Unix revivalist meeting

[sawyl]

So, today was the day of my conference presentation at the NEC User Group and despite all my nerves, it went well. Embarrassingly so, in fact.

Taking a lesson out of drspleen's book, I started off apologising for the fact that I hadn't given a presentation before and that I didn't think I had enough material for half an hour, to get everyone on side. I then gave my presentation proper, which I thought went reasonably well once I got into it - although I'm still a bit disappointed that no one jumped up and shouted, "Hallelujah" when I got into my spiel on the total greatness of Nagios. Somehow, I managed to finish dead on half past and knock all the audience questions on the head in time for Stephen to give a new version of his 4D VAR talk, which he'd previously presented at ECMWF.

I wasn't quite sure how it had all gone, I was pleased although it's always hard to tell how your thing, but quite a few people came upto me afterward and were really positive about it. One person said, "I didn't think a quite guy like you would be able to go on like that, but you started talking and it suddenly all came together didn't it? The ponytail, the gothic look, the whole Unix thing, it all worked..." Most worryingly of all, I had another conversation with the Master after God:

MaG:	That went well.
Me:	Really? I'm very glad you think so.
MaG:	Yes. A few people asked how many people were in your group.
Me:	[ slightly thrown by the non-sequitur ] Ah, I thought I mentioned that there were four people in our group at one point.
MaG:	No, they thought you were in charge. They thought you came across as a team leader. You're obviously someone who is going somewhere; someone who's star is rising...

It was all I could do not to run away screaming. There's no way I want the MaG to start meddling with my career, attempting to make me into a manager in his own image.

In other news from the NUG, there was a pretty interesting talk by Mathis Rosenhauer, who'd done a lot of super cool work on grovelling through SUX kernel data structures using C - pretty impressive given the abominable lack of documentation on system internals. The only way I was able to work out how to walk the same sort of structures with crash was by reading the header files and going through the crash source code - I gave up on C when I realised how retro the process table interface was (Ha! Interface! Like, as if).

I'm looking forward to a normal, albeit three day, week next week.