[A bunch of the team met up yesterday to hammer out some decisions.] I think we decided to try to not port the views system, but instead borrow as much code as possible from on top of it.
Some considerations that I think played into it (I had argued for this one, so I apologize if I mischaracterize it): - the stuff I laid out in my previous mail in explaining option 2 - current views code is very Windows-specific, and by generalizing it to two platforms you make it less useful on each - concerns about whether our drawing model (e.g. render everything into skia and then push it to display) will work well on X - we either can have one UI layer and require feature authors to make their code work on two systems simultaneously, or we can have separate UI layers and require catch-up between systems The huge negative, of course, is the code duplication/siloization aspect of this. The Mac has demonstrated we can share all of the "model" classes so hopefully it won't be so bad. (Historical note: Mozilla had separate frontends in this same way and eventually merged them with XUL, which I hear came directly from the pain of doing it this way. It remains to be seen whether this or something like XUL is more painful.) -- Note that this *doesn't* mean we're going a separate path for the look of Chrome. I hope we will be able to share the majority of the tab strip drawing/animations/etc. code. --~--~---------~--~----~------------~-------~--~----~ Chromium Developers mailing list: [email protected] View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-dev -~----------~----~----~----~------~----~------~--~---
