For me at least, this was one hell of a YAPC::NA.
For starters, begin with 24 hours worth of flying, including a flight where I landed before I took off, and zero bodyclock pre-loading, so utter and complete jetlag hell.
Next, toss in the adventure (that is, a significant endeavour where the outcome is unknown) of trying to keep planning far enough ahead on my homeless, jobless 3 month trip around the world that the trip won't fall in a heap, or I won't have to fall back on the emergency "adventure recovery" cash.
Finally, add the nightmare of not only having to write 2 talks between landing and 11am on the first day, but discovering after landing that the THIRD talk that should already have been written, was in fact somehow lost, most likely in a checkout expunging at some point.
The result is around 8 hours of sleep in 4 days, mostly spent either writing talks, or using precious high-bandwidth face time with a plethora of interesting and important Perl and CPAN people I only get to see in person (on average) 2 or 3 times a decade.
The trip is certainly living up to "adventure" so far, so I shall limit myself to providing the details of YAPC in a sort of point form, with items spread randomly over the 5 days.
1-2 job offers per day, but most half-hearted because the rate I'd need to compensate for ditching glorious Sydney and the awesome job offer waiting at home isn't feasible in this financial climate in the US. Unless something utterly incredible comes along, I doubt I'll have to consider not going back home.
Saw a grand total of 1 talk (by the Google guys). The Google speakers were polished but creepy, since I got the impression they weren't really here to talk technically or expected an audience who would answer their rhetorical questions. While interesting, most of their talk was superficial and contained points covered in better depth elsewhere (a number by me even). And the badly disguised acting of their tag-team speaking style just made the talk look rehearsed and more a marketting exercise than anything. They've probably done the identical talk a hundred times.
As usual everywhere around the world, the REAL Google people were quietly attending a few talks here and there, not engaging very much and not saying anything about themselves or their work, and then leaving early.
Did, however, see the lightning talks, which were ok, but two orders of magnitude less awesome than the excitement and insanity of the OSDC::AU lightning talks.
The exception, of course, was Ingy's strip tease during his leg of the "impromptu talks", for which Schwern and I quickly responded with (for me) the great American novelty of sticking $1 bills into the strippers underwear.
Unfortunately, that meant we were up near the stage in the wings and so saw behind the curtain... those with us needed some time to recover our sight.
My "Tiny" talk wasn't well attended, but then it was an ugly talk slot, competing with Ricardo Signes' amazing "Email Hates the Living". Lack of rehersal made the talk a bit too stilted in my opinion, and content was a little light for me... so I got the concept over but felt it lacked some depth. Feedback suggests an interesting talk for the audience, but somewhat lacking in wow factor. Must correct that for YAPC::EU.
Despite trying to write the entire "Strawberry Perl" talk in around 50 minutes... in the speaker slot between the Tiny and Strawberry slots (and failing to finish it) as a speaker I found the Strawberry Perl talk to be just outstanding.
There was a very full audience (I was in the second-largest of three rooms) and beyond just the size, the audience seemed to be bubbling with anticipation. For something that I made to my laptop would work, I remain amazed at the level of energy there is out there for Strawberry.
The talk was very well received despite the fact I basically freestyled the last half with no slides, the questions were so extensive I ran out of speaking time with half a dozen hands still raised. I got some offers of public IP virtual servers with legal Windows license I can use for testing, and a ton of great feedback with the most pressing need being more information on the website
That out the way, I had two days to prepare the final talk, late on the last day of the conference proper. With the use of an introduction video, I once again had the great joy of forcing the audience to close their laptops to avoid spoiling the darkness. The idea of airing the dirty laundry of the Perl community in a safe and controlled environment, while explaining the specific reasons behind WHY they failed, continues to generate a lot of buzz and lots of traffic over the IRC back-channel.
While the audience does sometimes sit there a little stunned with the occasional gasp, I consistently find that LOTS of people find the time to come up and thank me for these talks afterwards. People are clearly getting a lot from it.
The downside of the talks is not seeing anything. I'm looking through other people's Flikr feeds quite dissapointed with some of the stuff I missed. I can only hope that this time the video uploads actually happen properly.
Face time over the 5 days, however, appears to have generated hugely positive results, including:
Discovered from a corporate attendee (from a very very large bank) that took some of us out for great Indian food (and payed for it, w00t), that there is genuine demand for a $2000 corporate ticket for YAPC, because the cheap price actually makes it HARDER to get approval to go to YAPC. All they really want for the extra money is proper printed invoices etc to further ease the approval process.
Bribed Schwern into adding new_ok to Test::More (which I've been asking for for three years) by buying him a license for DEFCON. In my defense, I did have to spend copious time over the next day defending the addition of the function, and helping refine it's behaviour in corner cases.
Hung out with jcap, Ingy and Theory making extensive evil plans for the resurrection of OpenJSAN.org. I'll be doing some hacking to get it running again at my next stop in Philly.
Met some completely awesome POE guys and missed the entire dinner auction, bowling scheming up an idea for pulling parallel HTTP requests using POE under the covers, but without actually needing to be a POE application yourself, so you can do parellised fetching in a regular sequential program.
Once we finally get it released to CPAN, HTTP::Client::Parallel should let me do mirror detection and minicpan syncing massively faster.
Logged WAY too much time with Schwern gossiping about EVE Online...
But also spent some time discussing various Perl gossip, and generally syncing on changes to the QA and toolchain since Oslo.
Spent a great time with Brian, both before and after YAPC, generally getting up to date with the soap opera of the American Perl community
Dragged Andy Lester (which required no effort at all) back to the surprisingly good pizza of Lui Manadis (spelling) instead of having to choke down the horrid implementation of "ribs" at the speaker dinner.
Did, however, learn to play Flux at the speaker's dinner.
Picked up some clues that removing overloading from PPI (which is really just a convenience and serves no critically useful purpose) might reduce CPU usage by as much as 20%. Definitely plan to investigate this later.
Spent some great time talking to Cristoph Otto from Microsoft Open Source labs about what Microsoft might be able to do to help me out with Strawberry Perl (mostly it's force-multiplier stuff like licenses and VMs for CPAN authors to test with and access to expert advice on some particarly twisted issues like WiX).
Provided what I hope was valuable face time to 10-15 people about various specific behaviours or conceptual oddities of PPI or other stuff I look after. In particular, spent a great deal of time with Elliot Shank talking about how we can start moving some functionality from Perl::Critic that shouldn't really be there down into PPI itself.
Gave out 5-10 new commit bits to my repository.
Helped nudge hachi towards implementing the first statistics run ever over the Strawberry Perl CPAN mirror logs. Hacked late into the night, and into the morning, on getting it working. It's also the first ever real program to use ORLite, and so I discovered a few interesting subleties in the process.
For example, ORLite is SO good at generating the ORM, you really have no idea what the hell it does in the rare case it doesn't do the obvious
Also started experiementing with adding in Debian popularity_context data as well, but then the sun was coming up, so we went to bed.
Quiet dinner with the other International visitors on the last day of the hackathon, turned into an amazing photography session with Strawberries (real ones) and flash drives in preparation for the new Strawberry Perl front page when Perl-on-a-Stick is ready for beta. This was followed by an unrelated propoganda filming session for YAPC::EU to help Jose.
Discovered the worst ever example of something ostensibly named "Felaffel" at the conference dinner. It had the consistency of sand and tasted like Sambucca...
There's probably more that happened, but to be honest the lack of sleep and moderate levels of alchohol makes it all a little fuzzy. But clearly a ton of good stuff was resolved.
And finally, I've discovered my new favourite introduction for talks...
"Hello. My name is Adam, and I'm from the future... about halfway up on the right."