Libre Application Summit, Heroes of Fedora, commenting, and so much more!

Hmm, I didn't blog for a while and now I feel like I have three posts all coming along at once...

Best thing I read today: The Register poking the Apple PR department with a series of blunt sticks

Best thing I read yesterday: Jenn Schiffer on blog comments. It took me years to figure out what the article illustrates so perfectly: sometimes (especially if you're male...) you will want to comment on something not because you can actually enlighten, inform or entertain the author or other readers in any way, but just because, well, you're reading it and you have opinions and you're sure everyone else wants to know about that, because you're pretty great.

Don't.

All her other posts are pretty hilarious as well, but the comment sections take the cake. So many angry nerds with STRONG OPINIONS and absolutely zero sense of irony...

Anyhoo! In less random news: I'll be at the Libre Application Summit in Portland next week, in case anyone's interested. There may be some sort of openQA BOF there, depending on who turns up. It'll be fun! Come along! Also Portland is great, so long as you make sure not to get run over by a moustachioed hipster on a fixie on his way from the artisanal coffee shop to the craft beer bar. It's like New York - one of the few places that precisely lives up to its stereotypes...

I find myself with a bit less to blog about lately because I've sort of gradually shaded over to working more on a constant cycle of figuring out why tests are blowing up and trying to help keep the unstable branches composing and working every day, and less on 'community facing' events like Test Days and the like. And - much to my constant surprise - the hideous framework of hacks, bodges and StackOverflow-derived code that constitutes the release validation system seems to keep working more or less smoothly so I don't have to think about it very much.

Nevertheless, the Fedora QA community continues to rock and it's not like Test Days and manual tests and all that other fun stuff isn't happening! Geoff (aka coremodule) did a great job of writing up the Heroes of Fedora 24 blog posts:

It looks like there are some encouraging upward trends from Fedora 23, so that's great news - the wave of new contributors in the last few months has really been doing great work, so thanks to everyone.

Meanwhile, Sumantro and Petr are working on the Fedora 25 Test Days (plus I have to find some time to wedge a Wayland one in there at some point). We should be having events on cellular modem support, storaged - the replacement for udisks, and Fedora Media Writer, so look out for those coming up soon!

Fedora 25 Beta is slated for October 11th, with the freeze coming on September 27th, so those are the next dates we're looking towards for the release process. Current Beta nightly compose testing is looking pretty good so far.

openQA continues to expand its empire by degrees. In the last few weeks we've added tests for text installation, the Fedora Server database server role, KDE and GNOME default browser, the FreeIPA web UI, NFS installs and updates.img sourcing and upgrading an encrypted system. We also now have the capability to forward openQA test results to ResultsDB.

I did quite a lot of work in the last few weeks to try to fix some issues - including bugs in openQA, bugs in the tests, and bugs in Fedora itself - which could cause invalid failures and inconsistent results from day to day, with the aim of making the test results more reliably comparable between runs. For the last few days the results have been pretty consistent, pace a few remaining bugs which occur only intermittently (so sometimes a test will suddenly fail not because of a change in that day's compose, but just because it happened to run into an intermittent bug). This makes it a bit more feasible to think about plans like gating Rawhide pushes on test results, or running openQA on candidate updates and providing feedback via Bodhi.

Comments

No comments.