CleverBlogName

So there I am, listening to the radio, and what comes on but a creche ad for a chain of creches called “Paradise.” I’m sure it seemed like a great idea and all, but hearing repeated exhortations to “send your children to Paradise” just makes me think of cultists or baby suicide bombers.

This is not persuasive.

One of the best things about attending the Linuxconf is the renewed sense of enthusiam for my field (lightly dampened yesterday by battling with @#@^#%$!^% PulseAudio, which is the worst thing to be inflicted on desktop Linux in a long time, and today on arriving home to discover a household box had literally cooked itself, likely beyond repair).

A lot of the last near-decade has, for me, seen my interest in what I do wane, damn near finished off by two years of release management. I had become, well, barely a tradesman, practically an assemblyline worker; interested in doing my job properly, but largely devooid of a vast care factor beyond that.

The last while working on zLinux has helped considerably with that, and LCA has piqued my interest mightily; I want to get back into hand-rolling Postgres releases to play with the new features, I want to play with new technologies and tools and follow their development, not just because I want to be better as a craftsman of my job, but because I can glimpse my former interest in the art of what I do; of doing a thing for itself and that alone, rather than merely because it turns a crust. On that front, LCA is a huge personal success as well as providing me with a bunch of professional value, and that’s (hopefully) a genuine improvement in the whatsit of my life.

Today was the last (formal) day of linuxconf 2010, and since I was a “professional” delegate I went to the dinner, and took Maire and Ada along, too. It was a pretty decent night out, with a kapa haka group, a chance to catch up with gnat after way too many years, sitting with Tridge at our table (do coding skills, like celebrity, osmose through proximity? If so, I should be able to do the best work of my life after bathing in his aura), and having Liz give Ada a ride in her wheelchair (which was the coolest thing of the evening, I reckon).

Ada was magnificent. We left early, but still didn’t get home until 10, which is two and a half hours past Ada’s usual bedtime, and she was brilliant. She drew, she ran around Civic Square, she chatted to people, she sat in the lobby of the Town Hall with me and scrutinised the geometry of the tiles that made up the floor (“Un, deux, trois, quatre triangles in the square!”), and at quarter to ten she explained she wanted to go to bed, all with nary a grumpy moment for the evening. I feel so lucky with her.

I discovered that I’ve forgotten large chunks of Ka Mate.

The shame.

The quote of the conference is still “Twitter: Release early, release often—for thinking.”

L3wt soon, but first:

...<genuflect type=“waynes world”>not worthy, not worthy<genuflect>

Paul Mackerras

perf superceeds operfmon and similar tools for understanding.

• perf_event; initial proposal in Dec 2008; developerd through 2009 in the tip tree.
• In mainline in 2.6.31.
• Initially supported recent x86 processors and 64-bit POWER.
• Now have hardware support for SH and sparc64.
• Most have at least basic support, as long as you have high-res timers.

Andrew Tridgell, Samba Team (not a lawyer)

The talk is about how you can understand, as an engineer, how you can talk to patent attorneys, how to understand the language, do analysis.

Understand how to lower exposure to patent attacks, and how the community can avoid that; Tridge has a concern these will only become more common.

Patent lawyers are, like a platypus, shy, industrious creatures. Getting time with a patent lawyer can be a challenge, especially if you don’t have access to one through work. It’s great to understand how you can communicate efficiently with them.

Josh Berkus

A very excisting time for PostgreSQL, since the final commitfest for PG 9.0 development.

8.5 had gone by the by because the feature set is going to be SO DAMN HUGE. 64 bit windows, exclusion constraints, JSON/XML EXPLAIN output, host standby, sync replication.

What’s a commitfest? It’s a way of dealing with scarce committer and reviewers time; review patches faster, sooner, every patch, and train reviewers. People were getting patches knocked back or forgotten because of a lack of reviewer time, which was making people unhappy.

Every other month during the development period, we clear the queue of patches; this also makes it transparent as to what’s happening in the development world. It lets you help with failing patches if you care about them, too.

Four commitfests, followed by cleanup, beta testing, and then a final release. Version 9 should be released mid-way next year.

Dan Magenheimer

(This is a bit incomplete, I’ll update it Soon.)