Work sucks! I spent Wednesday afternoon in phone calls back in the hotel room, returning for Damian's Perl 6 talk. I enjoyed the trip to get buffalo wings, but bailed on the movie to write my lightning talk. Wednesday I was nervously rehearsing, but enjoyed a quiet lunch with Allison. The talk went well (as did Chip's Perl-as-religion talk), and then it was off to the auction. I was so knackered from my late Wednesday that I had to slip off from the auction and sleep again. I'm getting weak in my old age.
Whose talks did I enjoy? Damian is, as always great. Steve Hayman from Apple was as entertaining as Damian, and I don't say that lightly. Daniel Allen's style guide talk had a message that everyone should be exposed to. The two talks on test-driven development (by Ted Karitonov and Andy Lester) yet again made me think I should be doing it.
I think the world needs a set of kata, small tasks that you can write tests for first. Then you can see how an expert wrote tests, and compare. Everyone works on different projects, so it's hard when people like Schwern and Andy say "just start testing your code!" to know what aspects of your code to start with. Baby steps, training wheels ... that'd be good. You listening, chromatic? :-)
Anyway, Abigail's about to start talking about, I don't know, regular expressions on crack, which seems to have been his theme throughout the whole of his speaking career. I think I missed the talk where he proposed solving all NP-hard problems with regular expressions, but I don't want to mis this one. Until later ...
--Nat
01compile.t, 02import.t, 03constructor.t, and so on. the idea is that every module should be testing pretty much the same thing at the outset. and even if there is no need for an import() test, I expect the file to exist and print a skip message, just so everything is uniform to a certain part. I have personally found that dictating a few simple things like that help to get people to start thinking in "test."
Good idea. I'll take it to the QA list and we'll figure out something.
Re:Baby steps
2shortplanks on 2004-06-19T22:34:40
That's a damn good idea. Having a group 'adopt' a module too might be a good idea too. So there's more focus and people have a smaller goal they can work on than improving the whole thing.I often think we should use the power of London.pm for good as well as evil. I should take this to the QA list as well I guess...which I *really* should start reading more often (bad me.)
Re:Baby steps
petdance on 2004-06-21T03:39:36
London.pm is part of what inspires me so. You guys were such a presence at OSCON last year.
Re:Style Guides for Large Projects
gnat on 2004-06-23T12:28:52
There were a lot of #perlers whining "we already KNOW this" during your talk, which I took as a thumbs-up for the content and a thumbs-down for their ability to read talk descriptions:-) --Nat
Re:Style Guides for Large Projects
vsergu on 2004-07-06T20:19:46
You're neglecting the possibility that they read the description and went to the talk specifically to whine.Re:Style Guides for Large Projects
jmm on 2004-07-06T20:49:11
We stated the "already know this" bit during the Toronto PM meeting that recapped YAPC, but not as a whine. It was just that since Daniel had given the talk to tpm not too long ago, we didn't need to review it in detail for the people who didn't go to YAPC. The fact that a number of us went to the YAPC talk after having heard it before in Toronto could be taken as a thumbs-up for the content too.