Whoopsie, haven't blogged since OSCON, which was fun. I stayed around in San Francisco for a bit (wow, those hills are steep. also In-N-Out is tasty) and I'm now back in sunny London. I was actually interviewing for a few positions around San Francisco but my timing is off: USCIS have announced (PDF) that there are no more H1Bs (the "tech visa") left for the coming 14 months, which is sucky. Dear all countries of the world: people who come to work for you or visit you are good for your economy, please start thinking more long-term.
I quite liked OSCON. I enjoy meeting people and seeing what everyone is up to. Thesedays, I'm more interested in the enterprise track: more hits, more data. Things you should be using: MySQL (with InnoDB and constraints, replicate down for reads and partition data if you must), memcached (fast, simple caching is all you need, trust me), mogilefs (don't spend $$$ or implement your own distributed file system or I'll hit you), Perl (well, CPAN really - dependencies are your friend), apache (nothing better yet), xen (I hate hardware, so virtualise it all so you can have no scheduled downtime), svk (for development - even if you don't need the neat distributed stuff, it's faster and uses less space than the svn client). Got any tips?
I'm really looking forward to YAPC::Europe in Portugal in a week and a bit (arrgggh, slides!) and then Euroscon.
The Summer of Code is coming to a close (note to Google: having that not updated isn't very professional). I wonder how the projects are going, and how organisations/mentors get to decide if the project is finished (and the student gets their $$$$).
Also, I'm sick of people writing frameworks. Something I stressed in my interviews is that I'm all about getting things actually done. Spending six months writing a framework with no idea of what it'll actually be doing isn't my cup of tea: I'll do the project instead and ship it. Only if I do three or more things which share some similar code will I actually consider writing a framework. Frameworks all the way down is lunacy, and I recognise it in too many companies. Get it done instead! ;-)
So back to square one. Anyone want a cool Perl hacker in London?
Also, I'm sick of people writing frameworks. Something I stressed in my interviews is that I'm all about getting things actually done. Spending six months writing a framework with no idea of what it'll actually be doing isn't my cup of tea: I'll do the project instead and ship it. Only if I do three or more things which share some similar code will I actually consider writing a framework. Frameworks all the way down is lunacy, and I recognise it in too many companies. Get it done instead!;-)
Amen. People building for reuse before there are a waste of space. Rip your frameworks out of your products, don't try and guess what you're going to be doing in six months time.
