Re: [proteowizard-developer] Fixing TeamCity Builds

[Matthew C. <mat...@gm...>]

Before msparser, everything was passing except the two timing out builds. I will claim 
responsibility for those probably (it's the CLI bindings' MSDataFileTest that's timing out).

Msparser is causing some weirdness I don't understand in Boost Build and I really don't have time to 
debug it. So if Rob can't do it, I'd just disable msparser in identdata for Windows. We seem to be 
awfully close to having it working though.

-Matt


On 3/29/2012 12:47 PM, Brendan MacLean wrote:
> Hi all,
> I would like to call a moratorium on commits to the project that do not fix a broken test on
> TeamCity, and I would like to make sure that every broken test has someone actively working to fix it.
>
> At the point where we give up on getting everything back to green, then we will have to admit that
> the failing tests need to be moved into the ProteoWizard Experimental group until we can get them
> passing again.  As it stands today, we are just have things failing all the time, and it is hard to
> tell when a regression happens, and whether anyone is working on them when they do happen.
>
> I know we started it with the VS 2010 port, but that was a very difficult problem that would have
> been even more difficult without going through some TeamCity pain.
>
> But, now we really, really need to get serious about getting back to a stable state, especially if
> our manuscript is soon to be published in Nature Methods.
>
> So, please no commits that are not intended to fix the TeamCity builds.  I just reverted a change
> that appeared to break several tests.  Once that has gone through the system, I will start checking
> in with owners of the remaining failing tests.
>
> Thanks.
>
> --Brendan