I'll be vacationing from tomorrow to the 1st of July (be back in
California July 2nd.) My packages should all have open acls so feel
free to fix and I've added myself to https://fedoraproject.org/wiki/Vacation
There's some changes in the packagedb that are very nice but I don't
want to push them while I'm on vacation as I've noticed some slowness on
certain queries. :-( If someone else wants to do so, I'm totally fine
with it as long as you have the old packages available to rollback in
case we start getting timeouts.
python-fedora-devel has been updated to a new version. But there's a
few very strange things happening in the new jsonfas interface. I've
clued lmacken in on that so if he has time to work on it we could end up
pushing the new stuff to some test servers. Otherwise it will probably
wait until after I get back in July.
I'll stop by on IRC once in a while but I'm going to make a conscious
effort not to check my email ;-) See you all in July!
although I've been around in Fedora land for quite a while, lurking on
this list and in #fedora-admin (besides other lists and channels), it's
time for me to step up and do some work with you guys, The Great
The talk Mike McGrath has been giving today during FUDCon BarCamp,
regarding Community Services Infrastructure, inspired me that this is
something I can help with ;-) So, here's the officially required
introduction of /me, so you guys know who I am and where I come from.
My name is Jeroen van Meeuwen, aka kanarip, aka Dutch (as both the other
names are somewhat unpronounceable for some people), and I'm 24 years
old. I'm from the Netherlands (and proud of it), and I have done a
little something here and there within Fedora before. In short I'm a
developer of revisor, pyjigdo, a packager for several packages in Fedora
and EPEL, an Ambassador and advocate of Free and Open Source software,
the vice-president of Fedora EMEA, the composer of the Fedora Unity
Re-Spins, and I have a couple of things pending like the SUMS Update
Management System (SUMS), and puppetmanaged.org.
As far as my dayjob is concerned I'm a Senior Consultant, certified in
Linux, proprietary North American vendor crap and Cisco.
My plan for getting involved with the Infrastructure team is basically
two or three parts:
- CSI, for the greater good and because it'll be a huge help with
- Puppet, for Fedora Infra obviously, but also CSI and again
- Nagios & maybe cacti just for kicks
Thank you so far for doing a great job and I look forward to work with
Jeroen van Meeuwen
OK, now that Mike's let the cat out of the bag, let's review the
status of the Fedora Asterisk server:
1) Everything we currently use is fully puppetized (or at least I hope
so). That includes a cron job that syncs user information from FAS at
10 and 40 minutes after the hour.
2) There are two conferences set up sip:email@example.com and
3) Dialling in from the DID numbers work, and the directory appears to
1) Need content on talk.fedoraproject.org to help folks set up
clients. Max has a done a good job of getting the boilerplate ready so
now we need to do the hard work of writing the content.
2) Better system for provisioning conferences. Some of the
requirements I have in mind are:
a) password protected conferences for things like the Fedora
Advisory Board meetings
b) a way to stream audio using flumotion to a larger listen-only
audience (for the town-hall meetings)
c) an easy way to provision "normal" conference rooms for other groups
Some of #2 was implemented during our testing phase, so we'll be able
to re-use some of that work.
I have to run for the bus right now.
ricky noticed that hosted1 was down and we took a look. xen on
serverbeach4 was having issues. We tried to bring it back but no joy.
Currently sb4 is not making one of the disks available to hosted1 and
we're still getting corrupt database messages in /var/log/messages::
Jun 17 12:07:21 serverbeach4 xenstored: TDB error on read: Corrupt database
Jun 17 12:07:21 serverbeach4 last message repeated 3 times
Pastebin of ricky and I's IRC transcript here::
Backup of old xenstores is in /var/lib/xenstore.20080617*
Since we switched from Moin to MediaWiki, all the outgoing links from
the wiki have the attribute rel="nofollow", which instructs the search
engines to ignore the PageRank value and do not influence the link
This attribute was introduced in MediaWiki as a countermeasure against
spammers which tried to abuse Wikipedia, linking to their own malicious
websites. While there is some merit of using this at Wikipedia, where
even unauthenticated users are allowed to post, other people do not
think this is fair and consider Wikipedia acts like a black hole of the
web, absorbing PageRank without giving anything back, after it used all
the PageRank received to build its reputation (this is also my opinion
and when I link to Wikipedia in my websites I always use rel="nofollow").
The situation in our wiki is very different: we allow posting only from
authenticated users which have a FAS account, so the editors have to
pass a number of filters and leave enough identification data.
Personally, I hate to be treated as a spammer (such treatment is a
result of using rel="nofollow" by default). *If* I am found abusing the
system for personal gain, then kick me out and remove all my rights,
otherwise I believe I deserve to be considered an innocent user.
I also think there are legit uses for "dofollow" links in our wiki: we
want to give PageRank value to friendly or sister projects, like pages
from fedorahosted.org, fedorapeople.org or entities like GNOME, Red Hat
So, based on the argumentation above, my proposal is to trust the users
and remove the "nofollow" attribute.
nicu :: http://nicubunu.ro :: http://nicubunu.blogspot.com
Cool Fedora wallpapers: http://fedora.nicubunu.ro/wallpapers/
Open Clip Art Library: http://www.openclipart.org
my Fedora stuff: http://fedora.nicubunu.ro
For your consideration.
-------- Forwarded Message --------
> From: Mads Kiilerich <mads(a)kiilerich.com>
> To: webmaster(a)fedoraproject.org
> Subject: Wiki migration - lost content
> Date: Sun, 08 Jun 2008 00:19:24 +0200
> I don't know who to contact about wiki migration issues.
> doesn't mention any contact.
> Congratulations with the wiki migration. It was 99.9% perfect. Well
> done, no doubt about that!
> Comparing https://fedoraproject.org/wiki/YumUpgradeFaq and
> http://fedoraproject.org/wikiold/YumUpgradeFaq I can see that old "Tips"
> boxes with several lines has been migrated to literally
> "YumUpgradeFaq/code" and a couple of fragments on the talk/discussion page.
> It is OK that formatting wasn't migrated 100%, but it is not acceptable
> that content is lost. For this specific page I can fix it manually, but
> it would be very sad if content was lost randomly all over the wiki. I
> suggest that the scope of the problem is analyzed and an appropriate
> solution is found. A solution compromise could be to programmatically
> add a cleanup-request and a note with the lost content to the affected
Ignacio Vazquez-Abrams <ivazqueznet(a)gmail.com>
PLEASE don't CC me; I'm already subscribed
Why no people uses Ruby to write the Script ! In FI,we write script in use Python,Perl ,Shell etc. I think its is power like Perl in Sysadmin.
The Ruby creator says "I wanted a scripting language that was more powerful than Perl, and more object-oriented than Python." I suggest someone to learn Ruby!It is really very good!
Hello to other web app developers,
Xavier Lamien sent a message to me today about incompatibilities in
database schemas between FAS versions 0.8.4 and 0.8.5. This prompted me
to think about how we can help people update their databases if changes
occur between versions of the database. For this particular update I
added sql scripts to FAS in the updates/ directory and listed how to
invoke them in the README. If you are updating your database schema and
you are also packaging the software for general consumption (for
instance, packaged in rpm form in Fedora proper), please consider how to
manage this for your users. There's a variety of ways ranging from what
I did to the `migrate`_ project (I haven't seen this appear in Fedora
yet. Any takers?) leveraging tg-admin sql might work for simple cases
like adding a new table but I haven't found that it will update existing
In any case, remembering to document the procedure for other members of
the code will be great.
.. migrate: http://code.google.com/p/sqlalchemy-migrate/
I decided to get more involved in fedora project (i.e do more "technical
stuff" with the fedora project as i'm very interesting about fedora
infrastructure) and would like to contribute to FI.
I'm a system engineer working for an Open Source Software Company called
I'll try to resume a bit all i do all days at work and for fedora just
== Fedora services with what i'm already familiar and used to work: ==
* Buildsystem :
- Koji server/client : deployed at work in our infrastructure for best
support for our client.
Also working to make koji able to deal with
.deb builders (called pbuild) as i have collegues who work on the .deb part.
automate the database interaction with
- plague server/client : also deployed at work .
- CVS + Sources control : deployed at work with a look-aside-cache-like
settup on RPM Fusion
* Account system :
- FAS2 : Only on RPM fusion infra.
We use something else at work for this part :(. but my
employer is really interesting by FAS and FDS for our clients.
* and FDS
== System with which i'm familiar : ==
RedHat 8, 9, el4, el5
Solaris 8, 10 (we only update on release n+2 for technical reason)
fedora (of course!)
== sotfware knowledge ==
proxy (like squid) server
apache server (also v2)
all related repository tools
glpi (including puppet deployement) [one of my collegue is an puppet
zsf, svm, global domain (solars part)
== additional ==
used to work with python, bash, shell, zsh and ksh script
well, i don't know if that enough to you to know me more.
If you have any additional question, just ask.
PS: i'll be at Redhat summit (as hexibitor) if you would like to alk about
GPG-Key ID: F3903DEB
Fingerprint: 0F2A 7A17 0F1B 82EE FCBF 1F51 76B7 A28D F390 3DEB