Hah, you wish! These blog titles are getting way ahead of the progress I make with WebPositive. Or let’s say the title is truthful in some ways, but on the other hand perhaps suggesting more substantial progress than what was made. I did turn my attention to fixing a lot of little annoyances and bugs that were reported via various channels, the comments section of this series of blog entries being among the important sources of feedback. So keep the good feedback comming, it’s very useful for me! Before I recap what stuff got done, let me quickly say that I didn’t tackle most of the bigger TODO list bullet points I keep mentioning. I did start with getting affine transformation support into the BView graphics stack features, but BAffineTransform, which I reimplemented from it’s previous incarnation in order to get rid of introducing Anti-Grain Geometry classes into a public header, is not yet used anywhere. Also I didn’t upgrade the WebKit SVN revision that the port is based on. That’s in part due to some emails on the webkit-dev list which suggested stability may have seen some regressions, at least in intermediate revisions. I shall review the WebKit change log and make an informed decision about upgrading, otherwise stability could be worse when I just use any random version. Also I did not yet investigate into using the WebCore internal allocator in order to perhaps get rid of some memory misalignment related crashing bugs. No clipping paths, no alpha masks, no advanced compositing modes.
This is a companion discussion topic for the original entry at https://www.haiku-os.org/blog/stippi/2010-04-03_webpositive_gets_polishing/