Home > Archive > PHP PEAR Questions and Answers > February 2006 > Re: [PEAR-QA] Re: New package
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Re: [PEAR-QA] Re: New package
|
|
| Arnaud Limbourg 2006-02-10, 7:56 am |
| Pierre wrote:
> On 10 Feb 2006 10:24:43 -0000
> bolk@hitv.ru wrote:
>
>
> Can you please commit before any release?
>
> Why is it beta? It must be alpha.
>
> Please remove this release and release it again once you fixed these
> problems.
>
> --Pierre
>
Hello,
To the best of my knowledge the first release must not be stable but
nothing says it has to be alpha, right ?
Arnaud.
| |
| bertrand Gugger 2006-02-10, 7:56 am |
| Arnaud Limbourg wrote:
> Pierre wrote:
>
>
A reliable repository should be a rule, sure !
Also tagging it in the same time ... so we can measure evolutions.
>
> Hello,
>
> To the best of my knowledge the first release must not be stable but
> nothing says it has to be alpha, right ?
>
> Arnaud.
>
Perhaps I get it wrong but 0.x.0 should be alpha for the numbering point
of view.
I'm bad the same as I released Text_Wiki_BBCode as 0.0.x but alpha...
I would suggest, as I'm now some "illustrious" helper :) that pear
packager warns about such ill state/version. No, I don't have the patch
yet :)
à+
--
toggg
| |
| Florian Anderiasch 2006-02-10, 7:56 am |
| Hello folks,
I don't know exactly how it is handled in the main php source, but there
a great mass of people seem to have general cvs karma.
Now how about granting people the rights to other packages if the
maintainer isn't reachable for a while, without making them maintainer.
So they can fix bugs in cvs, but not release new versions, this should
still be done by an official maintainer.
The bad: I can't think of an automatic process of switching
"half-orphaned" packages to "write for some" or even "write for
everyone", but then I can't really evaluate if it's not better to stay
with the current process.
But when I look at several other projects like Debian stable (which I'd
compare a bit to PEAR stable, because no BC breakage and it should be
really reliable) - where security updates are often made on the same day
- and now I read something about "didn't answer in two w s, let's wait
another" - although I don't know if that are critical bugs that Anish
talked about.
I also know that PEAR has this strict maintainer for some reasons
(although I surely don't know them all) and not a policy that ever PEAR
developer can access any packages.
Well, much talking - but the most traffic here is imho the discussion
about when to appoint a new package maintainer, perhaps there could be
some improvements.
Greetings,
Florian
| |
| Lukas Smith 2006-02-10, 7:56 am |
| Florian Anderiasch wrote:
> Hello folks,
> I don't know exactly how it is handled in the main php source, but there
> a great mass of people seem to have general cvs karma.
thats for historical reasons mainly ..
anyways in your scenario you can push patches through pear-qa or even
get karma if required through the pear-qa/pear-group.
regards,
Lukas
| |
| bertrand Gugger 2006-02-10, 6:57 pm |
| Bonjour,
Florian Anderiasch wrote:
>Hello folks,
>I don't know exactly how it is handled in the main php source, but there
>a great mass of people seem to have general cvs karma.
>
>
Not a mass, but several. I belong to them
>Now how about granting people the rights to other packages if the
>maintainer isn't reachable for a while, without making them maintainer.
>
>
It's a QA responsibilty, discussed and granted a while ago.
I mean, nobody with general access is allowed to touch any other dev
package.
Only on request from QA (whose I don't belong), as I did on 2 packages
in the past.
>So they can fix bugs in cvs, but not release new versions, this should
>still be done by an official maintainer.
>
>
In this case, the general karma is only used to commit code.
The release is made by QA.
>The bad: I can't think of an automatic process of switching
>"half-orphaned" packages to "write for some" or even "write for
>everyone", but then I can't really evaluate if it's not better to stay
>with the current process.
>
>
Nothing is "automatic" here, it's each time disscussed in this very list.
>But when I look at several other projects like Debian stable (which I'd
>compare a bit to PEAR stable, because no BC breakage and it should be
>really reliable) - where security updates are often made on the same day
>- and now I read something about "didn't answer in two ws, let's wait
>another" - although I don't know if that are critical bugs that Anish
>talked about.
>
>
I believe security updates can be very quick, and for obvious reasons
not necessarly discussed here, unfortunately we did not have some
occasions to talk about some war. It's a guerilla process.
>I also know that PEAR has this strict maintainer for some reasons
>(although I surely don't know them all) and not a policy that ever PEAR
>developer can access any packages.
>
>Well, much talking - but the most traffic here is imho the discussion
>about when to appoint a new package maintainer, perhaps there could be
>some improvements.
>
>
Yes
>Greetings,
>Florian
>
>
>
à+
--
toggg
| |
| bertrand Gugger 2006-02-21, 6:58 pm |
| Bonjour,
Confirmed, some anarchists think they may use it.
So, yes, we need a real policy about that.
Sorry, I'm too ingenuous.
For this case (Text_Wiki), it's well known I'm not inactive, more I
flooded pear-dev and Text_Wiki lists with what was currently in preparation.
I cannot take into account wild commits in the middle of such change,
but I permanently call for some help and check back by colleagues.
I tend to think the thing is not mine, seeems some don't.
I stop checkout the cvs until it comes back coherent, so won't be able
to commit the same.
--
toggg
-------- Original Message --------
Subject: Re: cvs: pear /Text_Wiki/Text/Wiki/Parse/Default Code.php
Date: Tue, 21 Feb 2006 15:04:53 +0100
From: bertrand@toggg.com (bertrand Gugger)
To: pear-cvs@lists.php.net,Del Elson <delatbabel@php.net>
CC: pear-cvs@lists.php.net
Newsgroups: php.pear.cvs,php.pear.cvs
References: <cvsdelatbabel1140509459@cvsserver>
Del Elson wrote:
> delatbabel Tue Feb 21 08:10:59 2006 UTC
>
> Modified files:
> /pear/Text_Wiki/Text/Wiki/Parse/Default Code.php Log:
> Whitespace changes only.
>
Sorry, who are you ?
You're not a maintainer of Text_Wiki, please rollback.
--
toggg
Florian Anderiasch wrote:
>Hello folks,
>I don't know exactly how it is handled in the main php source, but there
>a great mass of people seem to have general cvs karma.
>
>Now how about granting people the rights to other packages if the
>maintainer isn't reachable for a while, without making them maintainer.
>
>So they can fix bugs in cvs, but not release new versions, this should
>still be done by an official maintainer.
>
>The bad: I can't think of an automatic process of switching
>"half-orphaned" packages to "write for some" or even "write for
>everyone", but then I can't really evaluate if it's not better to stay
>with the current process.
>
>But when I look at several other projects like Debian stable (which I'd
>compare a bit to PEAR stable, because no BC breakage and it should be
>really reliable) - where security updates are often made on the same day
>- and now I read something about "didn't answer in two ws, let's wait
>another" - although I don't know if that are critical bugs that Anish
>talked about.
>I also know that PEAR has this strict maintainer for some reasons
>(although I surely don't know them all) and not a policy that ever PEAR
>developer can access any packages.
>
>Well, much talking - but the most traffic here is imho the discussion
>about when to appoint a new package maintainer, perhaps there could be
>some improvements.
>
>Greetings,
>Florian
>
>
>
| |
| bertrand Gugger 2006-02-21, 6:58 pm |
| I rolled it back.
| |
| Arnaud Limbourg 2006-02-23, 3:56 am |
| There indeed should not be wild commits like that without checking with
the maintainer first.
As for Florian's idea, it would be great but the karma management is
manual atm.
There is no nice system to manage karma more easily atm. (the proposal
http://pear.php.net/pepr/pepr-proposal-show.php?id=279 can help to some
extent).
Remember to be nice and not commit blindly (even if it is just whitespace).
Arnaud.
bertrand Gugger wrote:
> Bonjour,
> Confirmed, some anarchists think they may use it.
> So, yes, we need a real policy about that.
> Sorry, I'm too ingenuous.
> For this case (Text_Wiki), it's well known I'm not inactive, more I
> flooded pear-dev and Text_Wiki lists with what was currently in
> preparation.
> I cannot take into account wild commits in the middle of such change,
> but I permanently call for some help and check back by colleagues.
> I tend to think the thing is not mine, seeems some don't.
> I stop checkout the cvs until it comes back coherent, so won't be able
> to commit the same.
> --
> toggg
>
> -------- Original Message --------
> Subject: Re: cvs: pear /Text_Wiki/Text/Wiki/Parse/Default Code.php
> Date: Tue, 21 Feb 2006 15:04:53 +0100
> From: bertrand@toggg.com (bertrand Gugger)
> To: pear-cvs@lists.php.net,Del Elson <delatbabel@php.net>
> CC: pear-cvs@lists.php.net
> Newsgroups: php.pear.cvs,php.pear.cvs
> References: <cvsdelatbabel1140509459@cvsserver>
>
> Del Elson wrote:
>
>
> Sorry, who are you ?
> You're not a maintainer of Text_Wiki, please rollback.
> --
> toggg
>
> Florian Anderiasch wrote:
>
>
|
|
|
|
|