On Wed, Jun 25, 2008 at 9:51 PM, Donnie Berkholz <dberkholz@g.o> wrote:
> On 18:22 Wed 25 Jun     , Nirbheek Chauhan wrote:
>> Here it is:
>> http://soc.gentooexperimental.org/news/show/5
>>
>> Cya next week! ;)
>
> Could you paste the content into a reply, just in case soc.ge.org ever
> disappears?

Ah, yes, I thought of that, but I didn't think anyone else would ;p

Here it is:

-----------------------
Last week's report[1] was a mighty long one because it was the first
one, so this one won't be <*sighs of relief*>.

The projections (last week) for this week's work were:

    * Chroot preparation
    * jobuild parsing via Jobuild() object and jobuild.sh

I've only been able to do Chroot preparation however, since jobuild
parsing has turned out to be more tricky than we had anticipated (and
we really don't want ugly solutions ;). Also, Patrick is out of town,
and I had an exam yesterday, so that hindered progress too. However,
I've not been idle in the past week, and the chroot preparation and
manipulation parts are completely done (more than what I had projected
for this week). See

slave/autotua/chroot/__init__.py

The second thing that I had to do was create some space so I could
test the slave out properly. This involved resizing my windows
partition to increase the size of my "experiment mount". As luck would
have it, my windows partition had bad blocks, so I could not move the
free space to the existing (albeit small) "experiment mount". The only
solution I could see was to delete my windows partition (8GB sitting
around doing nothing). Now I had two partitions separated by 60Gigs.
And I hate LVM. So I installed btrfs[2] ;p

Don't worry, the partition will only contain expendable test data, and
the unstable explosiveness of btrfs adds a level of excitement to the
project =)

The third thing that I did (after talking to robbat2 about hosting the
code on gentoo) was move the code to git. The reasons for this are
simple: Gentoo already has the required Infra to manage git
repositories, and bzr-http uses too many GETs to be a viable hosting
solution (that's what brought http-svn down apparently).

So, the code is now on git[3] (git-fast-import + bzr-fast-export
ftw!~), and so will the jobtage tree :)

The diff stat for the past week has been:

$ git diff --stat --find-copies-harder -M -C a609bd1d..
 .gitignore                                     |    2 +
 autotua/__init__.py                            |  167 ------------------------
 autotua/bin/jobuild.sh                         |    6 -
 autotua/sync/__init__.py                       |   38 ------
 scripts/minimal-portage-snapshot.sh            |   20 +++
 slave/TODO                                     |    3 +
 slave/autotua/__init__.py                      |  107 +++++++++++++++
 slave/autotua/bin/jobuild.sh                   |   13 ++
 slave/autotua/chroot/__init__.py               |  126 ++++++++++++++++++
 slave/autotua/const.py                         |   78 +++++++++++
 {autotua => slave/autotua}/fetch/__init__.py   |   10 +-
 {autotua => slave/autotua}/jobuild/__init__.py |    0
 slave/autotua/sync/__init__.py                 |   90 +++++++++++++
 slave/test_modules.py                          |   31 +++++
 14 files changed, 472 insertions(+), 219 deletions(-)

git-fast-import seems to have not detected the move (done in bzr) from
autotua to slave/autotua properly :)
This week:

    * jobuild parsing plz, kthx :)
    * And a complete slave, ftw!

1. http://archives.gentoo.org/gentoo-soc/msg_7b8cf5c6033d78f8afb3d5919b81560b.xml
2. http://btrfs.wiki.kernel.org/index.php/Main_Page
3. http://git.overlays.gentoo.org/gitweb/?p=proj/autotua.git;a=summary

-- 
~Nirbheek Chauhan
-- 
gentoo-soc@g.o mailing list