I’m still not rich, but I really don’t want to be as I’m able to do what I want, have a job and a house.

I still am not healthy or fit as I want to be, but I’m working on it.

I still have way too many interests and projects, but I would rather have that issue than be bored all day.

The biggest thing is that Elaine and I are still together and we seemed to have “turned the corner” and left behind the cloud that arrived over us in the final year before her father passed. We still miss him, her much more so than me.

So yea, 2011 is gone and guess what – there be dragons out there. Yep, 2012 is the year of the dragon!

Some things we now know to be good ideas:

• Writ­ing op­er­at­ing sys­tems in a com­piled ma­chine-in­de­pen­dent lan­guage
• Per­form­ing file I/O by read­ing, writ­ing, or over­writ­ing in­te­gral num­bers of bytes at in­te­gral off­sets.
• Cre­at­ing processes by du­pli­cat­ing ex­ist­ing processes.
• Null-ter­mi­nated byte strings.
• In­vest­ing a sub­stan­tial pro­por­tion of pro­gram­mers’ time in build­ing tool­ing to make them­selves more pro­duc­tive.
• When ex­plain­ing a new pro­gram­ming tech­nique, start­ing with “Hello, world”.

It’s hard to be­lieve that there was a time when any of these weren’t con­ven­tional wis­dom, but there was such a time. Unix is com­posed of more ob­vi­ous-in-ret­ro­spect en­gi­neer­ing de­sign choices than any­thing else I’ve seen or am likely to see in my life­time.

It is im­pos­si­ble — ab­solutely im­pos­si­ble — to over­state the debt my pro­fes­sion owes to Den­nis Ritchie. I’ve been liv­ing in a world he helped in­vent for over thirty years.

Not today ;) – thanks to a lot of work by the RelEng team to be on top of any issue my Monday is going fine

wait-times-report

We also landed a number of test infrastructure fixes that Joel over on the ATeam worked on which enable reftest, jsreftest and crashtest coverage and also fixed some small issues with the current Talos tests.

MOAR GREEN!

The awesome IT folk have installed PDUs (remote power switches) which will allow us to remote power-cycle Tegras when they get wedged. This will reduced the offline time for the average Tegra.

As we move forward with a faster development cycle, the automated testing should now be able to keep up :) and give the developers timely feedback.

update: Aki sent me a photo of the rack-o-tegra’s that we used before IT hooked us up with real racks

and a screen shot of the Tinderbox waterfall showing the tests that are live:

After a very accurate and firm review by a co-worker I have to realize that the “oh, I just need to tweak this” attitude isn’t working at all so i’m doing a personal reset – going to go back to small simple steps/changes and make them my priority. This isn’t to say that i’m going to stop all of my personal projects but it means that they will now be in the personal side of the task list and work will be in the work side.

For most folks this may sound silly or they may be shaking their heads, but for me i’m trying to stop the wild swings between “I can do anything” and “oh crap, why am I fucking everything up”. This reset is also going to be for my eating and excercise also – as I have seen others go thru a similar change with success.

So, step one is make a list of things i’m doing and start assigning time to them with the first being this:

Just because you can do something, doesn’t mean you should

well, at least that’s how I think it goes – Google wasn’t being helpful in finding a citation. So, back to the post.

I’ve run my own mail server ever since I could afford to pay for a remote server – heck, even before VM slices were common. That also means that i’ve managed my own anti-spam set of tools and have had to maintain them on a regular basis to keep up with the changes in tech.

No more. I’ve tossed in the proverbial towel!

I’ve been listening to Leo rave about Mailroute.net for a long time but it’s always gone in one ear and out the other as, you know: I don’t need that service – ha! After spending another couple of hours today updating my server, checking logs and looking thru my spam folders (yes, folders – one on the server and a couple on my local IMAP server) I realized that was time that I could be doing writing code or something, heck anything, else.

Sure it’s a subscription service and sure I am now losing some control, but it seems to make sense at this point in time. I’ll check back in later to see if it still seems like a good idea after a couple months.

Thanks to stpeter for the reminder of this aniversary.

Once Joel wrangled the many tests that had subtle dependencies on device and operating-system, we began to pair up getting each test suite to run from the command line. After getting that stage done, we started in on the tool changes needed with him fixing or filing bugs for the auto-tools side and me getting the existing Talos build steps to “know” about Android and mobile testing in general. Previous Talos testing was all on the same device as the test itself, but mobile introduces a new wrinkle – the device being tested often cannot also run a web server or any of the support tools that the testing framework requires.

Once I coded the buildbot proxy to control each buildslave based on the heartbeat signal coming from the Tegra and Aki pounded my patches to the buildbot factory and config into shape, we landed them to our internal staging environment last week and I’m happy to say that we have put the three Tegra’s we have available into a puddle (a bit too small to be called a pool ;) )

For those who enjoy the Tinderbox waterfall (which I do) – you can see them in all their Green glory here and one of the tests shown in gory graph detail here.

This is a huge win for the developers and it couldn’t have happened without a lot of cross-team work: Joel, Clint and Bob from the auto-tools team, the mobile dev team and a big shoutout to Aki on the releng team for mentoring me thru the new stuff and slogging thru the many last minute details to get this moved from Staging to Production.

We also realize that this is just the beginning, heck, it is working for three Talos tests and none of the reftests and it’s only running on three Tegra’s! But all of that is being iterated on and tests will be added as soon as they clear the auto-tools team :)

thanks!

Hmm, I know VCALENDAR and can hack-parse html …

So 3 hours later and I have a very basic program that reads the Schedule wiki page, sucks out all of the tables that represent the daily schedules and generates a .ics ready for importing into your favourite calendar program.

Requires Python 2.6 or 2.5 with ElementTree installed: generate_summit_calendar.py

enjoy!

\o/

This was an interesting task because it involved a little bit of each job in the RelEng world at Mozilla:

I had to update the puppet manifest which we use to distribute files and manage configurations for the build farm so the signing keystore and configuration file would get to all of the production build servers. I also had to write a small Python program to make the actual call to Jarsigner because if we just called it from within the Buildbot step, the output would include the keystore passwords and that appears on the build server’s web page – not good.

I also had to drill deep into the various parts of the Mozilla Buildbot configuration and custom code because the Android build was new and just different enough from the other mobile builds that it wasn’t just a cut-n-paste solution.

And all of this I was doing the Mozilla way for the first time and let’s just say they have some unique custom code :)

But today the first half of the task was completed and that allows the nightly signed builds to be made public. What remains is to remove some work-arounds (for the packaging and tests) and then to start building the steps required to generate a release version of Fennec for Android.