We have a package (mingw32-openssl) which has some tests that run
under Wine. Strictly speaking, running the tests isn't necessary, but
because it's such an important package I feel that we should run the
Anyhow, Wine needs an X server, although it doesn't really use it for
anything. To provide one, I'm using Xvfb like this:
Xvfb $display & xpid=$!
trap "kill -TERM $xpid ||:" EXIT
make -C test tests
This works, but there are two problems. Firstly there isn't a good
way to choose a free display number, so I'm just picking one at random
here. This assumes that port 6021 is free, also disallows any builds
that happen in parallel. Secondly, will Koji let me do such a thing
at all? Is there another / a better way?
Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones
virt-p2v converts physical machines to virtual machines. Boot with a
live CD or over the network (PXE) and turn machines into Xen guests.
Don't have much personal interest glib/gtk+ anymore, so I'll let them go
free to a new loving home.
Mostly, I didn't have much incentive or energy to fix a recent FTBFS gtk+
I'm going to be building RC3 of Mono 2.0 for release tomorrow. It's not
broken anything big time here, but it's worth mentioning that I've
backported a few extra patches into RC3 which haven't made it to the RC3
release, but are due in 2.1 (I think that's due end of 2008/start of
Updated : libgdiplus, mono, mono-basic, xsp, mod_mono, monodoc,
Please let me know if it kills anything big style!
Sie können mich aufreizen und wirklich heiß machen!
I understand why rawhide should be frozen at some points, but sometime
there is a package one really wants, that is built in koji, but not
available in rawhide because it is frozen. In my case, for example,
rpm-build is broken but aleady fixed. So maybe it would be worth having
a special repo that would not be enabled, but that would allow to have
the latest builds during a beta freeze. In my opinion this repo should
also be used for Matt rebuilds since there are alot of already fixed FTBS
in his reports due to the freeze, which is a bit annoying.
Does anyone want to maintain GEF? I don't have time for it and haven't
updated it recently. It needs to be updated to the latest (Ganymede
train) release. Nothing depends upon it so if no one wants it and we
remove it it's probably not the end of the world.
I've given up ownership in FAS so feel free to take it there if you want
I have some services found being activated by default that should be
removed for the following reasons:
1. sendmail: starts way too slow, is not usefull for any normal desktop
user I know. Making it usefull requires configuration so I assume
wohever uses that service _can_ activate it.
2. ip6tables: I do not know of any provider actually working with ipv6.
So I assume the mass of all users do not need it.
3. isdn: isdn requires configuration and thus should be set to start
when that config is actually done.
4. setroubleshootd: That service also takes long to boot, but its quite
usefull. I wonder if one could make auditd start setroubleshootd when
required - having two daemons working on base of the same informations
seems not very clever.
So, now go on and punsh me ;)
On Tue, Sep 30, 2008 at 09:37:52PM -0500, Don Harper wrote:
> Open a bug to put sbin *after* bin? I have had sbin in my path for
> years, and I do not have those issues, and I always set bin dirs before
> the sbin dirs.
Putting sbin after bin does not change anything.
Since you've had your own system set this way for years, you don't know what
you're missing with bash_completion.
Matthew Miller <mattdm(a)mattdm.org>
Senior Systems Architect
Computing & Information Technology
Harvard School of Engineering & Applied Sciences