Is there any guidance on minimizing the time is takes a load a caja guest
page? The attached files demonstrate a little test case I'm using to
measure the time it takes to load and run a trivial caja guest page.
Please let me know if there are any major issues with this methodology.
Here are samples of its output across three different browsers (all on a
2.2 GHz i7 MacBook Pro with 8GB of RAM):
Chrome 23.0.1271.64:
Total Time: 964ms (cold start in a new browser)
Total Time: 740ms (after a few reloads in the same browser tab)
Firefox 16.0.2:
Total Time: 740ms (cold start in a new browser)
Total Time: 762ms (after a few reloads in the same browser tab)
Safari 6.0.2:
Total Time: 502ms (cold start in a new browser)
Total Time: 443ms (after a few reloads in the same browser tab)
I certainly wouldn't claim that this a valid full-blown benchmark. But it
does seem to give a general indication of the startup overhead for a caja
guest on various browsers.
Are there any general practices that can be used to minimize this overhead?
Has anyone done more extensive tests along these lines? And if so, did
those tests produce similar results (such as the significantly better guest
page load times on Safari versus Chrome or Firefox)?
Thanks,
Jamey
Title: Caja host page
Caja host page
Title: Timing Guest Page
Total Time:
ms
