I'll check smf package today. It seems I skip your post, About #Maintainer and license - I add them but it will take some time...

2008/1/16, Callan Barrett :

On Jan 16, 2008 7:38 PM, Sergej Pupykin wrote:

...

I'll check smf package today. It seems I skip your post,

About #Maintainer and license - I add them but it will take some time...

Just so you know I didn't go through these by hand, it took all of 5 minutes to get a list of them and run grep.

I have suggested your removal because you aren't voting and this is exactly what the bylaws states must be done in a case like this, if you were unable to vote you must mark yourself as inactive. The fact is I doubt you're bad at packaging but I don't think it's possible to maintain 600 packages.

I think what would be best, especially if you're going to be inactive, is to orphan some of your packages, as many as possible so we can make this manageable. The voting is negligible and I don't see why we can't work something out but this huge amount of packages you maintain is the real issue.

I think the main concern here is because we already had a situation with lots of packages being managed by WillySilly and his disappearance. I didn't review sergej's packages, but I'm satisfied with the speed he updates and fixes bugs reported in them. I do agree that license information is very important for community and no illegal packages should be there. Neverthless none of these issues worth starting a discussion period IMO, because they all can be resolved by a communication. -- Roman Kyrylych (Роман Кирилич)

If anybody want to adopt my package, just mail me... I think Sergej should orphan some packages and some other TU should

adopt and maintain it. This will help to maintain the quality of packages.

16 Jan 2008, Sergej Pupykin wrote:

- 2 of 602 package marked out of date now

But what's your motivation for maintaining so many packages? I think maintaining is more than just bumping the pkgver. It includes following the upstream project, making sure the PKGBUILD conforms to the Arch Packaging Standards, testing the package ... I'm not sure if this is possible with 600 packages. Alex

I fixed most of maintainer tags. (hope I did not mistake and did not add myself to foreign package)) Moreover he seems to not have notification for comments for his packages in

the AUR, so writing a comment about a package is like writing it in /dev/null.

Current AUR sends notification to owner even if he was not subscribed. The last thing: seriously, who the hell needs stuff like

cross-arm-wince-cegcc-libstdcppdll ? More than 500 packages owned by sergej have less than 10 votes, so virtually useless.

People who have PocketPC and want to write program for own needs want to have cegcc package group. Yes, I don't know how many people want it... I have to admit almost all his packages are not out of date (still, I can't

understand how this is possible), kudos for that, but IMHO the quality is more important than the quantity.

Quantity becomes quality

On Jan 16, 2008 12:15 PM, Sergej Pupykin wrote:

...

Quantity becomes quality

I doubt that very much. I can make a mound of crap larger than my house, but that never makes it anything but a mound of crap

Hey I resemble that remark in as much as my house IS a mound of crap ! Not So Anonymous Liviu Librescu - În veci pomenirea lui. (May his memory be eternal.)

Alessio Bolognino wrote

The last thing: seriously, who the hell needs stuff like cross-arm-wince-cegcc-libstdcppdll ? More than 500 packages owned by sergej have less than 10 votes, so virtually useless.

Sergej, why don't you move all of your packages that have less than, say, 8 or 10 votes to unsupported and/or orphan them? This will bring the number of packages you maintain in community to a more reasonable figure and will help calm down the, IMO very understandable, irritation that is being addressed at you here. I think what we are asking you is clear: besides fixing the issue with the missing maintainer and license tags, please DO REDUCE the number of packages you maintain in community! François

2008/1/16, Sergej Pupykin :

there are 2 main problems: - Maintainer tag (which is mostly fixed already) - Missing license (~70 packages) Why I can not fix this issues and continue maintain my packages? I think I can maintain it successfully.

There are no opened bugs in FlySpray. There are 2 out-of-dated packages. - kernel26-xen - proper patch for 2.6.23 kernel not found - aqbanking - which have temporary build problems

Are you afraid the same as willysilly or try to reduce disk usage and mirror traffic?

I guess willysilly's story indeed has some effect here. The other thing, I guess, is that people seem to like some kind of clearness in repositories, i.e. don't want binary repos to contain very rarely used packages. On the other hand, I see nothing wrong in providing as many binary packages as possible (obviously with good quality and maintained), but in this case it's better to have 50 TUs with 10 packages than 1 with 500 packages, IMHO. sergej already agreed to share his packages:

If anybody want to adopt my package, just mail me... I don't think creating a 100-300 orphans in Unsupported now will be much better than maintaining them in Community. :-/ So, my suggestion here is: let sergej continue his job, but try to rethink the need of some packages for Community, as it's better to adopt a more voted package & drop some less needed IMHO.

P.S.: sergej, do not top-post please. ;-) -- Roman Kyrylych (Роман Кирилич)

On Jan 17, 2008 7:46 AM, Firmicus wrote:

I just looked at the TU Bylaws concerning TU removal http://dev.archlinux.org/~simo/TUbylaws.html#Removal and it looks like this discussion is not yet valid, for it says that

"A motion must be made by at least two active Trusted Users for the removal of a Trusted User."

AFAICS only one such motion has been made until now ...

The discussion period was started because of the "special case" and is meant to be "automatically triggered" anyway. Keep in mind that even if this discussion ends in a vote (which it looks like it won't) you're free to vote no, that's why it's a vote. After some of the replies I read yesterday I was feeling good about this thread and was hoping we could work something out where sergej lets go of at least some of his packages but it seems like that's not going to happen any time soon. Sergej, you're saying you're willing to give up some of your packages but you're doing it in a way that people won't want bother and where you still keep as many as possible, what is the deal? You're so dead set on keeping as many as possible but why? I know I'm also guilty of putting some probably unnecessary packages into community but how about we try work something out with the way we put new packages into community? At the moment TUs are allowed to put whatever they like into community without notice to anyone, perhaps this needs to stop and we need some order about how we do things? While we wouldn't need to be as stringent as the devs about this it might be a good idea that set up something like an email of packages we want in community as well as some grace period before that. Perhaps we need to start a new thread about this. I don't think I've seen that good of an argument yet as to why sergej should keep so many packages or why it is not a good idea to have orphans, even if it means a lot of them. Emailing sergej to see if you can grab a couple of packages is *not* the correct way to go about this, I don't want any and the reason I brought this up isn't because I'm jealous of what he has. Please, come up with some proper arguments instead of being so wishy-washy and backing both of us up. -- Callan 'wizzomafizzo' Barrett

On Jan 17, 2008 7:19 AM, Callan Barrett mailto:wizzomafizzo@gmail.com> wrote:

On Jan 17, 2008 7:46 AM, Firmicus mailto:Firmicus@gmx.net> wrote: > > I just looked at the TU Bylaws concerning TU removal > http://dev.archlinux.org/~simo/TUbylaws.html#Removal http://dev.archlinux.org/%7Esimo/TUbylaws.html#Removal > and it looks like this discussion is not yet valid, for it says that > > "A motion must be made by at least two active Trusted Users for the removal of a Trusted > User." > > AFAICS only one such motion has been made until now ...

The discussion period was started because of the "special case" and is meant to be "automatically triggered" anyway. Keep in mind that even if this discussion ends in a vote (which it looks like it won't) you're free to vote no, that's why it's a vote.

Hmm, I should have read the last paragraph obviously :) You were perfectly right. Sorry for the noise. I agree this discussion is important and necessary. I still hope we can resolve the issue. And for that Sergej needs to show some sign of cooperation. Ronald van Haren wrote:

For a lot of these packages I do not see any use for them being in [community]. Please drop a lot of them, orphan them, put them in unsupported and post a list on an announcement on bbs so at least some of them get picked up. The other packages will get adopted in time when somebody needs them (which is questionable seeing the amount of votes). This will give you time to properly maintain your remaining packages.

For all other TUs looking to adopt more packages, or upcoming TUs, there are enough packages with a reasonably amount of votes in unsupported left to adopt. Please choose to maintain them instead of packages you may use which have only a very small amount of votes. Callan if you want to start a new thread about this, feel free to do so this discussion does not get lost here.

I perfectly agree with the above. That was exactly the point I made in my first message in this thread. F

I really do not understand why you discuss count of my packages and why I should move some of them to unsupported. As an onlooker, I understood the reason. Let me try to see if I can tell you how I feel about it: I think the risk of running into a 2nd AUR disaster are big. Plus, I don't doubt your good intentions when you maintain that much

Hey, "Sergej Pupykin" writes: packages, but reality is just that not many people have more than 600 packages installed, let alone from community. So deducing from that I would say that you just _can't_ maintain all of them with the same dedication as someone that only has 30 or so.

Not only I have community packages with low votes. Well, percentage wise, even without looking, I suspect you take the pot.

Let us make _exact_ rules when TU can move package to community and when can not. Let us not. This isn't supposed to be a job droids to, it's supposed to be fun. Use common sense is the only thing people ask for, and people can't understand why you insist on holding onto so many packages.

Constructive way, I think, is notify me about problems (which already done in this thread) or make _exact_ rules and make ALL TUs (not me only) execute it. I hope that was constructive. Don't try to be superman, we know the endgame. (I'm referring of course to the 1st AUR disaster. ;-))

br, benny

On 1/17/08, Sebastian Nowicki wrote:

On Thu, Jan 17, 2008 at 02:10:01PM +0300, Sergej Pupykin wrote:

...

I really do not understand why you discuss count of my packages and why I should move some of them to unsupported.

Not only I have community packages with low votes.

I'd just like to point out that Sergej maintains around 1/3 of community. There are 25 other TUs according to the AUR, and for one TU to maintain a third of community is just, well, surprising I suppose.

I'm having mixed feelings about the whole thing. After all, if a single employee in a company would produce as much 8 others, he'd get a raise and a nice car, not be threatened to get fired. But then again, we don't pay our TUs, and there's no legal binding on anyone. And people can easily leave without any financial or juridical risk. Only thing that can be harmed is fame, and since we're kind of a lower-middle class (as it was revealed in that other interesting recent thread), the quantity of loss in that case might not be very large... --vk

On Jan 17, 2008 11:16 PM, Sebastian Nowicki wrote:

The point I was making is that in comparison to the rest of TUs, Sergej is maintaining a huge amount of packages. As others have said, it's hard to maintain such a large amount of packages, though Sergej is doing an amazingly good job, in terms of keeping them up to date.

I started the discussion period simply because sergej had also not voted for a long time (seems that is changing now), if he had been I would have just brought up his pkgbuilds for regular discussion. Regarding sergej's pkgbuilds, why did it even take a thread like this for you to fix them? Why weren't they fixed when they were uploaded or some time in between then and now? If you're so damn confident you can take care of 600 packages and be any good at it why did I have to start up this thread? I simply DO NOT think you're capable of this and you're kidding yourself by saying you can, it's not hard to bump a pkgver a few times a week, it IS hard to maintain quality for 600 packages at the same time and your work definitely reflects this. Again, stop kidding yourself. At this point I can't see this discussion going very far, just like all TU discussions. I've been doing this for a year now and *nothing* has happened, nothing has improved and stuff has just gotten worse and worse as people stop caring. I don't think I care anymore either because the only people who who have any real opinion aren't TUs and are unable to vote, the rest of you just agree with both sides and contribute nothing. Why even bother? -- Callan 'wizzomafizzo' Barrett

Again, a lot of fuss about almost nothing. The community repo and AUR are working, so why the useless talking? We should be thankful for Sergej to maintain such a high workload (and obviously he's prepared for it, some minor points aside that just needed prompting him) instead of trying to kick him out. Even if he should leave one day there are enough TUs on the team to solve the problem of the packages he leaves behind. Leslie -- My personal blog: http://blog.viridian-project.de/

"Leslie P. Polzer" writes:

The community repo and AUR are working, so why the useless talking? Yeah, why should something be improved. It's those damn stupid people with a stupid agenda.

We should be thankful for Sergej to maintain such a high workload (and obviously he's prepared for it, some minor points aside that just needed prompting him) instead of trying to kick him out. Nobody is trying to kick anyone out, it's about doing the right thing and try to _improve_ things.

Even if he should leave one day there are enough TUs on the team to solve the problem of the packages he leaves behind. Wait, you were a TU right? Did you even have 4% of the packages that Sergej currently has?

Don't talk for others about what they can do. But let me do the same: Nobody likes to clean up a mess.

Leslie br, benny

On Fri, Jan 18, 2008 at 05:44:32PM +0100, Leslie P. Polzer wrote:

Actually one of the main points of this discussion is exactly this. I have read a lot of people saying "no way Sergej is able to maintain this" -- of course without being able to prove it, as the current situation shows rather that he *can* maintain this.

Leslie

On Wed, Jan 16, 2008 at 05:45:30PM +0900, Callan Barrett wrote:

~70 of your packages are missing licenses. ~440 are missing either a maintainer tag of contributor tag. ~550 are missing a maintainer tag.

Given the amount of packages that did not follow the guidelines, I beg to differ. You wouldn't maintain 600 packages would you? And you most definitely would not like to take a share (~20) of those packages if Sergej was to disappear. This isn't the main reason why this discussion was started. It was due to Sergej not voting.

There is one special case for removal, removal due to unwarranted and undeclared inactivity, for which standard voting procedure deviates from the above. This motion is also automatically triggered by repeated quorum offenses, as described in the Quorum subsection of this document. For this special case, SVP is followed with a discussion period of three days, a quorum of 66%, and a voting period of 5 days.

2008/1/18, Callan Barrett :

At this point I can't see this discussion going very far, just like all TU discussions. I've been doing this for a year now and *nothing* has happened, nothing has improved and stuff has just gotten worse and worse as people stop caring. I don't think I care anymore either because the only people who who have any real opinion aren't TUs and are unable to vote, the rest of you just agree with both sides and contribute nothing. Why even bother?

Ok, you'll have my opinion here. But first, let me notice that the bylaws don't say we *must* participate in TU discussions. Even though discussing is useful, it is clearly stated that "All active TUs should be participating in discussions and voting procedures in order to continue meeting the quorums". So the accent is on the quorum, which is reached by voting. I never missed a voting procedure, just like others that didn't express their opinion in this or other threads. Now, back on track. The problems you highlighted on this thread surely exist, but I don't think they can't be fixed. In fact, before starting an official discussion period for Sergej's removal, I think you should have started an unofficial thread to have a free discussion about how can we solve them without being so drastic. * The 600 packages. They are a lot, maybe too many, I'd suggest we find a way to redistribute them a little and move some of them to unsupported (maybe with some exceptions - I'm going to start a new thread about it after I finish this e-mail). A small addition: we always assumed that a TU should take as many packages as he's able to maintain (I couldn't find this rule written anywhere), but we never set a "sane maximum", even though this was discussed after the willysilly mess. I think we should do it now. * The voting procedures. I think this requires a reform, too. We should clearly define how many consecutive voting procedures can be missed before being automatically proposed for removal. Voting is indeed a fundamental duty of a TU, and Sergej (and all the others) should take part in it, trying to limit exceptions as much as possible. I think we should also "ping back" after we miss a vote just to say we're there and we didn't disappear. The real problem here is that Sergej didn't do anything *clearly* wrong. Since the bylaws include many interpretable rules, we can't really blame him for doing what he thought was right (unless his final objective is total arch destruction :-D ). Either we *really* try to reform the bylaws or we should be prepared to handle situations like this one. There has been a lot of talking after willysilly disappeared, but no clear resolutions were made, if I remember correctly. Corrado

Alright, I think I'm beginning to see the other point of view in this.

From what I've seen so far Sergej has done everything I've asked of him in the beginning of the thread, he is voting again and he has acknowledged and is willing to fix up his packages. I don't think I can ask for much more than this even though he did not do it in a way I thought best.

Sergej: If you are sure you can maintain all of these packages I can accept that even if in my opinion it is way too much, provided you fix up your packages and pay closer attention to what you do upload I'm happy. I never actually had a specific list of packages that need a maintainer tag/license/etc. but it's just a matter of using something like grep to check. As long as you can do this yourself I don't think we need a bug report for it. Anyway, next time I want to bring up something like this I promise I'll start an unofficial thread before it before I start the longest thread ever to hit the ML again. I also suggest anyone who wants to talk about restricting what packages go in community to start a new thread about it rather than using this one. -- Callan 'wizzomafizzo' Barrett

On Sat 2008-01-19 18:22 , Darwin Bautista wrote:

On Jan 16, 2008 4:45 PM, Callan Barrett wrote:

...

[...]

Just a thought, the 'unknown' 'license' is actually in the Arch packaging HOWTO: "license: The type of license, if you do not know it please write down 'unknown'."

If the license is unknown, the software shouldn't stay in the repos (legal issues 'n' shitz) -- Alessio Bolognino Please send personal email to themolok@gmail.com Public Key http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xFE0270FB GPG Key ID = 1024D / FE0270FB 2007-04-11 Key Fingerprint = 9AF8 9011 F271 450D 59CF 2D7D 96C9 8F2A FE02 70FB

2008/1/19, bardo :

2008/1/19, Alessio Bolognino :

...

If the license is unknown, the software shouldn't stay in the repos (legal issues 'n' shitz)

For the record, we have 2 packages in [extra] (vim-buftabs, vim-bufexplorer) and 7 in [community] (gnome-python-docs, gtk2-docs, mg, vile, gdm-themes, linux-howtos and icewm-themes) with an unknown license.

I've fixed this in the next version of gdm-themes but had not uploaded it yet. license=('unknown') in extra should be reported as a bug. -- Roman Kyrylych (Роман Кирилич)

2008/1/19, bardo :

2008/1/19, Roman Kyrylych :

...

I've fixed this in the next version of gdm-themes but had not uploaded it yet. license=('unknown') in extra should be reported as a bug.

Ok, I'm going to submit a bug for those two.

Done as #9261 and #9262. Roman, should a bug be filed also for packages with pre-license-field PKGBUILDs? C.

2008/1/19, Roman Kyrylych :

You mean those without license field at all? Devs are aware of them and have a todo. The work is in progress.

Yes, those are the ones I meant, thanks. C.