browserd replacement
 

RE: http://lists.maemo.org/pipermail/mae...er/028697.html

I was wondering if anyone has any further news on this. I've done a quick search but can't seem to find any info.

The idea of updating the browserd gecko engine whilst retaining the current, closed source UI sounded quite appealing at the time. It could hopefully fix several rendering errors, increase HTML5 support (increasing important now adobe is finished with flash for mobiles) and, most importantly, close any security flaws that exist in the current version.

Re: browserd replacement
 

I've been looking over the patches for microb-engine on the garage page to see the extent of the modifications that are present vs. the original source tree and to compare them to Firefox 9.0.1. After going through some of them I found the actual source for the most recent microb package, but I haven't looked into that yet.

The good news is that although the line numbers are quite a way out on the code I have examined, the overall flow hasn't changed that much. In some cases, password manager for example, the code flow is actually more like that in the patches so they could be made simpler.

The bad. I don't have scratchbox installed for starters. I haven't got the time at the moment to get much further into this or the disk space on my pc. Not sure I fancy building it on my netbook either. Secondly, I don't know what effect the javascript changes will of had, could be able to drop the previous patches but some may need a rewrite.

Some patches are already updated (https://bugzilla.mozilla.org/show_bu...id=422540#c226 updated GStreamer patch from Maemo6 queue)

Even as a straight forward recompile of existing source, there would be the option of removing the following when thimb2 fix is in CSSU:to improve memory handling.


Finally, the requirements for building a stock version of Firefox/xulrunner: