I suppose your book explains which TLAs I should chant while coding.
]]>"If you have nothing constructive or kind to contribute to a discussion, please feel free not to comment."
]]>That is true. I often do this whenever i am in a situation where i can apply working like this.
> Tests are highly compatible with exploratory coding.
Not in all cases though. Example from my work: I have an XML web API (blegh), that is supposed to provide me with data about one-way flights. I have zero documentation or explanation of it however, as the thing is still being built on the other side. That means that i have no idea what individual data points can be extracted from it until i've written the code to request and parse apart the data.
I can apply some testing very early, in the form of regression testing, but TDD is extremely difficult in this situation, as it's basically a case of figuring out which comes first, chicken or egg?
]]>I recently wrote a FF automation script using MozRepl. It was challenging, fun and rewarding -- the source to WWW::Mechanize::Firefox was invaluable!
And by coincidence, I recently wrote an HTTP service request library using AnyEvent. There's a comment in one of the MozRepl modules about needing to rewrite Net::Telnet with AnyEvent, and I've thought about accepting that challenge.
Anyway, so I'm very curious about the AnyEvent and Coro bits of the presentation!
]]>