[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Debian needs more buildds. It has offers. They aren't being accepted.

Anthony Towns <aj@azure.humbug.org.au> writes:

> On Sat, Feb 14, 2004 at 11:09:52AM +0100, Mathieu Roy wrote:
> Erg, another non-developer? But hey, at least you're in the queue.
> > Anthony Towns <aj@azure.humbug.org.au> wrote:
> > > My conclusion is that in order to achieve our goals of supporting our
> > > users and the free software community, we can't give too much credence
> > > to people just because they complain a lot. One of the ways in which we
> > > do that is to say "it doesn't matter how much anyone complains; it's the
> > > maintainers decision. if you don't like that, convince them otherwise,
> > > prepare a patch, work around the problem, use a different program,
> > > or use a different distribution".
> > This make sense. But if I correctly understood the situation, it would
> > be like if the maintainer was refusing to apply a correct patch.
> No, it would be like a maintainer ignoring a patch because it's not
> particularly important, may not be desirable and there are other things
> to worry about. And that's the maintainer's prerogative.
> > Maybe I misunderstood the situation. Maybe the complainers should
> > write down exactly what they propose, on just one page, and then, the
> > maintainers should reply to each proposal why it is possible or
> > not. 
> Again, no. The maintainer's don't owe anything to the complainers. If the
> complainers want to do something, it's *entirely* their responsibility to
> make it happen. If they're not being helped, they need to work out some
> way of either working with the maintainer in a way that's acceptable to
> them, or they need to work out a way of making it so the maintainer's
> not affected at all by their desires.
> A lot of people like to analyse open source development as a "gift
> culture".  One of the corollories of that is that you don't get to place
> conditions on the gifts people give you, or demand more gifts. You accept
> them with a smile, and if they're not enough you work something else out.

If I have a patch and the maintainer just ignores it, if I then
contacted him and got ignored too. What else would anyone do then but
NMU the package? One can only assume the maintainer is MIA.

If the maintainer doesn't like a patch he can easily say so. Only then
it's the patchers problem to not impede on the DD.

> > >> Can't you keep off the discussion personal attacks? Can't you just
> > >> stick to facts and avoid any comments based your very own opinions
> > >> about others persons involved knowledge?
> > > No, because my goal isn't to justify James and Ryan's actions --
> > > I don't believe they need any justification. My goal is to try to get
> > > this community to stop ceaselessly attacking the people it relies on.
> > You are maybe in your good right, but to prove that, you must in some
> > way justify James and Ryan actions. 
> The only time it's appropriate to review maintainers' actions is when
> they're egregiously bad. In this case, all the buildds for all arches are
> operating within spec, so there are absolutely no problems to be solved.

That where we disagree. Its just a plain fact they don't all operate
within specs. Its your opinion they operate good enough but thats just
because the problem is still being worked around.

Look at m68k, I was gone for the weekend and already 0.5% drop again.
2 days work of 2 buildds lost for the port.

> Maybe it's fun to second guess people on things they know more about
> and have more experience with than you, but it in general just seems a
> waste of time to me.
> If you're worried about oversight in general, rather than in this specific
> case, then that's not an issue: we have mechanisms available if there are
> egregious problems anywhere, and we have multiple people in a position
> to review the effectiveness of the buildds. That you mightn't be one of
> them is not a problem for the project.
> > In other words, criticizing a work being done by Debian maintainers is
> > not necessarily a bad thing. 
> Well, yes, it is necessarily a bad thing. Nobody likes criticism,
> and when we don't offer anything for peoples' contributions apart from
> appreciation and the knowledge they're doing something good for the world,
> criticism isn't helpful.
> There are times when it can be mitigated because certain issues are
> important enough to overcome that drawback, and it can conceivably be
> unavoidable in certain circumstances, and in some cases having robust
> criticism can encourage much more healthy debate, but this case doesn't
> come under any of those headings. The issue isn't important, it could
> have been dealt with in a number of other ways, and there's nothing
> healthy about announcing to the world that people are incompetent,
> on crack, and need to be removed.

You brought in "incompetent" and "on crack" into _this_ thread.

> > If it has no ground, it should not be
> > very complicated to prove it and to drop the charges. 
> If it has no grounds the "charges" should never have been made in the
> first place.
> > If it has ground, the problem is not the critics, 
> Martin's already summarised the specific issues. Are the grounds for
> the decisions of the delegates to be overruled, or their delegations to
> be withdrawn?

There are grounds for changes, extend to be seen, and some changes are
already being done.

> > Telling to someone he is doing something wrong is not acting against
> > him,
> Your idiocy in this thread can only be explained by your utter
> incompetence, or an addition to cocaine. I call upon the DPL to ensure
> you're re-educated and admit the validity of my claims and support me
> in future threads, or to remove you from the project (or at least this
> list) immediately.

So we are back to personal attacks. Well done. Not even a smiley to
show irony or something.

> > in some way, it helps him. 
> Tell me, would you categorise the above paragraph as helpful or "acting
> against you"? How would you compare it to the rest of the mail? Do you
> think this email would be be better without the above paragraph?
> What do you think are the substantive differences between the mails
> complaining about this issue, and the above paragraph?

Telling someone when he is wrong hopefully stops him from being wrong
the same way next time. You can't learn from your mistakes unless they
are recognised as such.

> > Sure, such critics does not have to be
> > violent and aggressive, but if you refuse to discuss the whole
> > subject, it is likely to harrass the people that are trying to do
> > their best, just like James and Ryan. 
> There's an easy way to avoid all this harrassment: don't continually
> whine when a decision upsets you.
> > >> You are proving that  "the people 
> > >> who" should "do the judging are" not necessarily "the ones in charge
> > >> of the area", because technical knowledge in no way means management
> > >> capabilities. 
> > >
> > > If you make the right decision, and are independently judged to have made
> > > the right decision, but are attacked for it nevertheless, who is at fault?
> > > You, for not fully persuading your attacker in advance that you're utterly
> > > blameless; or your attacker for making an issue out of something they don't
> > > actually understand?
> > Well, the whole thread already looks like a trial. To go on, maybe it
> > should be good to forgot what the other said and done (on both sides)
> > and just tell what we could accept to do to reach a consensus about
> > buildd.
> What value, exactly, do you think your opinion on the buildds has?
> Personally, I don't think the opinion of the people who aren't actively
> working on them matters at all -- as I've already mentioned, it's likely
> to be highly underinformed and underexperienced and all the poorer
> for both things -- and the only thing it's likely to influence is the
> direction of pointless flamewars on -devel. 
> But frankly solving this one issue isn't going to make any difference on
> the larger scale. If it's solved, in a couple of weeks there'll be someone
> else complaining about James or Ryan's or someone else's incompetence,
> demanding they justify everything they've done, and people jumping in to
> say "Yeah, I agree. And while I might not know exactly what's going on,
> well, that's your fault too."
> We've had, eg:
>   Feb 9 2004:
>     "This is really an unacceptable situation. I'm asking you, the
>     DPL, to do something about it.  If the current delegates cannot
>     or will not do what they need to do, *you* need to replace them."

And it was and the DPL did something about it.

>     http://lists.debian.org/debian-devel/2004/debian-devel-200402/msg00329.html
>   Jan 19 2004:
>     "Subject: KDE 3.1.5 Status Update - 20040119
>     Everything is in sid now, but the buildds FUCKING SUCK! The buildd
>     admins must be incompetent or on crack."
>     http://lists.debian.org/debian-devel/2004/debian-devel-200401/msg01629.html
>   Dec 30 2003:
>     "Needless to say, such a contract does not exist.  Well, there's
>     the DMUP[1], but its focus is on punishing abuse of the system
>     by users[2].  As I said in a previous DPL platform of mine, in a
>     plank that not too many people seemed to care about :), the DMUP
>     treats the users more as wayward children than as the parterns and
>     colleagues of the Debian System Administrators."
>     http://lists.debian.org/debian-devel/2003/debian-devel-200312/msg02309.html

That sounds pretty off topic to the buildd issue.

>   Dec 14 2003:
>     "This is an official complaint about the current buildd situation.
>     I really feel discriminated by this situation.
>     It seems as if the archs that are managed by Ryan Murray are preferred
>     in their restoring process [...]
>     I get the impression that there is some sort of a "Debian clan"
>     that controls some important positions of the Debian project and
>     that is protecting itself from being influence by the outside."

Which is what happened and your own graphs should show that to
you. The fact that James and Ryan seem to talk to each other might
explain it.

>     http://lists.debian.org/debian-devel/2003/debian-devel-200312/msg01218.html
>   Nov 14 2003:
>     "Subject: MIPS port backlog, autobuilder machines and some arrogance
>     As a result and a sort of protest, I'll stopped my m68k buildd,
>     because I don't know m68k that much to be of any help for this port
>     anymore. Therefore my m68k isn't needed anymore as my offered mips
>     machine isn't needed for the mips port or the Debian project at all."

Could you post the mail causing that too please. Thats just a heated
reaction to another mail.

>     http://lists.debian.org/debian-devel/2003/debian-devel-200311/msg01114.html
>   Nov 12 2003: (one just for Mathieu)
>     "Do you really think that I'm the only person to find UNACCEPTABLE
>     that a package in mail suggests packages in non-free, which ARE
>     http://lists.debian.org/debian-devel/2003/debian-devel-200311/msg00917.html
>   Nov 11 2003:
>     "Subject: ftpmaster accepts packages that have been rejected a few
>     days ago I would like to ask other maintainers: What would you do
>     if you have had packages repeatedly rejected while others get the
>     very same packages [..] accepted without problems? Retire? Stop
>     uploading? Asking other people to upload for you? Try to be more
>     submissive towards powerful people on un-electable posts?"

Well, did you reject packages from one person but not when uploaded by
another? That realy sounds like a big screwup on your side.

>     http://lists.debian.org/debian-devel/2003/debian-devel-200311/msg00742.html
>   Oct 08 2003:
>     "Subject: Quote: Debian and Democracy at Advocato.org
>     Debian and Democracy
>     Two unrelated words. From experience.
>     James Troup gets to decide who can get an account all by
>     himself. Project leader changes, but he doesn't change. He is the
>     administrator of everything. He is Mr. Debian."

In the end its James who enacts the final step of making someone a
DD. He is that last hurdle (?) to overcome.  Only the point that he
could be replaced but no better (or near equal) candidate has
voluntered is missing.

>     http://lists.debian.org/debian-devel/2003/debian-devel-200310/msg00832.html
>   Aug 25 2003:
>     "Subject: Ping for DAM, DAM is MIA? Mail alias broken?"
>     http://lists.debian.org/debian-devel/2003/debian-devel-200308/msg02179.html

Thats mine. And i still haven't gotten a pong back. No "fuck off", no
"Gratulations, your a DD", no "hi you too" on irc.


All other people CCed on that mail have responded withing 2 days.

>   Aug 08 2003:
>     "In the next few weeks, I will replace my GPG key with a new
>     one. Since I suspect that the time it takes for a key to be exchanged
>     in the key ring needs to be measured in weeks or months[1], [...]"

Is he wrong? I heart it can take that long.

>     http://lists.debian.org/debian-devel/2003/debian-devel-200308/msg00761.html
>   Aug 06 2003:
>     "I am currently on NM process.
>     We believe we could be helpful. However, We are trying to be cut off
>     from that project. Totally this is agaist prejudice on Policies.. and
>     DFSG.
>     Debian Maintainers are becoming too elite. However, outside world
>     becoming more excluded. And Debian finally is becoming so obsolete."
>     http://lists.debian.org/debian-devel/2003/debian-devel-200308/msg00443.html
>   Jul 13 2003:
>     "Subject: Excessive wait for DAM - something needs to be done
>     This is absolutely crazy!  If there's a good reason for these
>     people to be waiting this long, update their status page and let
>     them know why."

Don't you think its crazy that

5 applicants have been waiting for two years or longer.
3 of them have a comment from DAM as to why.

24 applicants have been waiting for a year or longer. 
13 of them have a comment from DAM as to why.

32 applicants have been waiting for 6 months or longer.
16 of them have a comment from DAM as to why. 

I find that a well researched mail.

Also following the mail nothing happend leading to my cry for help
above for example. And finally I got help, guess from whom. Of course
from our DPL again (functioning as Front-Desk).

>     http://lists.debian.org/debian-devel/2003/debian-devel-200307/msg00839.html
>   May 10 2003:
>     "It's really fun to watch these large threads based on a reasonably
>     small lack of discernment being escalated to disconcerting incidents."
>     http://lists.debian.org/debian-devel/2003/debian-devel-200305/msg00665.html
> I tried to limit it to threads that specifically attack one of the roles
> James fills, or decisions James has made, although since I happened
> across one that Mathieu was involved in, I figured I'd include that too.

And showed several that have very good cause.

> Personally, I don't have a major problem with people loudly declaring
> things are unacceptable, or including some snide asides, but when you
> get to the point where people are being attacked every few weeks, on
> issues where sober judgement indicates they made a reasonable decision,
> well, I don't think that's the way we should be operating, and as long
> as they continue to make judgements that seem reasonable upon reflection,
> I think the changes we should make are to the people doing the complaining
> rather than the delegates.
> Further to that point, it's not reasonable to think all these sorts of
> overinflated complaints are related to James. See, eg:
>   Nov 15 2003:
>   "Subject: Example of really nasty DD behavior
>   This nasty behavior shows that being an officiel Debian developer does
>   not mean quality.
>   I was writing an ITP when you posted that and suddendly, saw all my
>   work preparation turned into ashes."
>   http://lists.debian.org/debian-devel/2003/debian-devel-200311/msg01218.html
> I don't think this sort of attitude is appropriate for our lists;
> I don't think we should tolerate it, let alone encourage it.
> > I'm not sure that Nathaniel is trying to impose something, I think he
> > truly sees himself as blocked for contributing by two persons.
> I'm sure he does; and as a consequence he's trying to force those persons
> out of the way.
> It's fine that he feels blocked. I feel blocked from getting testing
> working as well as it should because the security team aren't willing
> to support it. Every now and then I try to convince them to change their
> minds. So far they haven't, and don't look like ever doing so, but that
> doesn't make them bad people, and no matter what I want a difference
> answer, or how much I might know about their job, they're the ones in
> the best position to make that call. And until I do the job myself or
> convince someone else to do it, and demonstrate that it's doable, I've
> got no cause to expect _any_ assistance from the security team.
> > Writing
> > to this mailing-list is a way to go public, which is something that is
> > usually done when it is no longer possible to discuss in private.
> No, going to the mailing-list is a way to pressure the individual into
> doing something they don't want. Sometimes that's appropriate, but it's
> very rare; certainly much rarer than actually happens.

In many cases above its the lack of communication forcing the
complainy to go public since _NO_ private discussion is possible.

> > Unfortunately, in many case, going public does not help much, it is
> > just an escalation of the crisis. Said to say, crisis should be public
> > at first to avoid the problem.
> If nothing happens to improve the situation, and Debian doesn't collapse
> into ashes, it can't have been much of a crisis, can it? If "going public"
> doesn't help, there's not much point doing it at all, is it?
> > And then, James and Ryan should study their proposal and reply step by
> > step precisely, by avoiding any references to what was not possible
> > because of bla bla (the past). 
> "As long as James and Ryan are forced to do something they see as useless
> and a waste of time, I'm happy."
> Why do you think you need to know what James' and Ryan's response to
> individual suggestions are? Why do you think your competent to judge
> the competency of any responses?

Because otherwise you have a suggestion box thats the trashcan.

> > For instance, the title of the thread is: Debian needs more
> > buildds. It has offers. They aren't being accepted. There are 3
> > questions here, with apparently no clear answer. 
> Uh, those are three statements, one of which is an answer, and which
> contradicts the claim made in the first statement. For future reference,
> you can recognise questions by the use of what's called a "question
> mark". It looks like this: "?".
> > > Again: if a developer thinks Ryan or James has made the wrong decision,
> > > then they can present the case of why that should be overruled to the
> > > technical ctte. If a non-developer can't convince a developer to do that
> > > for them, then they should be ignored.
> > I agree with that. I think anybody should have the right to express a
> > point of view on Debian. 
> Certainly. Do you think that it's reasonable for some opinions to be
> expressed somewhere other than debian-devel@lists.debian.org? Do you
> think it's unreasonable to expect non-developers to use private emails
> to try to convince a developer of their case so that if they're wrong,
> it can be explained privately, rather than acting as a distraction for
> many developers?
> Cheers,
> aj

It has been expressed privately over and over without any reaction in
many cases you mentioned. Read the complains, many aren't about a
"wrong" decision but about _no_ decision being made.


Reply to: