[gentoo-dev] GLEP system worthwhile?

[gentoo-dev] GLEP system worthwhile?

[Grant Goodyear]

[-- Attachment #1: Type: text/plain, Size: 1655 bytes --]

It's been a bit over a year and a half since glep.gentoo.org came
online, and GLEPs became the standard method for proposing significant
changes to Gentoo.  I think it's time to assess what works, and what
doesn't, about this system.  To date 30 GLEPs have been approved by the
GLEP editors that involve something other than the GLEP process itself.
Six of those GLEPs have been marked "Final", meaning they have been
successfully implemented.  Seven more have been "Accepted", meaning
that the idea in the GLEP has been approved, but the implementation has
not yet been completed.  Five are currently in "Draft" status, not yet
having been submitted for approval.  One has been rejected.  Eleven
GLEPs failed to get enough traction to be either approved or rejected,
and are thus "deferred".  Do these statistics mean that the program is
working well, working poorly, or failed utterly?  My personal opinion
is that even with the substantial number of timed-out GLEPs, the
program is still a modest success because these GLEPs provide a record
of notable, generally well-written proposals, which is a vast
improvement on seeing the same half-thought-out ideas appear on the
mailing lists time and time again.  On the other hand, one might well
argue that fairly few substantial accomplishments have come about from
GLEPs, so perhaps GLEPs just add another bureaucratic impediment.

Thoughts / comments?  I promise not to bite anyone's head off this time!

-g2boojum-
-- 
Grant Goodyear	
Gentoo Developer
g2boojum@gentoo.org
http://www.gentoo.org/~g2boojum
GPG Fingerprint: D706 9802 1663 DEF5 81B0  9573 A6DC 7152 E0F6 5B76

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 719 bytes --]

On Fri, 11 Mar 2005 08:39:51 -0600 Grant Goodyear <g2boojum@gentoo.org>
wrote:
| Thoughts / comments?  I promise not to bite anyone's head off this
| time!

I don't think there's anything wrong with the system, except that
certain managers can in effect put a GLEP on hold indefinitely because
their favourite editor sucks.

I'd rather have GLEPs than half-baked kook schemes. We've all seen how
long they take to un-screw-up... Well, actually, we haven't, since
they're still not un-screwed, but we've at least seen how much mess they
make.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Daniel Ostrow]

On Fri, 2005-03-11 at 15:02 +0000, Ciaran McCreesh wrote:
> On Fri, 11 Mar 2005 08:39:51 -0600 Grant Goodyear <g2boojum@gentoo.org>
> wrote:
> | Thoughts / comments?  I promise not to bite anyone's head off this
> | time!
> 
> I don't think there's anything wrong with the system, except that
> certain managers can in effect put a GLEP on hold indefinitely because
> their favourite editor sucks.
> 
> I'd rather have GLEPs than half-baked kook schemes. We've all seen how
> long they take to un-screw-up... Well, actually, we haven't, since
> they're still not un-screwed, but we've at least seen how much mess they
> make.
> 

Agreed, the GLEP system is good as it stands. If you look at the recent
GLEPs that came through like the GLEPs 30 & 31 which went by (with one
minor exception *cough* nano *cough*) very quickly due to developer
interest. And others, like GLEP 19 for example, are under active
development. The problem is not with the GLEP system it's with the
developers who are either too lazy/busy to write up a proper GLEP for
things that need it, and thus don't want whatever it is to be
implemented too badly, or loose interest in one way or another in one
they have already written and neither of these can be fixed by changing
policy.

--Dan

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Ian Leitch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Grant Goodyear wrote:
| It's been a bit over a year and a half since glep.gentoo.org came
| online, and GLEPs became the standard method for proposing significant
| changes to Gentoo.  I think it's time to assess what works, and what
| doesn't, about this system.  To date 30 GLEPs have been approved by the
| GLEP editors that involve something other than the GLEP process itself.
| Six of those GLEPs have been marked "Final", meaning they have been
| successfully implemented.  Seven more have been "Accepted", meaning
| that the idea in the GLEP has been approved, but the implementation has
| not yet been completed.  Five are currently in "Draft" status, not yet
| having been submitted for approval.  One has been rejected.  Eleven
| GLEPs failed to get enough traction to be either approved or rejected,
| and are thus "deferred".  Do these statistics mean that the program is
| working well, working poorly, or failed utterly?  My personal opinion
| is that even with the substantial number of timed-out GLEPs, the
| program is still a modest success because these GLEPs provide a record
| of notable, generally well-written proposals, which is a vast
| improvement on seeing the same half-thought-out ideas appear on the
| mailing lists time and time again.  On the other hand, one might well
| argue that fairly few substantial accomplishments have come about from
| GLEPs, so perhaps GLEPs just add another bureaucratic impediment.
|
| Thoughts / comments?  I promise not to bite anyone's head off this time!
|
| -g2boojum-

The effort required to get GLEPs even posted on glep.g.o is enough to
put some people off.

I propose that a dev is free to upload a GLEP in any state of
completeness/correctness as they wish, devs can then vote on said GLEP.
If the GLEP author makes changes, devs then vote again on that revision.
This system also gives power back to the devs (and users if they're
allowed to vote) and away from the GLEP editors.

Regards,
Ian Leitch

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMb4pefZ4eWAXRGIRAktvAJ4r6tZAJyaaxtkCgtYK/qQ+il1KLwCeIpJ0
XPhsHBWN0H/zS+JhZKbtWWU=
=pU/t
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 960 bytes --]

On Fri, 11 Mar 2005 15:50:01 +0000 Ian Leitch <port001@gentoo.org>
wrote:
| The effort required to get GLEPs even posted on glep.g.o is enough to
| put some people off.

Lies! GLEP 31 took me half an hour to write, and I was mildly
intoxicated at the time.

| I propose that a dev is free to upload a GLEP in any state of
| completeness/correctness as they wish, devs can then vote on said
| GLEP. If the GLEP author makes changes, devs then vote again on that
| revision. This system also gives power back to the devs (and users if
| they're allowed to vote) and away from the GLEP editors.

'Any' state is asking for trouble. You can post non-finished GLEP plans
to the list in any state, but I'd expect things making it onto the site
to be at least in reasonable draft state.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ian Leitch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ciaran McCreesh wrote:
| | I propose that a dev is free to upload a GLEP in any state of
| | completeness/correctness as they wish, devs can then vote on said
| | GLEP. If the GLEP author makes changes, devs then vote again on that
| | revision. This system also gives power back to the devs (and users if
| | they're allowed to vote) and away from the GLEP editors.
|
| 'Any' state is asking for trouble. You can post non-finished GLEP plans
| to the list in any state, but I'd expect things making it onto the site
| to be at least in reasonable draft state.

OK, that's fair enough, but just so long as the author can post a
revised version and have devs vote on it again. The main reason for this
is that it's hard to track changes made to a GLEP in a massive long -dev
thread.

Regards,
Ian Leitch

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMcLVefZ4eWAXRGIRAlXHAJ9nVGcarUlJoEUjb57N3T9ZzDqWdwCfexRK
fY/QdqTl/GbAbwFOOhT9iY4=
=sz/5
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 712 bytes --]

On Fri, 2005-03-11 at 16:09 +0000, Ian Leitch wrote:
> | 'Any' state is asking for trouble. You can post non-finished GLEP plans
> | to the list in any state, but I'd expect things making it onto the site
> | to be at least in reasonable draft state.
> 
> OK, that's fair enough, but just so long as the author can post a
> revised version and have devs vote on it again. The main reason for this
> is that it's hard to track changes made to a GLEP in a massive long -dev
> thread.

You just described the current process.

A GLEP can be modified and have the revision voted on as it stands.

-- 
Chris Gianelloni
Release Engineering - Strategic Lead/QA Manager
Games - Developer
Gentoo Linux

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ian Leitch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Chris Gianelloni wrote:
| You just described the current process.
|
| A GLEP can be modified and have the revision voted on as it stands.

Yes, but:

1) The revision has to go through one of 2 GLEP editors, what's wrong
with letting a dev do this themselves. I agree with genone on letting
devs have CVS access. Pkus it's not obvious on the GLEP main page which
GLEPs have been updated and require votes.

2) Voting is done on -dev, I can't think of a worse place to conduct a
vote. We need a web interface to count and clearly display votes for
each revision.

Regards,
Ian Leitch

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMdgHefZ4eWAXRGIRAoN3AJ0V4QuQEoXvf7jPhgQcgIVSv4qMwQCggfP4
w1pt2X5pevf2SkbHaBoz154=
=DdBJ
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 916 bytes --]

On Fri, 11 Mar 2005 17:40:23 +0000 Ian Leitch <port001@gentoo.org>
wrote:
| 1) The revision has to go through one of 2 GLEP editors, what's wrong
| with letting a dev do this themselves. I agree with genone on letting
| devs have CVS access. Pkus it's not obvious on the GLEP main page
| which GLEPs have been updated and require votes.

So you Cc: glep@g.o whenever you send an update to -dev. Big deal.

| 2) Voting is done on -dev, I can't think of a worse place to conduct a
| vote. We need a web interface to count and clearly display votes for
| each revision.

Uh, no. Voting is done at the manager meeting.

Incidentally, at least half the people voting on your GLEP won't have
read it, but that's an entirely different issue.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ian Leitch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ciaran McCreesh wrote:
~ > So you Cc: glep@g.o whenever you send an update to -dev. Big deal.

Not good enough, mailing lists suck to begin with.

| Uh, no. Voting is done at the manager meeting.

OK, I was wrong.. there IS a worse place to conduct a vote. What
percentage of devs attend the meetings? How many stick around till they
- -m the channel? Hell, half our developer base are probably asleep in bed
when those meetings take place!

Regards,
Ian Leitch


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMf9fefZ4eWAXRGIRAiByAJ43y1w8610fHqagOJGjpXbwY6XCIQCfbCrL
UbPZxaPDGxgSA856guxjeoA=
=wEFj
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 587 bytes --]

On Fri, 11 Mar 2005 20:28:15 +0000 Ian Leitch <port001@gentoo.org>
wrote:
| | Uh, no. Voting is done at the manager meeting.
| 
| OK, I was wrong.. there IS a worse place to conduct a vote. What
| percentage of devs attend the meetings? How many stick around till
| they - -m the channel? Hell, half our developer base are probably
| asleep in bed when those meetings take place!

Developers don't do the voting.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ian Leitch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ciaran McCreesh wrote:
| Developers don't do the voting.

I thought we were supposed to be an open organisation...*sigh*

Regards,
Ian Leitch
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMgciefZ4eWAXRGIRAk1RAJ9Bpb4LRukqwvkhtDitpXvYmABDhQCghiFW
M+n/QjUIAA5JbrukLC6pxo0=
=/6Pr
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Martin Schlemmer]

[-- Attachment #1: Type: text/plain, Size: 550 bytes --]

On Fri, 2005-03-11 at 21:01 +0000, Ian Leitch wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Ciaran McCreesh wrote:
> | Developers don't do the voting.
> 
> I thought we were supposed to be an open organisation...*sigh*
> 

Right, but haven't you followed the stuff about voting in the other
sekret mailing list, or in the meeting on Monday ?  Wont give all devs
votes, but should hopefully make coverage better.


-- 
Martin Schlemmer
Gentoo Linux Developer, Desktop/System Team Developer
Cape Town, South Africa


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Donnie Berkholz]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ian Leitch wrote:
| Ciaran McCreesh wrote:
| | Developers don't do the voting.
|
| I thought we were supposed to be an open organisation...*sigh*

Open doesn't mean everyone gets to decide. It means everyone gets to see
what's happening. Look at the kernel for an example of open development
in this manner.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFCMhaHXVaO67S1rtsRAha7AKDYEFDpSWdLyI9koreNjTYa3XHYgQCfXgWr
uH3EsCCh5x5uRNqLoDknqJw=
=xh3i
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 1300 bytes --]

On Fri, 2005-03-11 at 17:40 +0000, Ian Leitch wrote:
> Yes, but:
> 
> 1) The revision has to go through one of 2 GLEP editors, what's wrong
> with letting a dev do this themselves. I agree with genone on letting
> devs have CVS access. Pkus it's not obvious on the GLEP main page which
> GLEPs have been updated and require votes.

I could agree with this only if there were something in place that keeps
a new GLEP from being posted until it meets the proper layout.  I don't
want to see completely worthless entries that are still a work in
progress.  I would like it to at least be somewhat complete before I
spend time looking through them.

I'm thinking something like a barrier to entry, but not for editing.
Give the GLEP authors and a group of people accepted by the author
access to edit the GLEP.

> 2) Voting is done on -dev, I can't think of a worse place to conduct a
> vote. We need a web interface to count and clearly display votes for
> each revision.

I agree with you here.  Voting over a mailing list just isn't very
efficient.  Perhaps a login-restricted web page?  How about an app on
toucan?  "vote glep19 yes" or something equally as insane...

-- 
Chris Gianelloni
Release Engineering - Strategic Lead/QA Manager
Games - Developer
Gentoo Linux

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 507 bytes --]

On Fri, 11 Mar 2005 13:39:44 -0500 Chris Gianelloni
<wolf31o2@gentoo.org> wrote:
| I agree with you here.  Voting over a mailing list just isn't very
| efficient.  Perhaps a login-restricted web page?  How about an app on
| toucan?  "vote glep19 yes" or something equally as insane...

Haha. Sounds like a job for an eclectic module :)

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Marius Mauch]

[-- Attachment #1: Type: text/plain, Size: 1445 bytes --]

On Fri, 11 Mar 2005 15:58:22 +0000
Ciaran McCreesh <ciaranm@gentoo.org> wrote:

> On Fri, 11 Mar 2005 15:50:01 +0000 Ian Leitch <port001@gentoo.org>
> wrote:
> | I propose that a dev is free to upload a GLEP in any state of
> | completeness/correctness as they wish, devs can then vote on said
> | GLEP. If the GLEP author makes changes, devs then vote again on that
> | revision. This system also gives power back to the devs (and users
> | if they're allowed to vote) and away from the GLEP editors.
> 
> 'Any' state is asking for trouble. You can post non-finished GLEP
> plans to the list in any state, but I'd expect things making it onto
> the site to be at least in reasonable draft state.

Agreed, but being able to have potential/unfinished gleps available (so
you can check them by url, but not by clicking a link on the index)
would be nice so people who are asked by glep authors to
proofread/comment/poop on a new glep don't have to remember all the
custom d.g.o URLs. Another thing is that I'd like to update it directly
and not to have to send diffs/new versions to some gatekeeper instance.
So maybe direct commit access for devs (as mentioned in GLEP1) with some
special magic for the index page?

Marius

-- 
Public Key at http://www.genone.de/info/gpg-key.pub

In the beginning, there was nothing. And God said, 'Let there be
Light.' And there was still nothing, but you could see a bit better.

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Grant Goodyear]

[-- Attachment #1: Type: text/plain, Size: 794 bytes --]

Marius Mauch wrote: [Fri Mar 11 2005, 10:28:03AM CST]
> Agreed, but being able to have potential/unfinished gleps available (so
> you can check them by url, but not by clicking a link on the index)
> would be nice so people who are asked by glep authors to
> proofread/comment/poop on a new glep don't have to remember all the
> custom d.g.o URLs. Another thing is that I'd like to update it directly
> and not to have to send diffs/new versions to some gatekeeper instance.
> So maybe direct commit access for devs (as mentioned in GLEP1) with some
> special magic for the index page?

Sounds good to me.

-g2boojum-
-- 
Grant Goodyear	
Gentoo Developer
g2boojum@gentoo.org
http://www.gentoo.org/~g2boojum
GPG Fingerprint: D706 9802 1663 DEF5 81B0  9573 A6DC 7152 E0F6 5B76

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Donnie Berkholz]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ciaran McCreesh wrote:
> On Fri, 11 Mar 2005 15:50:01 +0000 Ian Leitch <port001@gentoo.org>
> wrote:
> | The effort required to get GLEPs even posted on glep.g.o is enough to
> | put some people off.
> 
> Lies! GLEP 31 took me half an hour to write, and I was mildly
> intoxicated at the time.

Not everyone is as brilliantly fast at ReST/XML as you. I can write the
thing in no time, but it takes forever to figure out the proper markup
for what I want to do with it.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMfiKXVaO67S1rtsRAokbAJ9lryNC5MjFPiCX5wLlpcGEGf34vgCfT1HR
Eteru/FthURDlp6nJa120Cw=
=PQiW
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 616 bytes --]

On Fri, 11 Mar 2005 11:59:06 -0800 Donnie Berkholz
<spyderous@gentoo.org> wrote:
| Not everyone is as brilliantly fast at ReST/XML as you. I can write
| the thing in no time, but it takes forever to figure out the proper
| markup for what I want to do with it.

There's no XML, it's just RST, which is pretty straight forward. Plus, I
heard a rumour that if you send it to Grant in plain text format then
he'll do the markup for you anyway...

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Donnie Berkholz]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ciaran McCreesh wrote:
> Plus, I
> heard a rumour that if you send it to Grant in plain text format then
> he'll do the markup for you anyway...

Bah, I hate those black-box systems. You never know what you're getting. =P
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMfyVXVaO67S1rtsRApYSAJ9UisZcnr4qTZMRt+zUKHPrDUNkUQCfbSnR
ZBvpJAaUuDexXg51OUOsyzs=
=rY9v
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 1119 bytes --]

On Fri, 2005-03-11 at 15:50 +0000, Ian Leitch wrote:
> This system also gives power back to the devs (and users if they're
> allowed to vote) and away from the GLEP editors.

Another abuse of power by the management trying to keep the little man
down!  Oh the injustice!

Sorry, man, but if you can't take the 30 minutes to write up a
well-written proposal in a readable and already established format, then
how can you possibly expect others to take the idea seriously?  (This
isn't meant to be pointing against you, Ian.)

The truth is that there are a lot of good ideas that die on the vine
simply because there isn't enough interest in them, or they aren't
thought out to a usable state.  I think we have also seen that sometimes
even a good idea just doesn't have enough momentum to keep it going.
Couple that with the fact that most of the work done on Gentoo doesn't
affect Gentoo on a global scope, and hence, doesn't require a GLEP, and
I would say the system has worked pretty well.

-- 
Chris Gianelloni
Release Engineering - Strategic Lead/QA Manager
Games - Developer
Gentoo Linux

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Grant Goodyear]

[-- Attachment #1: Type: text/plain, Size: 2222 bytes --]

Ian Leitch wrote: [Fri Mar 11 2005, 09:50:01AM CST]
> The effort required to get GLEPs even posted on glep.g.o is enough to
> put some people off.

Really?  It's usually just a matter of e-mailing me.  I read through
it, and if I see that the GLEP fails to answer some questions that I
predict will arise, I will generally mention it to the GLEP author and
suggest that such issues be addressed, even it is just to add a
section labeled "open questions".  Very occasionally I will mention to
an author that I don't think a GLEP is actually needed, and suggest
that the author just talk to the appropriate person and implement it.
I don't believe that I have ever outright refused a GLEP, or even added
more than a day to the time for it to be posted.

Most of these issues are moot, however, because in many cases GLEP
authors prefer to post an informal GLEP on -dev first, gather comments,
and then submit a formal GLEP after an initial revision.

> I propose that a dev is free to upload a GLEP in any state of
> completeness/correctness as they wish, devs can then vote on said GLEP.
> If the GLEP author makes changes, devs then vote again on that revision.
> This system also gives power back to the devs (and users if they're
> allowed to vote) and away from the GLEP editors.

It's worth noting that a GLEP author need not be a dev, so those
people, of course, would always require a surrogate to submit/revise
GLEPs.  As for devs being able to upload/revise GLEPs, I'm not opposed.
The reason that devs cannot update their own GLEPs right now is purely
technical: the GLEP page is part of the www tree, and that tree has
fairly strict permissions.  If opening up the GLEP directory isn't too
much of a pita for infra, I certainly won't oppose it.  I would still
prefer that GLEPs be run by one of the editors before being posted,
since we may be able to help, but I wouldn't insist on it.  I would be
very sad, though, if people took advantage of a more liberal policy to
post poorly-thought-out junk.

-g2boojum-
-- 
Grant Goodyear	
Gentoo Developer
g2boojum@gentoo.org
http://www.gentoo.org/~g2boojum
GPG Fingerprint: D706 9802 1663 DEF5 81B0  9573 A6DC 7152 E0F6 5B76

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ian Leitch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Grant Goodyear wrote:
~ > Really?  It's usually just a matter of e-mailing me.  I read through
| it, and if I see that the GLEP fails to answer some questions that I
| predict will arise, I will generally mention it to the GLEP author and
| suggest that such issues be addressed, even it is just to add a
| section labeled "open questions".  Very occasionally I will mention to
| an author that I don't think a GLEP is actually needed, and suggest
| that the author just talk to the appropriate person and implement it.
| I don't believe that I have ever outright refused a GLEP, or even added
| more than a day to the time for it to be posted.

For example, I first proposed the Planet idea long before GLEP 30 came
along. Basically the idea got shot down on IRC because infra were
worried about storing unmoderated material on Gentoo hardware. At that
point it was pretty much a "no", so I didn't bother following it up with
a GLEP.

Moving forward 1 year, we have Planet Gentoo. How the hell did that
happen then? Daniel persevered and produced a GLEP but which looked to
be doing nothing (like most other GLEPS) until gentooexperimental.org
came along setup a non official Planet. This made it obvious the idea
was popular and the implementation of the GLEP soon followed, as did
infra's concerns seem to disappear.

If we'd had a voting system, the popularity of the GLEP would have been
obvious from the start, and no doubt we would have had the Planet
implemented _much_ sooner.

In a nutshell, we need a system where ideas can be put forward so that
they have maximum exposure AND accessibility to vote. Posting them to
- -dev sure gets them a lot of exposure (though mostly only to devs), but
the majority of people likely to respond are those not in favor. Having
a simple voting system will allow people to put their voice forward with
maximum ease.

The thing I dislike about the current system is that most devs/users
will never hear about an idea until it makes onto glep.g.o. GLEPs need
to be put our there for everyone to see at an early stage so people can
vote on the IDEA, not the specifics or the quality of the GLEP itself.

| Most of these issues are moot, however, because in many cases GLEP
| authors prefer to post an informal GLEP on -dev first, gather comments,
| and then submit a formal GLEP after an initial revision.

Having all those in a central place would be far better, I'm sure there
are many GLEPs sitting in devspace that I or you have never seen.

| It's worth noting that a GLEP author need not be a dev, so those
| people, of course, would always require a surrogate to submit/revise
| GLEPs.  As for devs being able to upload/revise GLEPs, I'm not opposed.
| The reason that devs cannot update their own GLEPs right now is purely
| technical: the GLEP page is part of the www tree, and that tree has
| fairly strict permissions.  If opening up the GLEP directory isn't too
| much of a pita for infra, I certainly won't oppose it.  I would still
| prefer that GLEPs be run by one of the editors before being posted,
| since we may be able to help, but I wouldn't insist on it.  I would be
| very sad, though, if people took advantage of a more liberal policy to
| post poorly-thought-out junk.

If someone posts a poorly thought out GLEP then people will vote against
it. If the author wishes, he/she can post a revised version and the
voting process beings again. It's up to the author to make sure their
idea is of decent quality first time round, unless they want to spend
hours revising it.

Regards,
Ian Leitch

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCMgaWefZ4eWAXRGIRAlq5AJ0abRJqZJBHsujCHgTGAqT56OZ+awCgjAAg
1sYVgg8bfD9xsKoI81VMPm4=
=Wlfb
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Mike Frysinger]

On Friday 11 March 2005 03:59 pm, Ian Leitch wrote:
> If we'd had a voting system, the popularity of the GLEP would have been
> obvious from the start, and no doubt we would have had the Planet
> implemented _much_ sooner.

if only we could kick jeff into enabling voting on bugzilla ... then our 
communist users would have a voice !
-mike
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Martin Schlemmer]

[-- Attachment #1: Type: text/plain, Size: 1437 bytes --]

On Fri, 2005-03-11 at 20:59 +0000, Ian Leitch wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Grant Goodyear wrote:

> | It's worth noting that a GLEP author need not be a dev, so those
> | people, of course, would always require a surrogate to submit/revise
> | GLEPs.  As for devs being able to upload/revise GLEPs, I'm not opposed.
> | The reason that devs cannot update their own GLEPs right now is purely
> | technical: the GLEP page is part of the www tree, and that tree has
> | fairly strict permissions.  If opening up the GLEP directory isn't too
> | much of a pita for infra, I certainly won't oppose it.  I would still
> | prefer that GLEPs be run by one of the editors before being posted,
> | since we may be able to help, but I wouldn't insist on it.  I would be
> | very sad, though, if people took advantage of a more liberal policy to
> | post poorly-thought-out junk.
> 
> If someone posts a poorly thought out GLEP then people will vote against
> it. If the author wishes, he/she can post a revised version and the
> voting process beings again. It's up to the author to make sure their
> idea is of decent quality first time round, unless they want to spend
> hours revising it.
> 

Given, but there should be a cutoff point I think ... 3-5 tries is OK,
but 20 ....


-- 
Martin Schlemmer
Gentoo Linux Developer, Desktop/System Team Developer
Cape Town, South Africa


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 753 bytes --]

On Fri, 11 Mar 2005 23:31:13 +0200 Martin Schlemmer <azarah@gentoo.org>
wrote:
| Given, but there should be a cutoff point I think ... 3-5 tries is OK,
| but 20 ....

And if 3-5 tries is enough to get all but a few small details figured
out? GLEP 29 might take a few more revisions just to get the one small
remaining issue ironed out... There's no objection to most of it, just
that the perfect solution hasn't been found for one small part of it
yet.

Now, if it goes past the 3-5 limit with lots of people still hating the
entire thing, that's an entirely different issue.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Martin Schlemmer]

[-- Attachment #1: Type: text/plain, Size: 902 bytes --]

On Fri, 2005-03-11 at 21:33 +0000, Ciaran McCreesh wrote:
> On Fri, 11 Mar 2005 23:31:13 +0200 Martin Schlemmer <azarah@gentoo.org>
> wrote:
> | Given, but there should be a cutoff point I think ... 3-5 tries is OK,
> | but 20 ....
> 
> And if 3-5 tries is enough to get all but a few small details figured
> out? GLEP 29 might take a few more revisions just to get the one small
> remaining issue ironed out... There's no objection to most of it, just
> that the perfect solution hasn't been found for one small part of it
> yet.
> 
> Now, if it goes past the 3-5 limit with lots of people still hating the
> entire thing, that's an entirely different issue.
> 

I was more talking about the re-voting thing only - not posting to -dev
and then revising after some comments.


-- 
Martin Schlemmer
Gentoo Linux Developer, Desktop/System Team Developer
Cape Town, South Africa


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Luis F. Araujo]

Grant Goodyear wrote:

>It's worth noting that a GLEP author need not be a dev, so those
>people, of course, would always require a surrogate to submit/revise
>GLEPs.  As for devs being able to upload/revise GLEPs, I'm not opposed.
>The reason that devs cannot update their own GLEPs right now is purely
>technical: the GLEP page is part of the www tree, and that tree has
>fairly strict permissions.  If opening up the GLEP directory isn't too
>much of a pita for infra, I certainly won't oppose it.  I would still
>prefer that GLEPs be run by one of the editors before being posted,
>since we may be able to help, but I wouldn't insist on it.  I would be
>very sad, though, if people took advantage of a more liberal policy to
>post poorly-thought-out junk.
>  
>
I agree with the GLEPs being run by one or several editors, i think
it will keep the proposals very organized among them, something important
for the whole glep.g.o structure.
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] GLEP system worthwhile?

[Benjamin A. Collins]

[-- Attachment #1: Type: text/plain, Size: 852 bytes --]

I think the GLEP system is pretty nice actually.  Despite the extra
bureaucratic layers, it's a good system for vetting serious ideas.
Mailing lists are good for discussions, but once an idea has matured
some, it's nice to have the more formal document: (a) it's easier to
find and a good way to make important discussions available to those
who don't follow the mailing lists closely; (b) it provides a sort of
"official" reference point --- an RFC, if you will.

I don't think that the number of GLEPS that have been implemented,
accepted, rejected, or deferred are really that important (unless the
whole point is to increase the number of features that get implemented
in Gentoo, and as I perceive it, that's not the point of the GLEP
system).

bc
-- 
Benjamin A. Collins <bencollins@tamu.edu>
http://people.cs.tamu.edu/bcollins/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Marius Mauch]

[-- Attachment #1: Type: text/plain, Size: 865 bytes --]

On Fri, 11 Mar 2005 08:39:51 -0600
Grant Goodyear <g2boojum@gentoo.org> wrote:

> Thoughts / comments?  I promise not to bite anyone's head off this
> time!

Suggestion: make a short (one word) descriptive name mandatory. Numbers
get easily confused and most of the more recent GLEPs already come with
a nickname. For example I wouldn't know offhand which GLEP would deal
with ACCEPT_LICENSE stuff, and I'm also not sure about the real title
(maybe it doesn't contain ACCEPT_LICENSE at all), so if I want to make a
reference to it I actually have to look it up so people know what I'm
talking about. Minor issue, but really annoying over time.

Marius

-- 
Public Key at http://www.genone.de/info/gpg-key.pub

In the beginning, there was nothing. And God said, 'Let there be
Light.' And there was still nothing, but you could see a bit better.

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Grant Goodyear]

[-- Attachment #1: Type: text/plain, Size: 1039 bytes --]

Marius Mauch wrote: [Fri Mar 11 2005, 10:34:08AM CST]
> Suggestion: make a short (one word) descriptive name mandatory. Numbers
> get easily confused and most of the more recent GLEPs already come with
> a nickname. For example I wouldn't know offhand which GLEP would deal
> with ACCEPT_LICENSE stuff, and I'm also not sure about the real title
> (maybe it doesn't contain ACCEPT_LICENSE at all), so if I want to make a
> reference to it I actually have to look it up so people know what I'm
> talking about. Minor issue, but really annoying over time.

I agree with the sentiment, although I think a one-word limit may be a
bit draconian.  For example, Ciaran's latest GLEP is titled
"Per-Category metadata.xml files".  I think that title is perfectly
succinct, and I lack the creativity necessary to come up with an
equivalent one-word version.

-g2boojum-
-- 
Grant Goodyear	
Gentoo Developer
g2boojum@gentoo.org
http://www.gentoo.org/~g2boojum
GPG Fingerprint: D706 9802 1663 DEF5 81B0  9573 A6DC 7152 E0F6 5B76

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] GLEP system worthwhile?

[Thierry Carrez]

[-- Attachment #1: Type: text/plain, Size: 671 bytes --]

Grant Goodyear wrote:

> Thoughts / comments?  I promise not to bite anyone's head off this time!

It's a good thing to have enhancement proposals clearly explained and
written and the GLEP system does a great job for that. The way I see it,
the system generally fails after the GLEP is written :

- They are voted at Managers meeting level and people don't feel their
views are represented there

- Voted enhancements that need coordinating several Gentoo teams at the
same time (Metabugs) don't fly very well

These aren't GLEP-specific problems, these are metastructure problems.
They should both be solved by the current effort to reform our
metastructure.

-- 
Koon

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]