The roads I take...

Here's a short summary of Mozilla-related work I've done in weeks 42/2011 (October 17 - 23, 2011):

• Mozilla work / crash-stats:
  Did some more improvements on the filters for per-component crash reports.
  Exchanged emails with Jesse on his stack-blame scripts and ended up filing a bug on making them an on-demand web service.
  Worked with the Socorro team to get a problem fixed that some crash aggregation got stuck over the weekend.
  I also gave some feedback on planned patches for Socorro so that the end result should be really helpful to us.
  As usual, I also watched new/rising crashes closely and filed bugs for a couple of those.
• SeaMonkey build&release:
  Followed all the news on the main SeaMonkey VM host going down, but not much for me to do there, Callek and the Mozilla IT team have things as much under control as possible under the circumstances.
• Themes:
  I did some more work on my themes and finally uploaded 2.4 versions of both EarlyBlue and LCARStrek which are awaiting reviews now.
• Various Discussions/Topics:
  "Native UI" for Firefox on Android, how/where to land experimental features, background update work, my tablet being faulty, N9 prices and availability, etc.

The sad story how Mozilla might let go of the superior web-style-markup-based add-on system on Android devices in the future to be more easily competitive on performance shows me once more what sorry state mobile devices are in right now and why we badly need B2G be successful soon and graduate from an experiment to a full-fledged product as soon as humanly possible. In my eyes, Mozilla's mission requires us to use as open as possible and as "webby" as possible technologies wherever we can, and turning away from Gecko-rendered UI back to the (IHMO) stone age of using OS-native UI is not just against that but also against innovation. And while I see the arguments of why Android phones right now seem to need this, I don't like it - and I don't see any reason to even think about doing the same on tablets. I just hope I'll never need to use that "native UI", but then I surely won't buy and Android phone anyhow - but probably use an Android tablet until I can switch to B2G. Making the whole user-facing stack be driven by the web renderer is a real Mozilla way to go, using Open Web technologies from the bottom up, being innovative, creating new opportunities. That's what I signed up for with putting my time into the project, and that's what I want to see succeed. If you're looking for that spirit, please help the B2G project and help to make it succeed and make mobile devices a better, more open, more people-friendly place!

Here's a short summary of Mozilla-related work I've done in weeks 41/2011 (October 10 - 16, 2011):

• Mozilla work / crash-stats:
  Did some refactoring of the categorization code in my experimental per-component crash reports, and finally improved categorization by some additional filtering. New reports like the Firefox 8 Beta one for yesterday should have a quite usable component split and top 10 crash signatures for those components.
  Created weekly bug statistics for both Socorro and crash bugs, which compare newly filed bugs to fixed and otherwise resolved ("triaged") bugs in every week.
  Updated numbers for categorized Socorro bug lists once again and made a spread sheet and graph (not public right now) on the development of those numbers.
  Had a meeting with Naoki on how to coordinate on mobile wishes and Socorro.
  As always, I investigated various new or rising crashes and made sure bugs get filed for them in the right places and get the needed attention.
• SeaMonkey releases:
  Finally, I also put up SeaMonkey update snippets for linux64 updates to 2.4.1 - so everyone should get up to the latest release now. I've been asked if I could do that for betas as well but I find this too much work to do manually. If someone can script it, that would be a way, but then, we hope that soon we'll get the machine power to run tests and localized builds on linux64, and then we'll be able to provide official builds with updates right away in the normal process.
• German L10n:
  Updated German localization for SeaMonkey and core on Aurora as well as Nightly (except the actual "toolkit" parts for the latter) to match the original.
  Also, addressed a number of German L10n bugs on "select add-ons" UI (review comments), translation of "corrupt", and "beinhalten" being a suboptimal word.
• Add-ons:
  I also refactored my "KaiRo.at Mandelbrot" add-on to fit in-tab UI a bit better as well as eliminating its XULRunner app version and making it be an add-on only. The new 4.0b3 version is up on AMO and waiting for reviews.
• Various Discussions/Topics:
  Mozilla and "webby" or "native" solutions for web gaming, "Native UI" for Firefox on Android, Sync server issues and delaying MU for 3.6 users, how people got involved with Mozilla, N9 being available in Austria, etc.

This week ended up pretty productive both in work I'm paid for and some I'm not - but my list of TODOs, while significantly reduced, is still long enough that I can't get bored.
That said, while I got new external and internal hard drives for my desktop running fine, that ASUS Transformer tablet I got from Mozilla is acting strangely and might have a failure of some kind, with applications or the whole tab crashing or freezing randomly. I wonder if that's a hardware or software problem and how easy it will be to correct, given that the device belongs to Mozilla but there's no office and/or Mozilla IT people anywhere near here.
It's a cool device even despite running Android, but it's much cooler when it actually works...

It's been a long time since I last posted anything here, including a status report, due to various reasons, but I figure it's still good to have a record of what I worked - even for myself.

So, here's a short summary of Mozilla-related work I've done in weeks 36-40/2011 (September 5 - October 9, 2011):

• Mozilla work / crash-stats:
  Tested some Socorro release that went online and communicated a lot with the Socorro team on things they have been working on, like fixing how connected bugs are reported or upcoming features like a hang pairs report or a signature summary as well as some skiplist additions.
  Talked the CrashKill team's Q4 priorities for Socorro over with leads of both teams and put them up on the wiki.
  Had some interesting discussions with various people at the Mozilla all-hands meeting on crash-related topics.
  Talked with metrics about the possibility of leveraging their systems to create more flexible graphs of crash data, the first step is getting access to the data.
  Invested more work in my prototype reports, next to updating them for various version changes, the components reports are still somewhat experimental, but start to get really useful.
  As always, I investigated various new or rising crashes and made sure bugs get filed for them in the right places and get the needed attention.
• SeaMonkey releases:
  I did put up SeaMonkey update snippets for linux64 updates to 2.3.3, the bug for 2.4.1 is filed but I didn't get around to this yet.
• German L10n:
  I did most needed updates needed for German SeaMonkey 2.6 (now in Aurora) and a number of core strings also needed by Firefox 9.
  On the first weekend of October, I also took part in the German Mozilla community meetup where we discussed a lot of matters surrounding the localization and community in German-speaking countries.
• Themes:
  More work on updating EarlyBlue and LCARStrek for 2.4 versions, there's a bit more stuff I need to check there still, though.
• Various Discussions/Topics:
  Mozilla all-hands, ESR proposal, Web APIs, B2G, etc.

Surrounding the Mozilla all-hands meeting in San Jose (where a mindboggling almost 600 people being paid by Mozilla gathered), working another week from the Bay Area and the comparatively tiny German community meetup, I didn't really get around to writing up those status reports, and I'm wondering how helpful it is in general to do those.
I obviously found it reasonable to do another one now, mostly because it's also nice for me to have some log of what I have achieved and what I spent my time with - but I wonder how useful those reports are nowadays to other readers of this blog, and if every week is the right format or if I should always combine multiple weeks in a report. Any opinions?