FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Gentoo > Gentoo Development

 
 
LinkBack Thread Tools
 
Old 01-12-2010, 08:39 PM
Denis Dupeyron
 
Default Last rites: net-nntp/inn

2010/1/12 Tomáš Chvátal <scarabeus@gentoo.org>:
> Dont be joking,
[...]

Mmmh? Take a deep breath, a long walk, a large beer, or whatever
works. Because you need it.

Denis.
 
Old 01-12-2010, 09:24 PM
Duncan
 
Default Last rites: net-nntp/inn

Ben de Groot posted on Tue, 12 Jan 2010 21:35:45 +0100 as excerpted:

> 2010/1/12 Markos Chandras <hwoarang@gentoo.org>:
>> If you feel like it, become a proxy-maintainer and poke a developer to
>> put your ebuilds on tree. Have you ever heard of that ?
>
> Proxy-maintainership should be given a MUCH higher profile in Gentoo, in
> my opinion. It is a virtually unknown option.

Yay! =:^)

> Another thing that works in my experience, but this is up to the
> herds/projects, is having an official overlay where devs and users can
> work closely together.

Again! =:^)

> But I also believe we need a better structure to handle
> maintainer-needed, maintainer-wanted and nominally maintained but
> ignored packages. Maybe we should form a team, which would be dedicated
> to take care of such things, and which would have a review policy for
> user submitted ebuilds and patches in bugzilla. A bit like treecleaners,
> but bringing life instead of death. What do you think?

Well, sunrise was supposed to be that for packages not yet in the tree.
The devs shepherding that work hard with the users doing the ebuilds to
get them up to tree quality, so they're ready for devs to take on with
little work (or to go into proxy maintainership), when it's that
package's time. I know I've used a handful of sunrise packages that
ultimately ended up in the tree, which is pretty good, considering I
don't even have the sunrise overlay enabled unless there's something I
specifically want from it.

--
Duncan - List replies preferred. No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master." Richard Stallman
 
Old 01-12-2010, 09:37 PM
Duncan
 
Default Last rites: net-nntp/inn

Richard Freeman posted on Tue, 12 Jan 2010 14:07:38 -0500 as excerpted:

> On 01/12/2010 01:30 PM, Markos Chandras wrote:
>> IMHO ( this is not a treecleaners@ opinion, i m just talking for my
>> self ), announcing and masking a package is a good way to inform and
>> wake up everybody to take care of this package if they really really
>> want to stay on portage.
>
> I agree with the announce part, and the THREAT of masking. I just don't
> think that the masking should happen at the same time as the
> announcement.

FWIW, I feel for the treecleaners. It's a job with little thanks and
lots of chance to make someone mad at you, but I'm glad /someone's/ doing
it! =:^)

So going with this idea... Isn't the treecleaner masking 30-day at
present? What about extending that just a bit, to 5 weeks total, while
reducing the actual masking to 4 weeks, with the extra week a wait time
between the traditional last-rites mail and the masking?

In the case of the INNs of the tree, that should prevent masking
entirely, since popular packages will certainly have someone raising the
roof on just the warning, within a day or two. That was certainly the
case here. No masking means ordinary users won't have to ever know it
happened.

Or is that extra step going to throw a spanner into the works for
treecleaners? As I said, I definitely appreciate the job they're doing,
and wouldn't want to make their life harder. But this could well reduce
the fallout when the INNs of the tree come up, and that just might make
it easier to handle, even if tracking that extra step /is/ a bit more
work.

Treecleaners?

--
Duncan - List replies preferred. No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master." Richard Stallman
 
Old 01-13-2010, 12:18 AM
Arnaud Launay
 
Default Last rites: net-nntp/inn

Le Tue, Jan 12, 2010 at 10:37:19PM +0000, Duncan a crit:
> FWIW, I feel for the treecleaners. It's a job with little
> thanks and lots of chance to make someone mad at you, but I'm
> glad /someone's/ doing it! =:^)

Yeah. I'm glad each time I see old things getting deleted,
abandoned software and such. So, yeah, thanks, treecleaners.
Your job is as easy as a sysadmin's one: no one knows you exist,
except when someone needs to scream at you...

> In the case of the INNs of the tree, that should prevent
> masking entirely, since popular packages will certainly have
> someone raising the roof on just the warning, within a day or
> two. That was certainly the case here. No masking means
> ordinary users won't have to ever know it happened.

Well, as it happens, I knew it was being masked because I got the
mail warning from gentoo-dev-announce, which is unfortunately a
bit badly advertised.

Anyway, I would have scratched my head far, far more if I had to
understand WTF portage would complain about an inn masked... I
don't care when it's small, unused package -- last time it was
lprof, and I didn't really care, it was a bit still here from a
package I installed years ago, and which passed through depclean.

So, what about something like:
* mail on gentoo-dev-announce, saying "heads up. mask in one week"
* one week later, mask and "classical" mail "foo/bar masked"

I have absolutely no idea how much work it requires, so I won't
complain if TC says it's too complicated/unpratical/etc.

BTW, I have no knowledge of the concept of proxy-maintainer, I'll
look at it tomorrow, it's 2am here... I don't even think I
ever heard of it before, but I didn't brush my gentoo-fu for a
few years, that may explain...

Arnaud.
 
Old 01-13-2010, 04:45 AM
Jeroen Roovers
 
Default Last rites: net-nntp/inn

On Tue, 12 Jan 2010 21:51:28 +0100
Tomáš Chvátal <scarabeus@gentoo.org> wrote:

> > you need to fix your filter then. an "open bug" is not an
> > acceptable reason for masking a package. if you're going to clean
> > a package, you need to research actual reasons to mask & punt.
> > -mike
> Dont be joking,
> Your approach of adding new packages to main tree is that you add them
> with empty metadata.xml and we have to remove them in few years
> because they are steaming piles of bugs...

Er, say whah? Flinging mud?

Mike's got a very valid point, in that you don't mask a package because
of an open bug. All the rest of what you added below is about
--as-needed, which doesn't apply to the bug in case and which is still
no valid reason to mask a package anyway. End of discussion, plz?


Yah,
jer
 
Old 01-13-2010, 04:48 AM
Jeroen Roovers
 
Default Last rites: net-nntp/inn

On Tue, 12 Jan 2010 22:37:19 +0000 (UTC)
Duncan <1i5t5.duncan@cox.net> wrote:

> So going with this idea... Isn't the treecleaner masking 30-day at
> present? What about extending that just a bit, to 5 weeks total,
> while reducing the actual masking to 4 weeks, with the extra week a
> wait time between the traditional last-rites mail and the masking?

No, masking for removal should take 30 days. I strongly feel that
before treecleaner@ does any masking, an announcement should go to
-dev@ and -announce@ a pretty long time in advance, maybe two months,
especially with the two cockups a month that I am counting now.


jer
 
Old 01-13-2010, 04:52 AM
Jeroen Roovers
 
Default Last rites: net-nntp/inn

On Wed, 13 Jan 2010 02:18:59 +0100
Arnaud Launay <asl@launay.org> wrote:

> I have absolutely no idea how much work it requires, so I won't
> complain if TC says it's too complicated/unpratical/etc.

rm -r * is easy.

> BTW, I have no knowledge of the concept of proxy-maintainer, I'll
> look at it tomorrow, it's 2am here... I don't even think I
> ever heard of it before, but I didn't brush my gentoo-fu for a
> few years, that may explain...

It should probably be documented at the official places [1][2].


[1] http://devmanual.gentoo.org
[2] http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml [3]
[3] Why weren't these merged years ago?
 
Old 01-13-2010, 07:05 AM
Duncan
 
Default Last rites: net-nntp/inn

Jeroen Roovers posted on Wed, 13 Jan 2010 06:48:18 +0100 as excerpted:

> On Tue, 12 Jan 2010 22:37:19 +0000 (UTC) Duncan <1i5t5.duncan@cox.net>
> wrote:
>
>> So going with this idea... Isn't the treecleaner masking 30-day at
>> present? What about extending that just a bit, to 5 weeks total, while
>> reducing the actual masking to 4 weeks, with the extra week a wait time
>> between the traditional last-rites mail and the masking?
>
> No, masking for removal should take 30 days. I strongly feel that before
> treecleaner@ does any masking, an announcement should go to -dev@ and
> -announce@ a pretty long time in advance, maybe two months, especially
> with the two cockups a month that I am counting now.

30-day masking /does/ give the folks updating once a month at least one
warning, so I can see the case for that.

But... IMO extending the pre-mask warning another full 30 days... is
asking for trouble going the /other/ way. It's not urgent enough to
require immediate action... which can unfortunately cause people to put
it off and forget about it. Then it's masked 30 days later... and we're
back where we were!

So I'd say a week to 15 days pre-mask warning... and 14-15 days is
stretching it. A week is just about right, short enough to require
urgent action and thus front-burnering, long enough that if there's a
clear objection to be made, it should be very clear within 2-3 days, and
there's another 4 days to actually do something about it before the
masking.

Just my (non-gentoo-dev) opinion, however. I'm acutely aware I'm not the
one doing the work, so if that opinion doesn't match dev-reality, feel
free to ignore it.

--
Duncan - List replies preferred. No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master." Richard Stallman
 
Old 01-13-2010, 01:24 PM
Mike Frysinger
 
Default Last rites: net-nntp/inn

On Tuesday 12 January 2010 15:51:28 Tomáš Chvátal wrote:
> Your approach of adding new packages to main tree is that you add them
> with empty metadata.xml and we have to remove them in few years because
> they are steaming piles of bugs...

not that this is relevant at all to my point, but when i notice such bugs,
i'll fix them. but if no one cc's me, then i dont generally notice.

> Lack of maintainer

incorrect

> open bugs are valid reasons.

the *type* of bug matters, not the existence of an open bug. take any number
of the trivial QA bugs open related to an extra empty dir or split docs --
none are valid for punting an otherwise perfectly working package.

> And since WE want to enable as-needed as default at some time we need to
> work on the bugs

which isnt going to happen
-mike
 
Old 01-13-2010, 02:06 PM
Arnaud Launay
 
Default Last rites: net-nntp/inn

Le Wed, Jan 13, 2010 at 06:52:09AM +0100, Jeroen Roovers a crit:
> > BTW, I have no knowledge of the concept of proxy-maintainer, I'll
> It should probably be documented at the official places [1][2].
> [1] http://devmanual.gentoo.org
> [2] http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml [3]
> [3] Why weren't these merged years ago?

Well, nothing about "proxy-maintainer" here.
I just found one doc at
http://dev.gentoo.org/~antarus/projects/proxy-maint/

which kind of explain what is a proxy maintainer (more or less),
but does not explain how to become one...

Le Tue, Jan 12, 2010 at 09:49:10PM +0200, Markos Chandras a crit:
> If you feel like it, become a proxy-maintainer and poke a
> developer to put your ebuilds on tree. Have you ever heard of that ?

No problem. Just tell me who I need to poke Would that be
antarus, saying "hey, I'm mostly in servers, how may I be of
service" ?

Arnaud.
 

Thread Tools




All times are GMT. The time now is 11:37 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org