On Wed, Aug 29, 2012 at 10:19 AM, Daniel Veillard <veillard(a)redhat.com&gt; wrote: > I was a bit afraid of the .0 effect on release name, but we really > have a number of problem with 0.10.0 that ought to be fixed in a > "brown paper bag" release. I would urge people to report and try to > fix the problem being raised in 0.10.0, let's try to get them fixed > today or tomorrow and I will cut a new release on Friday, > > Sorry about that, but somehow we didn't managed to catch even serious > problem during the freeze, at some point we need to fix the problem of > testing the code that we push on git on a daily basis, we have many > tools but we lack at doing the continuous testing :-\ > > Daniel > Daniel, Would it be worth creating a 0.10.x branch and just keeping it as a stable branch until next month's release?

On Wed, Aug 29, 2012 at 11:19:42PM +0800, Daniel Veillard wrote: > I was a bit afraid of the .0 effect on release name, but we really > have a number of problem with 0.10.0 that ought to be fixed in a > "brown paper bag" release. I would urge people to report and try to > fix the problem being raised in 0.10.0, let's try to get them fixed > today or tomorrow and I will cut a new release on Friday, > > Sorry about that, but somehow we didn't managed to catch even serious > problem during the freeze, at some point we need to fix the problem of > testing the code that we push on git on a daily basis, we have many > tools but we lack at doing the continuous testing :-\ Would setting up Jenkins to trigger libvirt-tck either on commit or nightly help?

I could try to find some free time to do this.

It's a bit tricky since we'd be testing libvirt, libvirt-tck, Sys-virt and Qemu at once but it might be worth a try to get us more continous testing.

On Fri, Aug 31, 2012 at 02:36:07PM +0800, Daniel Veillard wrote: > On Thu, Aug 30, 2012 at 08:44:57PM +0200, Guido Günther wrote: > > On Wed, Aug 29, 2012 at 11:19:42PM +0800, Daniel Veillard wrote: > > > I was a bit afraid of the .0 effect on release name, but we really > > > have a number of problem with 0.10.0 that ought to be fixed in a > > > "brown paper bag" release. I would urge people to report and try to > > > fix the problem being raised in 0.10.0, let's try to get them fixed > > > today or tomorrow and I will cut a new release on Friday, > > > > > > Sorry about that, but somehow we didn't managed to catch even serious > > > problem during the freeze, at some point we need to fix the problem of > > > testing the code that we push on git on a daily basis, we have many > > > tools but we lack at doing the continuous testing :-\ > > > > Would setting up Jenkins to trigger libvirt-tck either on commit or > > nightly help? > > Definitely !!! > > > I could try to find some free time to do this. > > To be honnest i was thinking for months of doing something like that > but i don't have a good infrastructure. Where would Jenkins run ? > what kind of hardware set would be needed etc ? It'd start simple with a single VM running jenkins and the builds: http://honk.sigxcpu.org:8001 We can then add build slaves for other distros and operating systems step by step. For the moment I have enough resources here. The above currently only does make, make check and make syntax-check. I'll try to add libvirt-tck during the next days. There's also some duning to be done.