Back to newsletter 072 contents

We had some good entertainment this month, with Tim Bray and theServerSide making lots of noise but not really much light (about PHP vs Java scalability, see the news items below). In both cases they should really have known better - in Tim's case he should have known that his position means that anything that casts even a slightly negative light on Java will get blown out of proportion. TheServerSide probably has less blame - the odd sensational headline is difficult for a media site to turn down and, as long as its not the start of a trend, is fair enough. In all the noise, the fact that these were opinions and not measured analyses seems to have got lost.

A note from this newsletter's sponsor

As for Ruby on Rails (RoR), the third element in Tim Bray's comparison, well anyone who has spent a decade in IT has seen the RoR 4GL type of phenomena happen again and again. Someone comes up with a new way of doing development that seems way faster. Its taken up in a wave of enthusiasm. It turns out you can do the simple stuff in it, but as soon as you do anything slightly off the beaten path it becomes a world of pain (as apparently is the case with RoR). Disillusion sets in. After 5 years no one is developing in it, after 10 years projects have almost all been migrated off it. But at some point, one of these 4GLs will work for the non-simple cases too. Then someone's going to become richer than Bill Gates and our industry will (finally) move on. Meanwhile, Java is usually the best option for the job (my point of view, of course).

Now on to our newsletter. And we have Javva the Hutt continuing to tell us all about what he was doing while he was away. (For those of you who don't know Javva, or want a reminder, we've added a page for you to access all his old columns here.) And of course we have our other news, this months selected articles, tools, and our many new extracted performance tips . Out of the articles from this month, Tony Printezis's blog is essential reading if you are using JNI - and a little birdie tells me there will be quite an interesting JNI performance related analysis in the comments, so if its not there now, check back to Tony's blog in a couple of days.

A note from this newsletter's sponsor

News

Java performance tuning related news.

• Tim Bray actually says "After all, Java runs EBay while PHP runs Wikipedia and Yahoo! Finance. Both clearly scale well enough", but the ad hoc pictures (not based on any evidence) favour PHP
• TheServerSide decides to go for a sensationalist headline that doesn't reflect any actual evidence of any kind (though feasibly may reflect what Tim Bray said at a PHP conference)
• JSR 277 supports ahead-of-time compiled native code
• A Java game implementation, warts 'n all. I've included this here because he doesn't mention performance at all - for Java, performance is no longer an issue. Of course he has other niggling issues, but I suspect there would have been issues whatever platform he had used.
• clustering Lucene with Terracotta

Tools

Java performance tuning related tools.

• Faster Java Serialization
• Java2D Benchmark: J2DBench
• ProfilerJ profiling tool for Java 5.0 applications
• Visual Performance Analyzer (VPA) is an Eclipse-based performance visualization toolkit
• JChav is a way to see the change in performance of your web application over time, by running a benchmark test for each build you produce.

A note from this newsletter's sponsor

Articles

• The Garbage Collector is too aggressive!!! (Tony Printezis/Sun)
• Java2D Gradients Performance (Romain Guy/java.net)
• Ask the Experts on Swing (Scott Violet, Shannon Hickey, Chris Campbell, and Oleg Sukhodolsky/Sun)
• Scaling Enterprise Java on 64-bit Multi-Core X86-Based Servers (Michael Juntao Yuan, Dave Jaffe/OnJava)
• Are you really Multi-Core? (Heinz Kabutz/cretesoft)
• Real-Time Tracking and Tuning for Busy Tomcat Servers (Edmon Begoli/devX)
• Garbage collection tuning in Java 5.0 (Peter Mikhalenko/Builder.com)

Jack Shirazi

Back to newsletter 072 contents