I spent a lot of time today trying to make a small, reproducible example that behaves differently on Mac/*nix versus Windows. But my problem is that in my large application in which the problem originally manifested, I'm using an awful lot of 3rd-party JavaScript code (jsMath). That code does tons of stuff that I'm not totally aware of, which may have a big impact on my ability to reduce the bug to something I can reproduce.
I tried to make a simple Qt app that just creates a QWebView and then in it runs some JavaScript code that tries to make an XMLHttpRequest to the local filesystem; this is what was behaving differently in my large application. However, I must be doing something wrong because now I can't get it to succeed on any platform. Perhaps jsMath was doing some helpful setup of the JavaScript engine/space that I don't know about that was enabling this to work on some platforms. I'm sorry that I don't have time to continue to plug away at this. I guess it could be a jsMath bug, because jsMath makes a lot of queries about what browser it's running in, to try to be cross-platform; it may not expect whatever answers it's getting from a QWebView, and thus be making bad decisions. Nathan Carter On Apr 26, 2010, at 3:01 AM, Benjamin Poulain wrote: > ext Carter, Nathan wrote: >> Just for closure's sake, I'll report that this invasive and inelegant >> solution solves my problem for now, but is definitely not preferable. >> If XMLHttpRequest began behaving the same on all platforms, that >> would be much better. Is this a known bug or should I try to create >> an example I can submit? > > It is definitively worth creating an example showing the problem of > XMLHttpRequest. > > I have added Markus in CC, he is the more able to solve networking issues. > > Benjamin _______________________________________________ webkit-qt mailing list [email protected] http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt
