Lies, Damned Lies and Benchmarks
Benchmarks like statistics can be incredibly misleading in ways that
are only obvious with detailed analysis. Recently the apache
HTTPCore
project released some benchmark results whose headline results read
as having better performance than jetty in 3 out of 4 scenarios, and that jetty NIO sucked! So is HttpCore really an faster
than Jetty and does Jetty NIO suck?
[Read More]
Posted at 11:19AM Jun 18, 2010 by gregw in General | Comments[6]
Combining Javascript and CSS content
At the Rich Web Experience conference this week, there were many talks about performance and tools like ySlow. Everybody acknowledges that downloading lots of javascript and css with multiple script or link tags is a key reason that Ajax sites are slow to start, due to the latency of multiple round trips over the network. Yet almost every example I saw at the show had multiple script tags and css links.[Read More]
Posted at 10:02PM Sep 09, 2007 by gregw in General | Comments[9]
