Catalyst is Gentoo's Release Engineering team's metatool used for creating installation stages, LiveCDs, and GRP

Sebastian Pipping | 4 Jul 2011 01:28

Picon

Re: Migrating man page to asciidoc?

On 06/27/2011 05:55 PM, Sebastian Pipping wrote:
> On 06/27/2011 05:58 AM, Jorge Manuel B. S. Vicetto wrote:
>> We use 2.0.6.916 on official releng releases.
>> I tried 9999 on my private tests because [..]
> 
> To me that leaves the question who uses 9999 for serious stuff at all.
> In case it's no one (which we should find out) we could trash that
> branch and fully concentrate on calalyst_2.
> 
> I guess that sounds a bit radical at first.  It doesn't have to be a
> quick decision.  Also, version control allows us to bring it back if needed.
> 
> Ideas on find out who is using 9999:
> 
>  - Removing 9999 ebuild from the tree and see who's complaining
> 
>  - Resetting branch master to nothing but a README announcing
>    the possible death of that thread and a request to join
>    this mailing list and speak up about it if there is need.
> 
>  - Asking on one/some/all of gentoo-dev, gentoo-user, gentoo forums,
>    planet gentoo.
> 
> After such action I imagine a time window of 2 to 4 weeks to give people
> a chance to react.
> 
> What do you think?
> 
> 
>> I can confirm that the build with the master branch fails as it doesn't

(Continue reading)

Matt Turner | 4 Jul 2011 02:10

Picon

Re: Migrating man page to asciidoc?

On Sun, Jul 3, 2011 at 11:28 PM, Sebastian Pipping <sping@...> wrote:
> On 06/27/2011 05:55 PM, Sebastian Pipping wrote:
>> On 06/27/2011 05:58 AM, Jorge Manuel B. S. Vicetto wrote:
>>> We use 2.0.6.916 on official releng releases.
>>> I tried 9999 on my private tests because [..]
>>
>> To me that leaves the question who uses 9999 for serious stuff at all.
>> In case it's no one (which we should find out) we could trash that
>> branch and fully concentrate on calalyst_2.
>>
>> I guess that sounds a bit radical at first.  It doesn't have to be a
>> quick decision.  Also, version control allows us to bring it back if needed.
>>
>> Ideas on find out who is using 9999:
>>
>>  - Removing 9999 ebuild from the tree and see who's complaining
>>
>>  - Resetting branch master to nothing but a README announcing
>>    the possible death of that thread and a request to join
>>    this mailing list and speak up about it if there is need.
>>
>>  - Asking on one/some/all of gentoo-dev, gentoo-user, gentoo forums,
>>    planet gentoo.
>>
>> After such action I imagine a time window of 2 to 4 weeks to give people
>> a chance to react.
>>
>> What do you think?
>>
>>

(Continue reading)

Sebastian Pipping | 4 Jul 2011 02:18

Picon

Re: Migrating man page to asciidoc?

On 07/04/2011 02:10 AM, Matt Turner wrote:
>>>> We use 2.0.6.916 on official releng releases.
>>>> I tried 9999 on my private tests because [..]
>>>
>>> To me that leaves the question who uses 9999 for serious stuff at all.
>>> In case it's no one (which we should find out) we could trash that
>>> branch and fully concentrate on calalyst_2.
>>>
>>> I guess that sounds a bit radical at first.  It doesn't have to be a
>>> quick decision.  Also, version control allows us to bring it back if needed.
>>>
>>> Ideas on find out who is using 9999:
>>>
>>>  - Removing 9999 ebuild from the tree and see who's complaining
>>>
>>>  - Resetting branch master to nothing but a README announcing
>>>    the possible death of that thread and a request to join
>>>    this mailing list and speak up about it if there is need.
>>>
>>>  - Asking on one/some/all of gentoo-dev, gentoo-user, gentoo forums,
>>>    planet gentoo.
>>>
>>> After such action I imagine a time window of 2 to 4 weeks to give people
>>> a chance to react.
>>>
>>> What do you think?
>>>
>>>
>>>> I can confirm that the build with the master branch fails as it doesn't
>>>> seem able to find the spec files or doesn't accept them - the official

(Continue reading)

Sebastian Pipping | 4 Jul 2011 02:22

Picon

[rfc] Deleting catalyst 3.x code (was Fwd: Re: Migrating man page to asciidoc?)

Second try as a new thread:

On 06/27/2011 05:58 AM, Jorge Manuel B. S. Vicetto wrote:
> We use 2.0.6.916 on official releng releases.
> I tried 9999 on my private tests because [..]

To me that leaves the question who uses 9999 for serious stuff at all.
In case it's no one (which we should find out) we could trash that
branch and fully concentrate on calalyst_2.

I guess that sounds a bit radical at first.  It doesn't have to be a
quick decision.  Also, version control allows us to bring it back if needed.

Ideas on find out who is using 9999:

 - Removing 9999 ebuild from the tree and see who's complaining

 - Resetting branch master to nothing but a README announcing
   the possible death of that thread and a request to join
   this mailing list and speak up about it if there is need.

 - Asking on one/some/all of gentoo-dev, gentoo-user, gentoo forums,
   planet gentoo.

After such action I imagine a time window of 2 to 4 weeks to give people
a chance to react.

What do you think?

> I can confirm that the build with the master branch fails as it doesn't

(Continue reading)

Matt Turner | 4 Jul 2011 02:32

Picon

Re: [rfc] Deleting catalyst 3.x code (was Fwd: Re: Migrating man page to asciidoc?)

On Mon, Jul 4, 2011 at 12:22 AM, Sebastian Pipping <sping@...> wrote:
> Ideas on find out who is using 9999:
>
>  - Removing 9999 ebuild from the tree and see who's complaining

I actually use the -9999 ebuild with EGIT_BRANCH="catalyst_2", so I'd
prefer not to do this.

I would probably just ask if anyone is using catalyst3 through a few
mediums. I suspect that no one is.

In order to make any sort of decisions, we should probably know what
the significant changes in catalyst-3 are to begin with.

Matt

Sebastian Pipping | 4 Jul 2011 02:45

Picon

Re: [rfc] Deleting catalyst 3.x code (was Fwd: Re: Migrating man page to asciidoc?)

On 07/04/2011 02:32 AM, Matt Turner wrote:
> On Mon, Jul 4, 2011 at 12:22 AM, Sebastian Pipping <sping@...> wrote:
>> Ideas on find out who is using 9999:
>>
>>  - Removing 9999 ebuild from the tree and see who's complaining
> 
> I actually use the -9999 ebuild with EGIT_BRANCH="catalyst_2", so I'd
> prefer not to do this.

To improve on that workaround please add a new live ebuild 2.9999.

Sebastian

Zachary Bedell | 6 Jul 2011 19:28

Missing /usr/src in installcd-stage3-minimal

Greetings all,

I'm just beginning to play with Catalyst to create custom installcd images.  I've started with the Gentoo
releng spec files from svn://anonsvn.gentoo.org/releng/trunk/releases/weekly/specs/amd64.  I
managed to get through stage1-3 and installcd-stage1 without any trouble.  

Running installcd-stage2-minimal dies when trying to run Genkernel with "ln: failed to create symbolic
link `/usr/src/linux': No such file or directory".  I hacked the kmerge.sh script to drop to a shell right
after that happened and found that there was no /usr/src directory in the chroot.  Doing a `mkdir -p
/usr/src` right before the ln call, and everything seems to work out after that.

I'm assuming something probably went wrong in an earlier stage, but I don't know enough about Catalyst yet
to know quite where to look.

I can post spec files if they'd help, but I'm using stock from SVN with just some version number's changed
around.  Does anyone have a guess where I might look, or is there any chance this might be a bug in the
kmerge.sh script?  I'm using catalyst-2.0.6.916 at this point, but I may just give the -9999 build a try to
see if that fixed the issue.

Any pointers would be appreciated.

Best regards,
Zac Bedell

Peter Stuge | 6 Jul 2011 20:10

Picon

Re: Missing /usr/src in installcd-stage3-minimal

Zachary Bedell wrote:
> found that there was no /usr/src directory in the chroot

The latest stage3 also does not add udev to the sysinit runlevel.
Lost about an hour to that.

> I'm assuming something probably went wrong in an earlier stage,

Not neccessarily, think about who should create /usr/src. IMO it's
not at all really neccessary, it's just by (deprecated) convention
that there are kernel sources there.

> but I don't know enough about Catalyst yet to know quite where to
> look.

It has little to do with catalyst itself. In any case, if you make
sure that your livecd-stage1 includes kernel sources, then I think
you'll sidestep around the problem.

//Peter

Sebastian Pipping | 6 Jul 2011 20:14

Picon

Re: Missing /usr/src in installcd-stage3-minimal

On 07/06/2011 07:28 PM, Zachary Bedell wrote:
> Greetings all,
> 
> I'm just beginning to play with Catalyst to create custom installcd
> images.  I've started with the Gentoo releng spec files from
> svn://anonsvn.gentoo.org/releng/trunk/releases/weekly/specs/amd64.  I
> managed to get through stage1-3 and installcd-stage1 without any
> trouble.
> 
> Running installcd-stage2-minimal dies when trying to run Genkernel
> with "ln: failed to create symbolic link `/usr/src/linux': No such
> file or directory".  I hacked the kmerge.sh script to drop to a shell
> right after that happened and found that there was no /usr/src
> directory in the chroot.  Doing a `mkdir -p /usr/src` right before
> the ln call, and everything seems to work out after that.
> 
> I'm assuming something probably went wrong in an earlier stage, but I
> don't know enough about Catalyst yet to know quite where to look.
> 
> I can post spec files if they'd help, but I'm using stock from SVN
> with just some version number's changed around.  Does anyone have a
> guess where I might look, or is there any chance this might be a bug
> in the kmerge.sh script?  I'm using catalyst-2.0.6.916 at this point,
> but I may just give the -9999 build a try to see if that fixed the
> issue.

Afaik the code in 9999 differs by more than just bugfixes.  Let me quote
from the -9999 ebuild:

  "The git master branch (what you get with this -9999 ebuild) for

(Continue reading)

Sebastian Pipping | 6 Jul 2011 21:14

Picon

Re: Missing /usr/src in installcd-stage3-minimal

On 07/06/2011 08:10 PM, Peter Stuge wrote:
> Zachary Bedell wrote:
>> found that there was no /usr/src directory in the chroot
> 
> The latest stage3 also does not add udev to the sysinit runlevel.
> Lost about an hour to that.

Now that you mention it: I ran into the same problem.  It was ~20
something hours wasted over here I guess before I found that Xorg didn't
find my graphics card because of udev not being started.

Best,

Sebastian

Group

gmane.linux.gentoo.catalyst.

Advertisement

Project Web Page

Catalyst is Gentoo's Release Engineering team's metatool used for creating installation stages, LiveCDs, and GRP

Search Archive

Page

1 | 2

Articles in period: 15

July 2011

Language

Change language

Options

Current view: Threads only / Showing only 20 lines / Not hiding cited text.
Change to All messages, whole messages, or hide cited text.

Post a message
NNTP Newsgroup
Classic Gmane web interface
XML

XML

RSS Feed
List Information

About Gmane

Recent entries

Recent changes to my rewite branch (1 Jun 2013)
PKGCACHE_PATH and os.makedirs() usage (29 May 2013)
Python 3 support (20 May 2013)
Support for custom memory settings in AUFS (26 Apr 2013)
Catalyst Python reimplementation (21 Apr 2013)
[PATCH 0/2] Blacklisting binary packages (17 Apr 2013)
make catalyst call emerge --deep (10 Apr 2013)
make catalyst call emerge --deep (10 Apr 2013)
[PATCH] Add support for m68k (4 Apr 2013)
[PATCH] Make the use of preserved_libs optional by setting the options (31 Mar 2013)
[PATCH 1/2] Fix update_seed use by not using nor building binary packa (27 Mar 2013)
sem_open inconsistency in installed Python versions (11 Mar 2013)
[PATCH] livecdfs-update.sh: Escape ampersands in STARTX sed expression (3 Mar 2013)
[PATCH] livecdfs-update.sh: Fix '/etc/sshd' check for sshd_config twea (3 Mar 2013)
[PATCH] livecdfs-update.sh: Set XSESSION in /etc/env.d/90xsession (2 Mar 2013)
Stale AUTHORS? (1 Mar 2013)
Bug #355995, depscan.sh and update-modules (1 Mar 2013)
[PATCH 0/4] Some general cleanup, remove some hardcoding of paths. (28 Feb 2013)
[PATCH] Fix broken seed stage update... (27 Feb 2013)
[patch] mixed spaces/tabs and indent cleanup. (27 Feb 2013)
[patch] Fix a bug that portage didn't get rebuilt with the build use f (27 Feb 2013)

Archives

4	June 2013
22	May 2013
58	April 2013
38	March 2013
78	February 2013
21	January 2013
14	December 2012
3	November 2012
2	October 2012
1	September 2012
22	August 2012
17	July 2012
9	June 2012
2	March 2012
1	February 2012
15	December 2011
2	November 2011
5	October 2011
15	July 2011
88	June 2011
2	May 2011
11	March 2011
6	November 2010
5	October 2010
7	July 2010
12	June 2010
2	May 2010
2	April 2010
2	March 2010
5	December 2009
14	November 2009
27	October 2009
7	September 2009
33	August 2009
2	July 2009
13	June 2009
5	May 2009
2	April 2009
10	March 2009
8	February 2009
6	December 2008
8	November 2008
2	October 2008
20	September 2008
20	July 2008
3	June 2008
9	May 2008
8	April 2008
28	March 2008
66	February 2008
23	January 2008
3	December 2007
7	November 2007
31	October 2007
29	September 2007
47	August 2007
37	July 2007
67	June 2007
6	May 2007
123	April 2007
53	March 2007
22	February 2007
47	January 2007
19	December 2006
10	November 2006
9	October 2006
18	September 2006
161	August 2006
73	July 2006
2	June 2006
50	May 2006
58	April 2006
57	March 2006
23	February 2006
8	January 2006
50	December 2005
32	November 2005
41	October 2005
123	September 2005
10	August 2005
159	July 2005
63	June 2005
68	May 2005

Design

Zawodny | Right Menu | Left Menu

Your Own Design

Paste the URL of your CSS below. Download CSS template