Jonas Sicking wrote:
One argument is that it's simply impossible to work around an XHR
implementation that changes the casing in a way that the server doesn't
expect. For example if the server wants a 'doit' method and the XHR
implementation case folds to uppercase, the script author will simply
not be able to use that method.
So in this case I think both mozilla and opera does the wrong thing.
And here we run into the issue I keep meaning to bring up.... The HTTP
implementation may not be under the control of the XHR implementation.
Certainly in Mozilla all XHR does is pass the method on as-is to the networking
library; it has no control over what the networking library does with it. Now
it so happens that the networking library in question is also in mozilla.org
cvs, so it _might_ be possible to fix it (though it may also not be; the APIs
involved are frozen and have plenty of other consumers). But there's no
guarantee that this is the case for XHR implementations in general.
-Boris