Another good technique is to save the source of the page, and stick a base tag in the head to help you narrow it down.

<base href "http://www.blah.com">

Some other good tips can be found at http://webkit.opendarwin.org/quality/reduction.html

-Adele

On Jun 7, 2005, at 9:58 PM, Maciej Stachowiak wrote:


On Jun 7, 2005, at 6:39 PM, Steven Riggins wrote:


hi folks :)

I built today's webkit and was taking it for a spin.  I have one site (a dating site) in which I get replicable crashes in an unknown thread.

I saved an archive of the page, converted it to xml, but it is still unreadable as far as reducing goes.

Any tips on the best way to reduce a third-party's page?


One good way to get source to use as starting point on a bug reduction is to use the Firefox "save as complete HTML" feature.

Another option is to save off just the HTML file, and then use the Safari Activity window to find out and get at the other resources of the page, but this requires more manual steps.

Regards,
Maciej

PS Have you filed this bug in bugzilla yet? With a crashing bug, it's often possible to figure it out just by looking at the backtrace and reproducing the crash in the debugger, without the need for a reduction.


_______________________________________________
webkit-dev mailing list
webkit-dev@opendarwin.org
http://www.opendarwin.org/mailman/listinfo/webkit-dev