Hi Bryan, this all sounds pretty cool and I think it would be way more effective than a rendering engine switch, if we want to try one time again to take over the world ;)
Just a few comments: - I agree we should try this in an experimental branch and that it should possible to easily change/rewrite every part of the UI. Though basing it on pymozembed or similars could be too low level, a lot of needed features may not work properly, and that would stop people to try it out. I dont know much about python bindings works but... wouldnt be possible to just write a different EphyWindow in python (even by patching the code in a branch, I dont think that would be a big problem). - I think we should keep this easy to try out for people, even if experimental/incomplete code wise. For example I'd avoid to add dependencies like mono/lucene. IHMO better to use something simple even if it will not have a cool indexer. This is something we can deal later, when we found ideas that are worth spending a lot of time for a good implementation. - I must have said it a few times but... I think history/bookmarks are totally obsolete/unused tools. Replacing them in a mature project is going to cause disasters like we had for our somewhat different bookmarks system. But experimenting new ways of solving some practical problems, like those Bryan enumerated is going to be fun and I'd hope productive - A lot of work is gone in extensions and bindings lately. I think this would be a good way to use some of that stuff to actually change people browsing experience. - These sort of extensions, or whatever they end up to be, will interest a lot more people than, say, developer extensions. So they should bring more feedback, ideas and hopefully patches. Marco _______________________________________________ epiphany-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/epiphany-list
