I figure what prevented me from building the ffdb was the need to get the data to play with, so I've decided to crawl archives to get the raw data so I'll have something to show real people I approach in soliciting help. Sounds like a job for Perl and regular expressions, right? I've read my jwz, so I know better. Also, perl is ugly. Python and BeautifulSoup, on the other hand, is… amazing is the only word that seems fitting. I've crawled FFA, and have done great progress in processing and normalizing the data. Of course, my data is the metadata, thus my metadata is the mata-meta-data? Harry Potter is character #1, anyway. LOL.
The interpreter-driven edit-compile-test loop doesn't scale to multi-class modules, and decent ORM is still hard. I have not even begun to contemplate user interface implementation (typical me).
A coworker asked me today how does a jpeg gets sent in an email, since SMTP is a text based protocol. I said: "enlarged by 33.3% via base64 and with the correct mime type, as specified somewhere under rfc822", and he thought I was totally geeking out. I've only remembered the number because I've (re-)read about it recently (in relation to parsing dates), but simply knowing about mime and that there's an rfc is like… in the job description, for a web apps team. I mean, the world doesn't start and end in the JVM, for crying out loud!
I'm way too judgmental of people.
No comments:
Post a Comment