Thanks for comments and pointing out potential issues. I am not an expert in Tapestry nor JSF and I did this benchmark because I wanted to chose one of them, thus I could've made very obvious mistakes. Looks like I need to rerun the Tap4 test with Hivemind 1.1.1. I hope this will improve Tap4 results.
On Tue, 14 Mar 2006 23:33:04 -0600 Robert Zeigler <[EMAIL PROTECTED]> wrote: > Interesting. > Some thoughts and/or questions: > 1) Why are the page classes for success and fail abstract? They can be > concrete. > 2) For testing 4.0 stuff, why not use the 4.0 dtd? > 3) You've implemented the home page name,password, etc. properties as > simple properties... that is, you're not allowing tapestry to do the > legwork for you. a) I hope you're not doing this in production b/c > you'll run into issues in terms of data leaks. b) for benchmarking, > too, you should do things the "tapestry way" since otherwise, the > times don't reflect "real" processing times. > 4) The simple example is nice for a variety of reasons, but where you > will (theoretically, at least :) see tapestry shine over JSF is with > complex component-tree rendering/handling. Why not create a very > complex form with multiple nested components, etc. and compare jsf to > tapestry? That's more "true to life" and illustrates performance in > more realistic situations. > > Robert > > Alexander Varakin wrote: > > Hi, > > > > I ran a simple Tapestry vs JSF benchmark, results are posted on my > > blog: > > http://www.resupedia.com/blojsom/blog/Java/2006/02/28/Tapestry-vs-JSF.html > > > > > > Alex > > > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: > > [EMAIL PROTECTED] > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
