Comments on ..: hannosch :..: Plone trunk continues to get faster

Very impressive. I do hope you'll think about bac...

2009-03-02T02:02:00.000+01:00

Very impressive. I do hope you'll think about backporting what can be easily backported to 3.x. I know it will make your graphs less pretty, but it will also be really nice to be able to say "Plone 3.4: now 2x faster!" :-)

Tom: The language selector change in itself probab...

2009-03-01T22:00:00.000+01:00

Tom: The language selector change in itself probably has an impact of half a millisecond. It was more an example for what kind of problems our former "template-script" soup brought us. Count a dozen of those mistakes and you have a noticeable difference, though.

Laurence: All these numbers are ms/request, just turned into request/s for the presentation. 1000 divided by number isn't too hard to calculate, should you be interested.

P.S. I just made a couple of more intrusive experiments and I think I know, how to get the next 5ms or so down (yes, that means more than 50req/s). CMF Expressions apparently really suck performance-wise.

I think ms/request is a more useful measure than r...

2009-03-01T21:10:00.000+01:00

I think ms/request is a more useful measure than request/s

awesome! but the given example of the language por...

2009-03-01T19:25:00.000+01:00

awesome! but the given example of the language portlet should be easily be backportable to 3.x, no?

of course, I doubt, that that alone would generate such impressive graphs such as yours here :)