That's very close to my notes, minor correction inline:

On Apr 13, 2010, at 11:17 PM, Adam Barth wrote:

I think Maciej took notes, but my recollection is as follows:

1) Convert as many of the unit tests to LayoutTests as possible.
2) Land GURL in svn.webkit.org as is.
3) Convert GURL to WebKit style.
4) Make an ifdef that lets ports switch between KURL and GURL.

Actually this ifdef already exists, so we don't need to add it. But I had a step 4 of:

4) Identify or create relevant performance metrics for URL speed and memory use.


Once we get to this point, it will be easy to evaluate the performance
and correctness impact of switching from KURL to GURL.  Ideally, we'd
have a "bake-off" and pick the best parser based on some objective
criterion.  If KURL wins the bake-off, we should consider refactoring
KURL's parser not to depend on WebCore::String (or really any of
WebCore).  If GURL wins the bake-off, we should document (either with
comments or tests) which parts of GURL's API is needed by Chromium but
not necessarily need by other ports.


Regards,
Maciej

_______________________________________________
webkit-dev mailing list
webkit-dev@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev

Reply via email to