The Second Annual YAPC

Day 1 | Day 2 | Day 3 | Wrap-up

Back in Cambridge after three days of Perl talk, I took some time to put together my overall impressions of Yet Another Perl Conference (YAPC). These topics are matters I wasn't able to squeeze into my daily reports.

After my impressions, I've included some impressions of the last day of the conference from my O'Reilly associate, Joe Johnston.

At the end of this article, I will conclude with a few Pittsburgh social notes.

One of the recurring issues of this year's conference was the question of how Perl users treat others, including other Perl users. For example, a number of people were concerned that new Perl users were treated harshly when they asked questions in certain newsgroups and mailing lists. There are two sides to this issue, however. It is sometimes the case that newbies ask questions in forums reserved for important Perl development work, like When they do so, the folks trying to get work done in that newsgroup can be brusque or even a little abusive. There are times, also, when subscribers to comp.lang.perl.misc can be unpleasant, too, although that kind of response is less justified on that newsgroup. (Jon Orwant tells me that a recent Perl Journal article said that over 96% of questions to comp.lang.perl.misc get answered, so it is worth a try, if you have a thick skin.)

There are some safe places for newbies to get answers to their questions, however. Verbal abuse is expressly forbidden in, and is also a good place to try. And a new site,, a sponsor of YAPC, is also dedicated to helping new users get up to speed on Perl. Try them out.

Another topic of concern had to do with the occasional flame wars with Python adherents. Nat Torkington gave a rioutous Lightning Talk that attacked Python, but it was all in good fun. He said in another of his talks (entitled "Be an Advocate, not an A***ole") that, "Python is not the enemy; Microsoft is." The first part of that statement is certainly true. The number of users of both Perl and Python is growing, so whatever rivalry exists has had a beneficial effect on both communities. The uses of Perl and Python overlap, but the environment needed to make both of them flourish is identical. There are differences, certainly, that let serious programmers choose between them; but one's success doesn't hurt the other.

Another nice feature of this conference was the presence of talks demonstrating how people have successfully used Perl in practical applications. Steve Jenkins, for example, demonstrated how he used Perl in conjunction with a number of other languages in a highly diverse environment (NT, Unix, QNX...) to provide dynamic data display from wind tunnel tests. Steve's presentation was especially interesting because he chose to use the Web and standard Web browsers for the data transmission and display. His presentation was an example of the benefits of the increased speed of the Web. Although the Web is certainly not fast enough for the display of many real-time projects, this project, which required updates every few seconds, could use the Web. So, rather than having to program in C and write his own display code, Steve used Web standards and Perl modules, reducing his development time and maintenance while giving his users a familiar interface.

Andy Murren demonstrated how Oven Digital is using Perl with XML and WAP to deliver email and other messages to cell phones. Andy used CGI::WAP, a rewrite of CGI::WAP, not yet in CPAN, makes the modifications to (like the use of lowercase tags, essential to WML and WAP) to make its power applicable to the low bandwidth, low power, low computation environment of cell phones. (Andy promises the end to all the "e-" prefixes, too. He predicts that the new buzzwords will begin with "m-" for "mobile": m-commerce, m-mail, and so forth. You read it first here.)

As promised, here is what O'Reilly Perl programmer Joe Johnston had to say about the last day of YAPC:

As with any good meal, the best of YAPC was saved for last. The afternoon began with the Lightning Talks, a staccato series of very short lectures presented by various people. For many of the speakers, this was their first opportunity to present and some had difficulty finding the appropriate pace.

There can be little doubt that the highlight of this series was the two talks given by Nathan Torkington. The first of his rants provided a caricature of the Perl 5 Porters mailing list, which is devoted to ironing out the semantics of Perl's syntax and new features. Much of the crowd enjoyed his humorous portrayal of the chaotic beehive that produces such practical honey.

His second talk was a mock polemic scorning Perl's competitor Python. It was the worst form of advocacy, but a positive joy to hear. There is little doubt that Mr. Torkington has a future in entertainment should this computer thing peter out.

The day concluded with Jon Orwant's "Post Apocalyptic Perl, Episode 2", in which he denounced Object-Oriented programming as too often needlessly and wrongfully complex. He first related the workings of the human brain (!) and showed a model of how autism oversensitizes its host. This hypersensitivity often overwhelms the host. One way to compensate for this flood of sensations is to shut down as many sensations as possible and focus only on a very small set of stimuli.

Jumping to a seemingly unrelated topic, Orwant showed a series of questionable art objects and showed, while they clearly had a design behind them, that design wasn't necessarily good. This led to a tangent about copyright law. Design choices can be protected under copyright, but basic functionality cannot.

This segued into a discussion of pain and the hedonic calculus. How can we measure pain? Are shorter bursts of intense pain perceived to be worse than longer periods of less pain? Fortunately, an MIT professor received a grant to cause pain to people to find out. At the junction of art, law, pain and autism lies Perl. Perl allows a programmer to move beyond the rote functional code of other languages to add a creative dimension to his source code. Perl code is artful, but often expressed in such a focused way that an observer might not understand the zen of the source. Finally, Perl was born out of the pain caused by using blunt tools. Perl helps programmers put an end to painful cross-application muteness.

I learned more neurology and quantum physics at YAPC than in four years of college. I might need a few university night classes before the next one!

Thank you, Joe; I applaud your effort to summarize Jon's talk, a task I shirked yesterday.

Finally, as a born-and-bred Pittsburgher, every trip back is lots of fun. Here are a few social tips:

That's it for the YAPC. See you next year, whenever, wherever.


Frank Willison

Visit for a complete list of O'Reilly's Perl books.

Return to: Frankly Speaking