[gentoo-dev] maintainer-wanted bugcount

2007-11-26 Thread Markus Ullmann
Hi fellows,

when taking a look at the open bug count for bugs assigned to
maintainer-wanted (2450 at the time of writing), it seems pretty obvious
that we really can't handle all of them, at least not without growing at
least two dozen devs to maintain it properly.

As I highly doubt this will happen within a week, we have to make a
decision how to proceed with this stuff. So what options do we have?
These come to mind:

a) WONTFIX them within 4 or 8 weeks without picking them up
b) reassign them to herds (some herds are on CC) and have them
   respond withing 4-8 weeks and give a yey or boo.
c) let interested users move it to sunrise (some of them are there)
   so that the ebuilds are at least at our QA level we maintain for
   gentoo-x86 and are there to be picked up by devs if they're
   interested

If you have more options or comments, I'd like to hear about them, as we
definitely have to do something there.
F'up is set to gentoo-project as this is more a political thing.

Greetz
-Jokey




signature.asc
Description: OpenPGP digital signature


Re: [gentoo-dev] maintainer-wanted bugcount

2007-11-26 Thread Andrey Falko
On Nov 26, 2007 4:46 AM, Markus Ullmann <[EMAIL PROTECTED]> wrote:

> Hi fellows,
>
> when taking a look at the open bug count for bugs assigned to
> maintainer-wanted (2450 at the time of writing), it seems pretty obvious
> that we really can't handle all of them, at least not without growing at
> least two dozen devs to maintain it properly.
>
> As I highly doubt this will happen within a week, we have to make a
> decision how to proceed with this stuff. So what options do we have?
> These come to mind:
>
> a) WONTFIX them within 4 or 8 weeks without picking them up
> b) reassign them to herds (some herds are on CC) and have them
>   respond withing 4-8 weeks and give a yey or boo.
> c) let interested users move it to sunrise (some of them are there)
>   so that the ebuilds are at least at our QA level we maintain for
>   gentoo-x86 and are there to be picked up by devs if they're
>   interested
>
> If you have more options or comments, I'd like to hear about them, as we
> definitely have to do something there.
> F'up is set to gentoo-project as this is more a political thing.
>
> Greetz
> -Jokey
>
>
> Hi everyone,

I did the "Collective Maintenance" project this summer:
http://code.google.com/soc/2007/gentoo/appinfo.html?csaid=2881CA66D3587EA2

My vision was to move things in to a direction where the "maintainer-wanted"
packages would be collectively maintained without the need for a specific
herd or maintainer. To get to this vision, I set out to create a web
interface that would allow work to be coordinated in an organized,
efficient, and by-demand manner. Here is where I am currently at with that
interface: http://afalko.homelinux.net/tskdemo/

I think the current state of the interface is not going to allow for enough
organization. My fear is that it will turn into an unnavigable
dumping-ground when a lot of tasks/"maintainer-wanted" bugs are added to the
list. I am one year from earning at B.S. degree in Industrial Engineering
and my requirement this school year is to complete an IE design project. I
was able to get a group of four together to work on redesigning the
interface. My group members have been generating some excellent ideas so
far. I am optimistic that the end product will allow my vision to be
achieved. I will push to get the interface ready before the end of May 2008.

So in sum, I guess you can add a fourth option to the list: wait until the
end of May 2008 for the completion/deployment of an interface for
collectively maintaining packages. Or even a fifth option: use the current
interface right now and upgrade in May 2008 to a re-designed revision.

Best regards,
Andrey Falko


Re: [gentoo-dev] maintainer-wanted bugcount

2007-11-26 Thread Robin H. Johnson
On Mon, Nov 26, 2007 at 10:46:12AM +0100, Markus Ullmann wrote:
> As I highly doubt this will happen within a week, we have to make a
> decision how to proceed with this stuff. So what options do we have?
> These come to mind:
d) In addition to c), keep them open, flagged with sunrise in the status
   board, so that when a developer does want some package not in the
   tree, they can search first.
e) Encourage existing developers to review and commit this stuff more
   often.

I really would not want to see them closed unless there is a good reason
for them to not be committed to the tree.

-- 
Robin Hugh Johnson
Gentoo Linux Developer & Infra Guy
E-Mail : [EMAIL PROTECTED]
GnuPG FP   : 11AC BA4F 4778 E3F6 E4ED  F38E B27B 944E 3488 4E85


pgpCytXgGNDVk.pgp
Description: PGP signature


Re: [gentoo-dev] New USE flags documentation

2007-11-26 Thread Alec Warner
On 11/24/07, Jose Luis Rivero <[EMAIL PROTECTED]> wrote:
> Hi all:
>
> I've read on the planet the recently included support to document USE
> flags in metadata. Seems like it was an idea from flameeyes and cardoe,
> discussed on the planet [1][2] and performed by -infra (bug #199788).
>
> While planet is a good medium to share ideas and get contributors, seems to
> me like we need a more official way to discuss this kind of 'global'
> ideas before make them real. Or at least drop a note on -dev-announce
> explaining the new feature and telling devs and users this is now
> officially supported.
>
> I'm not asking for an extra overhead of 'bureaucracy' (write specs,
> mailling @dev, send to the council, etc.) but a bit more of communication
> would be appreciated:

Maybe you aren't but I am.  It takes all of 20 minutes to write a
short GLEP.  It takes 5 minutes to write an e-mail telling everyone
that you are doing something new before you do it.  Thanks for
considering everyone in your changes.

>
> Is the feature ready to be used? Is there any kind of documentation
> (aside of DTD)? It will replace use.desc?
>
> Thanks.
>
> [1] 
> http://farragut.flameeyes.is-a-geek.org/articles/2007/11/24/proposing-more-use-flag-documentation
> [2] http://blog.cardoe.com/archives/2007/11/23/metadataxml-updates-examples/
>
> --
> Jose Luis Rivero <[EMAIL PROTECTED]>
> Gentoo/Doc Gentoo/Alpha
>
> --
> [EMAIL PROTECTED] mailing list
>
>
-- 
[EMAIL PROTECTED] mailing list



[gentoo-dev] Re: maintainer-wanted bugcount

2007-11-26 Thread Ryan Hill
Robin H. Johnson wrote:

> d) In addition to c), keep them open, flagged with sunrise in the status
>board, so that when a developer does want some package not in the
>tree, they can search first.
> e) Encourage existing developers to review and commit this stuff more
>often.
> 
> I really would not want to see them closed unless there is a good reason
> for them to not be committed to the tree.

++

-- 
looks like christmas at fifty-five degrees
this latitude weakens my knees
EFFD 380E 047A 4B51 D2BD  C64F 8AA8 8346 F9A4 0662 (0xF9A40662)



signature.asc
Description: OpenPGP digital signature


Re: [gentoo-dev] New USE flags documentation

2007-11-26 Thread José Luis Rivero (yoswink)

Jose Luis Rivero wrote:


I'm not asking for an extra overhead of 'bureaucracy' (write specs,
mailling @dev, send to the council, etc.) but a bit more of communication 
would be appreciated:


Seems like everyone who contact me/us about this thread is agree about 
the needed of write a GLEP before doing this kind of global changes. So 
I would like to add a couple of considerations:


1. Dear gentoo devs, in the future, please, write a GLEP for global 
changes affecting all of us and post them to -dev. This way you can get 
some feedback, improve the original idea and inform everybody about the 
new feature (all-in-one).


2. With respect to the new metadata USE flag implementation, I think 
that reverting the changes could be quite radical so my vote goes for: 
leave the thingy as it (stop adding use flags to metadata for now), 
write a quick GLEP and go through the process. When it's done, migrate 
the current data (if needed) and enjoy.


Thanks.

--
Jose Luis Rivero <[EMAIL PROTECTED]>
Gentoo/Doc Gentoo/Alpha
--
[EMAIL PROTECTED] mailing list



[gentoo-dev] Re: maintainer-wanted bugcount

2007-11-26 Thread Markus Ullmann
Robin H. Johnson schrieb:
> On Mon, Nov 26, 2007 at 10:46:12AM +0100, Markus Ullmann wrote:
> d) In addition to c), keep them open, flagged with sunrise in the status
>board, so that when a developer does want some package not in the
>tree, they can search first.

That's done already

> e) Encourage existing developers to review and commit this stuff more
>often.
That would be the best option

> I really would not want to see them closed unless there is a good reason
> for them to not be committed to the tree.

See the reason's I've given here:
http://article.gmane.org/gmane.linux.gentoo.project/179

Greetz
-Jokey



signature.asc
Description: OpenPGP digital signature


[gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread Hanno Böck
Hi,

The mailman ebuild was a pain in the past, installing to non-fhs-locations 
(/usr/local), doing lot's of strange stuff, not able to use etc-update...

mailman-2.1.9-r2 tries to fix lot's of those issues, it's much more 
configurable through some variables. It's currently masked, but yesterday I 
committed a bunch of changes and now I'm pretty satisfied with it.

So I'd like to unmask it soon. Please, if you're using mailman test it, tell 
me if it suits your needs or just give me feedback like "worksforme", I 
actually don't have a clue how many people really use this ebuild.

-- 
Hanno Böck  Blog:   http://www.hboeck.de/
GPG: 3DBD3B20   Jabber: [EMAIL PROTECTED]


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] New USE flags documentation

2007-11-26 Thread Alec Warner
On 11/26/07, Diego 'Flameeyes' Pettenò <[EMAIL PROTECTED]> wrote:
> CCing council so that the other members can express their feeling about this;
> basically the only people I actually care about getting the feeling about
> this at all.
>
> On Monday 26 November 2007, you wrote:
> > Seems like everyone who contact me/us about this thread is agree about
> > the needed of write a GLEP before doing this kind of global changes.

Er, I think your statement here is wrong, as Diego pointed out ;)

>
> Everyone who contacted who? Why would people contact YOU about this thread?
> They should, if anybody, contact me and Doug, if they think we did it the
> wrong way, or the council, if they thought the proposal was to be stopped (in
> which case I would have notice, being in the council myself). Even better
> they should have opened a bug for the council to stop it.
>
> So. let's face it, the only people who agree on the need for a GLEP is Alec
> and Ciaran. The rest of the thread (which by the way I had to dig up on Gmane
> because I didn't really give a damn) is composed by Thilo who think is a
> great idea, and by Jer who answered to Thilo saying, afaics, that he doesn't
> see the reason to make it *mandatory*. Then there is Doug.

I never said I disliked your idea.  I disliked the way it was rolled
out as it sets bad precedent for future ideas, which can now use this
as an example to do random crap to the tree and not get feedback about
it until it is too late.

>
> Oh wait, everybody who contacted _me_ thinks it's a great idea, or nearly
> everybody. The difference is that people who contacted me did so in the blog,
> so you can read the comments at [1]. Yes I know they don't really count much,
> but counts more of your "everybody who contacted me/us" to me, considering  I
> see only Antarus agreeing with you.
>
> On the proper matter, whether a GLEP is needed or not, well, I already said
> before I think the GLEP process is totally broken, and I don't think that
> waiting for months to get the GLEP approved would help users at all. For what
> it's worth, there is already a GLEP on metadata extension, GLEP 5.. yes FIVE.
> Status: deferred.
>
> I think a markup change is not a problem of GLEPs, I don't think a lot of
> stuff that gone into GLEP process should have, and should just have been
> realised.
>

So to address the only real problem I have with this aside from
setting poor precedent is updating the tools that parse metadata.  If
this, as Doug alluded to in an earlier comment, begins to replace
use.local.desc then I'd like to see the tools fixed to support it.
But maybe this is just another one of those pesky process problems
where someone releases a new change over a holiday and I will wake up
tomorrow with a bug filed against gentoolkit that requests parsing
this new metadata ;)

> And as for Alec's "20 minutes" comment, I would like to remind him that we
> have a lot of people getting obnoxious when you make even a spelling mistake,
> so for a non-native English speaker like I am, the 20 minutes figure is
> totally wrong. And this is also my reason not to write GLEPs ever in my life,
> I don't want to spend two weeks just to get the spelling right. That's a
> waste of my time, and as I'm not devoting my whole life to Gentoo, it ends up
> hurting users again.

Then e-mail the glep to one of the GLEP editors (hey thats me!) and
they will fix all the grammar problems and you can focus on the
content of the thing.

>
> At any rate, if you have any comment regarding the way some dev act, I'd
> suggest you, mostly for good life of both you and the dev involved, to ask
> him BEFORE crapping on him in public. The announcement thing you referred to,
> as Doug explained, was just a time problem, and as we're all volunteer, I
> don't think Doug was forced to find the time to fix the stuff.
>
> So, as I don't really want to waste even more time on this thing that I think
> it's totally a non-issue and just a time wasting thing, I would just ask the
> opinion of the other members of the council. If they think we can proceed, I
> won't stop to add documentation that users can use; if they want to discuss
> it next meeting, I'll wait for it before doing anything; if they think it has
> to be removed and discusse, I'll comment out my metadata (I won't REMOVE
> them, users needs to have proper documentation of USE flags, so as I don't
> find it good for them to remove it, I'll remove it from the semantic of
> metadata until a new syntax could be made official - note that we NEED such
> documentation; if going through GLEP process means making this another
> deferred GLEP and thus giving up on documenting the USE flags for another
> year or two, then I'll be ready to fight the decision until devrel removes me
> from my position).

I don't think a rollback is necessary.  If necessary I will write up
the GLEP and get it approved as I'd like to have some record of the
change besides the cvs logs for the metadata.dtd (things

Re: [gentoo-dev] New USE flags documentation

2007-11-26 Thread Ciaran McCreesh
On 11/26/07, Diego 'Flameeyes' Pettenò <[EMAIL PROTECTED]> wrote:
> So. let's face it, the only people who agree on the need for a GLEP
> is Alec and Ciaran. The rest of the thread (which by the way I had
> to dig up on Gmane because I didn't really give a damn) is composed
> by Thilo who think is a great idea, and by Jer who answered to
> Thilo saying, afaics, that he doesn't see the reason to make it
> *mandatory*. Then there is Doug.

Er, I don't think it's a bad idea. Heck, I suggested it several years
back. What I do think is that there are several less than ideal aspects
of the way it's proposed that could easily be fixed by someone spending
a few hours going through the GLEP process. These aren't difficult
changes, but they will substantially improve the end result.

And no, the GLEP process isn't broken. Well thought out proposals can
be approved very quickly, and the GLEP process helps iron out any
problems in the proposals.

-- 
Ciaran McCreesh


signature.asc
Description: PGP signature


Re: [gentoo-dev] Re: maintainer-wanted bugcount

2007-11-26 Thread George Shapovalov
Monday, 26. November 2007, Markus Ullmann Ви написали:
> Robin H. Johnson schrieb:
> > On Mon, Nov 26, 2007 at 10:46:12AM +0100, Markus Ullmann wrote:
> > d) In addition to c), keep them open, flagged with sunrise in the status
> >board, so that when a developer does want some package not in the
> >tree, they can search first.
>
> That's done already
So what's the problem? 

The only onse where there is a valid issue, I think, are the ones where 
upstream has gone missing or no longer supports the package. (Even "better 
alternatives" thing is questionable IMHO. There are always people who have a 
different idea of better :)). These can be dealt with by having a dedicated 
force, perhaps comprised of people most annoyed by these bugs ;), who would 
simply scan the bugzilla for maintainer-wanted bugs, check the upstream and, 
if it is dead, close the bug with INVALID and a note of a dead upstream..

>
> > e) Encourage existing developers to review and commit this stuff more
> >often.
>
> That would be the best option
Except the issue with all these packages is not committing but maintaining 
them. Sure, we could go on a committing spree and never touch any of these 
packages again. We even had such a situation at one point (rather long time 
ago) and it is specifically discouraged now.
The "really best" option would be recruiting new devs so that we have enogh 
people to maintain everything and add more, as was already suggested. This in 
fact does happen, it is just that the rate of new joins cannot be infinite 
plus we seem to experience our growing pains periodically..

George
--
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] New USE flags documentation

2007-11-26 Thread Fernando J. Pereda
[ I'm answering diego's mail here because I didn't get a copy of it
through gentoo-dev, I guess the list missed me or my spam filter killed
it ]

On Mon, Nov 26, 2007 at 07:10:40AM -0800, Alec Warner wrote:
> On 11/26/07, Diego 'Flameeyes' Pettenò <[EMAIL PROTECTED]> wrote:
> > CCing council so that the other members can express their feeling about 
> > this;
> > basically the only people I actually care about getting the feeling about
> > this at all.

So you do not care about the rest of the dev community... great.

> > Oh wait, everybody who contacted _me_ thinks it's a great idea, or nearly
> > everybody. The difference is that people who contacted me did so in the 
> > blog,
> > so you can read the comments at [1]. Yes I know they don't really count 
> > much,
> > but counts more of your "everybody who contacted me/us" to me, considering  
> > I
> > see only Antarus agreeing with you.

Your proposal is no better just because people agree with you. Moreover,
it doesn't even address some of the concerns that were raised when it
was first brought up. *Brillant*.

> > On the proper matter, whether a GLEP is needed or not, well, I already said
> > before I think the GLEP process is totally broken, and I don't think that
> > waiting for months to get the GLEP approved would help users at all. For 
> > what
> > it's worth, there is already a GLEP on metadata extension, GLEP 5.. yes 
> > FIVE.
> > Status: deferred.
> >
> > I think a markup change is not a problem of GLEPs, I don't think a lot of
> > stuff that gone into GLEP process should have, and should just have been
> > realised.

This means that anobody who thinks that the 'put your favourite rule
here' rule is broken is hereby allowed to break it. Right?

> > And as for Alec's "20 minutes" comment, I would like to remind him that we
> > have a lot of people getting obnoxious when you make even a spelling 
> > mistake,
> > so for a non-native English speaker like I am, the 20 minutes figure is
> > totally wrong. And this is also my reason not to write GLEPs ever in my 
> > life,
> > I don't want to spend two weeks just to get the spelling right. That's a
> > waste of my time, and as I'm not devoting my whole life to Gentoo, it ends 
> > up
> > hurting users again.

And not caring about the whole development and user community by
bypassing its rules hurts all of us even more.

> > At any rate, if you have any comment regarding the way some dev act, I'd
> > suggest you, mostly for good life of both you and the dev involved, to ask
> > him BEFORE crapping on him in public. The announcement thing you referred 
> > to,
> > as Doug explained, was just a time problem, and as we're all volunteer, I
> > don't think Doug was forced to find the time to fix the stuff.

You are the only one that crapped yourself in public with this
completely uncalled for response.

You acted without asking the community so if people raise concerns about
how you acted, just face it. Don't play the "contact me before saying I
fucked up in public" thing.

- ferdy

-- 
Fernando J. Pereda Garcimartín
20BB BDC3 761A 4781 E6ED  ED0B 0A48 5B0C 60BD 28D4


pgpMZdr7aZPVF.pgp
Description: PGP signature


[gentoo-dev] Re: maintainer-wanted bugcount

2007-11-26 Thread Markus Ullmann
K, to sum it up then, everything stays like it is atm.

Thanks for your comments :)

Greetz
-Jokey



signature.asc
Description: OpenPGP digital signature


Re: [gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread Wolfram Schlich
* Hanno Böck <[EMAIL PROTECTED]> [2007-11-26 15:39]:
> Hi,
> 
> The mailman ebuild was a pain in the past, installing to non-fhs-locations 
> (/usr/local), doing lot's of strange stuff, not able to use etc-update...
> 
> mailman-2.1.9-r2 tries to fix lot's of those issues, it's much more 
> configurable through some variables. It's currently masked, but yesterday I 
> committed a bunch of changes and now I'm pretty satisfied with it.

Nice!

> So I'd like to unmask it soon. Please, if you're using mailman test it, tell 
> me if it suits your needs or just give me feedback like "worksforme", I 
> actually don't have a clue how many people really use this ebuild.

Any special hints/advice?
-- 
Regards,
Wolfram Schlich <[EMAIL PROTECTED]>
Gentoo Linux * http://dev.gentoo.org/~wschlich/
--
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread Wolfram Schlich
* Hanno Böck <[EMAIL PROTECTED]> [2007-11-26 15:39]:
> [...]
> So I'd like to unmask it soon. Please, if you're using mailman test it, tell 
> me if it suits your needs or just give me feedback like "worksforme", I 
> actually don't have a clue how many people really use this ebuild.

pkg_postinst() says...
--8<--
 * Please read /usr/share/doc/mailman-2.1.9-r2/README.gentoo.gz for additional
 * Setup information, mailman will NOT run unless you follow
 * those instructions!
--8<--
...but that README actually has .bz2 instead of .gz on my system :)
-- 
Regards,
Wolfram Schlich <[EMAIL PROTECTED]>
Gentoo Linux * http://dev.gentoo.org/~wschlich/
--
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread Wolfram Schlich
* Hanno Böck <[EMAIL PROTECTED]> [2007-11-26 15:39]:
> [...]
> So I'd like to unmask it soon. Please, if you're using mailman test it, tell 
> me if it suits your needs or just give me feedback like "worksforme", I 
> actually don't have a clue how many people really use this ebuild.

I get this using hardened-sources with activated grsecurity
trusted path execution feature:

2007-11-27 02:15:47 +01:00; alpha; kern.alert; kernel: grsec: From 127.0.0.6: \
denied untrusted exec of /usr/lib/mailman/bin/mmsitepass by \
/bin/bash[bash:14178] uid/euid:280/280 gid/egid:280/280, \
parent /bin/bash[bash:14173] uid/euid:280/280 gid/egid:280/280

That's because /usr/lib/mailman/bin/ is group-writable.
Is that necessary at all?!
-- 
Regards,
Wolfram Schlich <[EMAIL PROTECTED]>
Gentoo Linux * http://dev.gentoo.org/~wschlich/
--
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread Wolfram Schlich
* Wolfram Schlich <[EMAIL PROTECTED]> [2007-11-27 02:24]:
> * Hanno Böck <[EMAIL PROTECTED]> [2007-11-26 15:39]:
> > [...]
> > So I'd like to unmask it soon. Please, if you're using mailman test it, 
> > tell 
> > me if it suits your needs or just give me feedback like "worksforme", I 
> > actually don't have a clue how many people really use this ebuild.
> 
> I get this using hardened-sources with activated grsecurity
> trusted path execution feature:
> 
> 2007-11-27 02:15:47 +01:00; alpha; kern.alert; kernel: grsec: From 127.0.0.6: 
> \
>   denied untrusted exec of /usr/lib/mailman/bin/mmsitepass by \
>   /bin/bash[bash:14178] uid/euid:280/280 gid/egid:280/280, \
>   parent /bin/bash[bash:14173] uid/euid:280/280 gid/egid:280/280
> 
> That's because /usr/lib/mailman/bin/ is group-writable.

Ok, that's not true :]

Using this configuration...
--8<--
CONFIG_GRKERNSEC_TPE=y
# CONFIG_GRKERNSEC_TPE_ALL is not set
CONFIG_GRKERNSEC_TPE_INVERT=y
CONFIG_GRKERNSEC_TPE_GID=1005
--8<--
...I have to add 'mailman' to group 1005.
-- 
Regards,
Wolfram Schlich <[EMAIL PROTECTED]>
Gentoo Linux * http://dev.gentoo.org/~wschlich/
--
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread Wolfram Schlich
* Wolfram Schlich <[EMAIL PROTECTED]> [2007-11-27 02:31]:
> * Wolfram Schlich <[EMAIL PROTECTED]> [2007-11-27 02:24]:
> > * Hanno Böck <[EMAIL PROTECTED]> [2007-11-26 15:39]:
> > > [...]
> > > So I'd like to unmask it soon. Please, if you're using mailman test it, 
> > > tell 
> > > me if it suits your needs or just give me feedback like "worksforme", I 
> > > actually don't have a clue how many people really use this ebuild.
> > 
> > I get this using hardened-sources with activated grsecurity
> > trusted path execution feature:
> > 
> > 2007-11-27 02:15:47 +01:00; alpha; kern.alert; kernel: grsec: From 
> > 127.0.0.6: \
> > denied untrusted exec of /usr/lib/mailman/bin/mmsitepass by \
> > /bin/bash[bash:14178] uid/euid:280/280 gid/egid:280/280, \
> > parent /bin/bash[bash:14173] uid/euid:280/280 gid/egid:280/280
> > 
> > That's because /usr/lib/mailman/bin/ is group-writable.
> 
> Ok, that's not true :]
> 
> Using this configuration...
> --8<--
> CONFIG_GRKERNSEC_TPE=y
> # CONFIG_GRKERNSEC_TPE_ALL is not set
> CONFIG_GRKERNSEC_TPE_INVERT=y
> CONFIG_GRKERNSEC_TPE_GID=1005
> --8<--
> ...I have to add 'mailman' to group 1005.

Ok, it get's worse: for the mailman webinterface, I'd have to add
'apache' to group 1005 as well, opening up even bigger holes.
No way! So, emerge -C mailman, that is :(
Too bad.
-- 
Regards,
Wolfram Schlich <[EMAIL PROTECTED]>
Gentoo Linux * http://dev.gentoo.org/~wschlich/
--
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] net-mail/mailman-2.1.9-r2: Request for testing

2007-11-26 Thread René 'Necoro' Neumann
Wolfram Schlich schrieb:
> * Hanno Böck <[EMAIL PROTECTED]> [2007-11-26 15:39]:
>> [...]
>> So I'd like to unmask it soon. Please, if you're using mailman test it, tell 
>> me if it suits your needs or just give me feedback like "worksforme", I 
>> actually don't have a clue how many people really use this ebuild.
> 
> pkg_postinst() says...
> --8<--
>  * Please read /usr/share/doc/mailman-2.1.9-r2/README.gentoo.gz for additional
>  * Setup information, mailman will NOT run unless you follow
>  * those instructions!
> --8<--
> ...but that README actually has .bz2 instead of .gz on my system :)

Depends on what PORTAGE_COMPRESS is set to ;) (Don't know WHERE this is
actually being set - but different systems seem to have different values
here).

- Necoro
-- 
[EMAIL PROTECTED] mailing list



Re: [gentoo-dev] Good-bye

2007-11-26 Thread Peter Weller
On Sun, 2007-11-25 at 16:31 +0100, Sven Vermeulen wrote:
> It seems like Gentoo is
> undergoing a generation switch - many of the elderly are taking on
> less responsibilities (or even retire) to give place to the energetic
> new developers ;-)
Yay for kids! :D

Bai bai seemant :( I don't think I ever got to know you as well as a
probably should've... Too busy being a kid, I suppose :P

welp


signature.asc
Description: This is a digitally signed message part