over the inclusion of systemd, with threats of a fork being issued, along with the possible cancellation of the GNU/kFreeBSD port and the possible dropping of support for the SPARC architecture. After so much strife and disruption, can Debian afford to have such a serious bug affect such a critical package so soon before such a major freeze?
This discussion has been archived.
No new comments can be posted.
Wine Fails to Launch under Debian Jessie, Just Days before the Freeze Deadline
|
Log In/Create an Account
| Top
| 119 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
In Tulsa, Oklahoma, it is against the law to open a soda bottle without
the supervision of a licensed engineer.
(Score: 2) by Konomi on Saturday October 25 2014, @03:47PM
There are always bugs in Debian like this, and checking the bug report page there is already a patch. Freezing isn't something that implies "Oh no the software is broken and it's a freeze1111!", freeze just means no new versions are likely to enter the repos unless it's needed to fix some bug that can't be patched. I don't know why this is over dramatised as some sort of topic that needs detailed discussion. Seriously community, everyone grow the hell up and stop devolving into some 16 year old sleep over party, fucking shesh.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @03:49PM
I use wine to play some games on my Debian system. I just updated wine a few minutes ago, and ran into this bug. I don't care if there's a patch. It's useless to me if I can't immediately pull it in as an update. Besides, it's a really stupid and obvious bug when you look at the patch. How did it even happen?! Didn't they test it at all?!
(Score: 2) by Konomi on Saturday October 25 2014, @03:52PM
There is not guarantee you will not run into bugs in testing, it is called testing for a reason. If you have problems with that I suggest you fall back to stable. Even if the bug may seem obvious to you it is easy for bugs to not show up for the maintainer building the package because of something different on their system that they don't notice.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @04:53PM
Have you looked at the patch? This is a pretty inexcusable bug, even for an unstable or testing version. This isn't an obscure problem that only happens on certain systems. This is a shell script with completely fucked up paths to binaries!
(Score: 2) by Konomi on Saturday October 25 2014, @05:05PM
Yes, I read the patch before I replied to you and my previous reply still stands.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @05:25PM
No, it doesn't stand. This is the kind of bug that even minimal testing by the package maintainer should have detected right away. If this did slip through, then Debian needs to do a massive and thorough review of their entire testing procedure.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @06:57PM
join the wine team and help fix it, you entitled kid...
in free software you receive a gift from developers and you are allowed to participate on the creative process
you are not a costumer and they are not trying to sell you anything... grab this into your forehead
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @07:41PM
Joining the Wine team won't help. This isn't a problem with Wine. This is a problem with Debian. Debian needs to fix their shit.
(Score: 1, Informative) by Anonymous Coward on Saturday October 25 2014, @08:12PM
Join the debian WINE team. Derp.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @08:52PM
Why would I waste my time and effort volunteering for a failed organization like Debian? The whole systemd debacle should never have happened. And this bug never should have happened. I have better things to do with my time than work with inept groups of people.
(Score: 4, Insightful) by Marand on Saturday October 25 2014, @04:33PM
I use wine to play some games on my Debian system. I just updated wine a few minutes ago, and ran into this bug. I don't care if there's a patch. It's useless to me if I can't immediately pull it in as an update.
This will sound harsh, but it's only useless to you because you're choosing to be useless.
You've chosen to use the testing distro, and now you're upset that a bug is affecting you. That's like using the Windows 10 preview that's out and then complaining that you found a bug in it. If you're not willing to accept that there will be problems like this and either wait for a fix or try to find a workaround, then you clearly installed Debian from the wrong repository.
You have options you can choose to deal with the problem:
1. Don't use the testing repository if you can't handle occasional temporary breakage. It's called "testing" for a reason. Stick to the stable release and use the backports repository.
2. Wait for the update.
3. Mix repositories. You can often cherry pick pieces out of stable or unstable as needed. If you really need wine to not change abruptly, start using it out of stable perhaps.
4. Try using the repository from winehq.org. It says it's for sid (unstable) but outside of a freeze unstable and testing don't differ much.
5. Compile it from source. Wine is dead simple to compile. Getting games to work in wine is more difficult than compiling wine itself
6. Buy CrossOver.
I personally like option #5 because it's easy to keep multiple versions of wine around if needed for wider compatibility, plus I can always have the newest version when I want it.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @05:07PM
This isn't an obscure bug, though. This is Wine not starting up at all! It's obvious with even the most minimal level of testing. This is as bad as the JVM not starting up, or the Python interpreter not starting. This bug makes it so that thousands of other programs now can't be run. It's a very serious bug.
(Score: 2) by Marand on Sunday October 26 2014, @04:37AM
This isn't an obscure bug, though. This is Wine not starting up at all! It's obvious with even the most minimal level of testing. This is as bad as the JVM not starting up, or the Python interpreter not starting. This bug makes it so that thousands of other programs now can't be run. It's a very serious bug.
Why does it matter that it's obscure or not? It's a bug, it's testing, it will get fixed in time. This isn't the first time something major has broken in testing, and it won't be the last. I switched from stable to testing in 2005 and have been using it since, and while uncommon, this sort of thing does occasionally happen. Sometimes it's dependency screw-ups that make a package remove stuff it shouldn't, other times it's packaging errors that just break a program, but they get ironed out, especially for the bigger packages. One example that stands out in my memory is Compiz, because it used to break quite often between updates.
I'm not saying the bug isn't a bad one, but it's not a world-ending problem. Hell, the previous version (1.6.2-8) is still in the repository so it's not like you're stuck with the bad one. The first response to "oh crap wine broke" should have been "I better downgrade to the last working version" rather than "I will complain on SN about this".
These sorts of things are part of using the testing repo, though I'll admit it's easy to forget that considering how solid the testing repo usually is. I started using it as a rolling-release Debian on my main system because the major problems are so rare, but occasionally something like this happens and I just work around it for a few days.
Also, not directly relevant, but wine breakage vs python breakage isn't even remotely comparable, considering Python and Perl are actually needed for parts of the system but wine isn't. The JVM comparison is closer, though.
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @09:30AM
If you want to use Debian unstable or testing, then consider using apt-listbugs [debian.org] which notifies you of known critical bugs in packages you are attempting to install/update. Maintaining a system that's running a non-stable distro requires some sysadmining effort. If you don't want to expend that effort, run a stable distro. (Personally, I've switched to running Ubuntu on all but my primary desktop which runs Debian unstable.)
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @03:56PM
I didn't expect such a news story on SN
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @05:09PM
This is a very relevant submission for SN. It highlights yet another serious problem affecting Debian these days. These problems just keep adding up, and it has gotten a lot worse since the whole systemd debacle.
This is just the kind of news I want to find at SN. This is the kind of news that nobody else covers, but it still impacts thousands and thousands of people.
This is a lot better than yet another shitty submission about that man who robbed that store in Ferguson and then attacked the cops and got shot to death. This submission is at least relevant and useful!
(Score: 2) by Marand on Saturday October 25 2014, @04:16PM
There are always bugs in Debian like this, and checking the bug report page there is already a patch. Freezing isn't something that implies "Oh no the software is broken and it's a freeze1111!", freeze just means no new versions are likely to enter the repos unless it's needed to fix some bug that can't be patched. I don't know why this is over dramatised as some sort of topic that needs detailed discussion.
It's also worth noting that this is the entire point of the testing repository. Outside of a freeze, packages start in unstable, and if they go ten days without any massive bugs getting noticed, they get dumped into testing automatically. Other bugs might show up during testing, but they eventually go away with new packages trickling down from unstable.
During a freeze, the testing repo stops getting those automatic new versions and goes into what is basically bugfix mode. What this means is that outside of freeze, a package might go from version 1.2 to 1.3 to 1.4, but if it freezes at 1.4, updates will be 1.4.1, 1.4.2, etc., just fixing bugs in version 1.4 without adding new features.
This is also how the stable repository works, so in essence, a freeze on testing just means it's acting like the stable repository for an arbitrary period while they prepare to make it the next stable. Debian is very strict about not adding new features to software in the stable repo, so the testing freeze is just setup for the next stable release.
Debian's strict avoidance of version changing in stable is why its version of Firefox (iceweasel) follows the ESR release instead of the more rapidly-moving normal Firefox source: because the ESR one gets bugfixes without new features. It also kept chromium from getting into Debian for a while due to similar concerns, but that eventually got ironed out.
It's also the reason Debian has a reputation for having outdated software. A lot of time is spent in freeze ironing out bugs before going stable, and unless you use backports, you won't see a new version beyond bug fixes, ensuring a consistent experience for the lifespan of that release.
---
TL;DR: Unless this still occurs when jessie hits stable, this is a non-news example of a testing repository being used for testing. Submitter AC is likely either a troll or upset that WoW stopped working briefly and used SN's frontpage for his soapbox rant.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @05:04PM
I don't think this submission is a "troll". Wine isn't just one application. It's a system that allows many thousands upon thousands of other apps to run. If you break Wine, you've broken a lot of other software. It would be like the Java VM not running any software, or the Python interpreter no longer starting up. It's a very, very serious bug.
This bug comment [debian.org] is very ominous suonding, too:
That makes it sound like a fix needs to be integrated very soon otherwise this problem could be present for a long time. Maybe it's just worded poorly, but that's how I interpret it.
As it stands, it's now October 25 and there's still no update to that bug's tread of discussion indicating that it has been fixed.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @04:25PM
"SoylentNews is powered by your submissions, so send in your scoop. Only 11 submissions in the queue."
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @04:42PM
> "SoylentNews is powered by your submissions, so send in your scoop. Only 11 submissions in the queue."
If only this were a scoop instead of the normal way debian testing works.
It isn't that this is an off-topic story (I have no patience for people who make that complaint), it is a non-story.
It kind of has a systemd-troll vibe about it though.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @05:28PM
It's not a non-story. It's a major story. I've been using Debian unstable for a long time. I've got thousands upon thousands of packages installed. And I've never had one break this severely, from such a stupid problem that should have been instantly detected by any normal testing procedure. This is big news. I'm glad that SoylentNews is reporting it. At least now I know not to update wine any time soon!
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @08:37PM
> It's not a non-story. It's a major story.
Only for someone who has a very self-centered perspective.
Come back when a bug like this ends up in stable.
(Score: 0) by Anonymous Coward on Saturday October 25 2014, @11:12PM
It's a bug that never should have happened in the first place. I'm not even certain that the package maintainer did a minimal level of testing. You know, like installing the package and seeing if it works. Because clearly it's really fucking broken.
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @12:15AM
Why don't your volunteer to do something about it, smartass? Get a dictionary and look up "unstable" and "testing". Fucktard.
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @01:11AM
In the past, I would have. I've contributed a number of fixes to Debian packages in the past. I used to maintain several packages, too. But after the whole systemd incident, and the tyranny that it involved, I refuse to contribute to the Debian project any longer. When FreeBSD 10.1 is out in a couple of weeks, I'm switching to it, and I'm not looking back. I'm done with Debian as a contributor, and I'll soon be done with it as a user.
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @02:15AM
But you'll continue on as a whiner . . .
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @01:00PM
Pointing out Debian's idiocy is much more productive than trying to contribute to Debian. It's not worth contributing until the idiocy has been cleared away.
(Score: 2) by neagix on Sunday October 26 2014, @02:46PM
I am also very much surprised about the tone and most of the comments here.
Sorry if somebody will get offended, but I think this article is full of shills and clueless people.
There's no drama, and TFA is *NOT* a valid reason to bash Debian.
Looks like somebody wants to release all their rumblings and FUD about Debian and took the chance to do it here..
(Score: 2) by neagix on Sunday October 26 2014, @02:48PM
Actually why did this even get through the moderation queue? Where is the news?
Shall we publish news about each and every bug of Debian before the release? And why not doing giving the same service to other distros? Hopefully they will get some manpower to do more testing and development.
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @06:10PM
Are you intentionally trying to be stupid?
This is obviously more than just about a single bug.
This is about the sort of bug: an unjustifiable one. Even the most basic testing would have caught this problem right away. You know, like installing wine and trying to run a program with it. You know, the kind of testing that should have been done before this update was ever made public.
The real news here is about how the Debian project is falling apart before our very eyes. Read the other stories that are linked to later on in the submission. This isn't an isolated incident. There's been the systemd disaster. There's been the talk of killing GNU/kFreeBSD. All of this is turning the Debian project and its community upside down. There's been talk of forking Debian, even!
Debian's standards are slipping, and they're slipping badly. This is massive news, with far-reaching consequences for thousands upon thousands of people. Millions of people could be affected, if you want to consider everybody who uses a Debian system directly or indirectly.
This news is perhaps the most important I've seen in days. This is news that will have a real impact.
(Score: 2) by neagix on Sunday October 26 2014, @08:20PM
No, I am really this stupid.
Aside from your sarcasm, I perfectly agree that the most basic test toolsuite should have caught such a bug, but I do not agree on why it is of such big impact, nor why this would be a further proof of Debian bad shape.
What is the expectation of quality of testing/jessie?
Personally I would be AOK with the criticism if it were *already* marked stable and this happened, but it's not what has happened.
I think this story should be discussed with a very critic tone on the Debian development mailing list, and topic could be about enforcing stricter testing rules and/or fix some tools, but not as in here, like: "Look kidz! Debian is rotting!"
Maybe Debian is in bad shape, I am not here to discuss that, and probably I agree with some of the other reasons why one could and should (constructively) complain. But specifically this one, e.g. a bug close to freeze time, is not valid IMO: Jessie is still in development, it has not been released.
I was in particular being critic of the mass of comments that were talking like if this had happened on an officially released and distributed stable version, like Wheezy. It ain't happened. No news here, move along.
(Score: 0) by Anonymous Coward on Sunday October 26 2014, @06:18PM
Your attitude is what I'd expect from a Microsoft Windows user.
You're saying, "Yeah, it's a bug. So what. It's totally fine."
That's not how the Linux community sees things. We have these things called standards. When it comes to software, one of these standards involves doing basic testing. It appears that wasn't done in this case, because if it had been, then this bug would have been discovered immediately, well before the updated package ever went live.
Your shitty standards have no place in the world of software development. They have no place within the Debian project.
The Debian project's standards are slipping, and this should definitely be news. This is the most important news I've seen here at SN in days.