On 14 August, 2007, the Israeli Perl Mongers held their regular monthly meeting. The program:
- 18:30-19:30 -- Gabor Szabo - How Test Automation is Done in Various Open-Source Projects
- 19:30-20:30 -- Yuval Yaari - Regular Expressions "Internals" (2nd Part)
Location: F5 Networks' offices at Ramat HaXayal in 24 HaBarzel St., Tel Aviv
- Ran Eilam
- Liror Rein
- Galel Mekori
- Eden Unicorn
- Michael Railburd
- Shlomi Fish
- Micha Nasriachi
- Gabor Szabo
- Zohar Snir
- Sagiv Barhoom
- Kfir Lavi
- Uri Bruck
- Shmuel Fomberg
- Pinkhas Nisanov
- Thomas Maier
Sagiv gave me a lift to the meeting. He invited his cousin who came with his friend, and we met them outside. Uri Bruck also came there later on. We talked about the fact Actcom (an Israeli ISP) was bought by Bezeq Bein-leumi and what it would mean for the Linux support.
We then entered F5, reserved our seats in the presentations' room, and went to the kitchen to drink and munch on the cookies. More people came, and we had some nice socialising.
We waited for Gabor to arrive, and he eventually did, he set up his laptop, connected it to the video output (worked so-so), and started presenting. Gabor prepared the slides as yet-unpublished pages on his blog, and used them as a reference. He said he is going to give the presentation on the upcoming YAPC::EU in Vienna, but that he doesn't expect a large audience, because he was scheduled simulatenously to Mark Jason Dominus.
In any case, Gabor talked about how various open source projects deal with running automated test periodically and with Quality Assurance. He covered perl 5, Parrot, pugs, PostgreSQL, the Linux kernel and (briefly) Mozilla and Subversion. It was interesting, but felt a bit unfocused to me, and the ad-hoc-ish slides he prepared did not help much. Someone then suggested he'll cover the various philosophies instead of going over what each project does, but I prefer Gabor's original method.
Then Yuval Ya'ari gave the second part of his presentation about regular expressions in Perl. He went over the various nodes of the regex engine, and covered the "trie" (pronounced "tree") data structure used to optimise some alternation expressions. He then demonstrated the "re" module with its flags, that can be used to debug regular expressions and understand how Perl compiled them. It was very interesting and engaging.
Then, many of us (nine, in fact) went to the Irish Pub after a long time where we were too lazy to do so. It was already quite late and we stayed in the pub until 23:00. There we discussed whether programmers who are good in one technology will write good code in other technologies, on version control, on whether it is OK to write quick and dirty code for an urgent task, and then to either throw it away, or refactor or rewrite it later on.
This review was written by Shlomi Fish.