
Moderatore: Newser
Fogg_80 ha scritto:Fate l'amore non fate la guerra![]()
![]()
![]()
![]()
.
Timberwolf: Progress on an important project
Note: Since this blog is a channel for OS4 developers to tell about their projects, I think it's a good idea to post something about Timberwolf here, even though it's not related to Hyperion Entertainment or a specific AmigaOS development.
I think it's safe to say that Timberwolf (i.e. the port of Firefox to AmigaOS) is an important project for AmigaOS 4, since today, a system that wants to be taken seriously will have to have Firefox in it's software repertoire. A lot of people argue that OWB fills in the gap of a modern web browser and think that we should rather concentrate on that and ditch Firefox. However, the Mozilla software is more than a web browser. It's technology. It offers a platform independent way of writing applications, and AmigaOS can very much profit from all the developments going on in the Mozilla community.
As a web browser, Firefox is cutting edge technology, offering things like html5 and WebGL, In fact, Firefox has become part of the definition of the World Wide Web, driving it's development. It's importance in that respect matches or surpasses any other browser. It is also still actively developed by a large group of individuals.
As a technology, Mozilla allows the creation of applications that run on a variety of platforms. One such application is, for example, Thunderbird, an email client that also became the de-facto standard on most systems.
So, what about Timberwolf ? After the release of the early alpha preview last year, we've been working on getting the source code updated to the new version of Firefox, 4.0. Early versions (up to beta 6) produced rendering errors within the box layout on Windows (I didn't test the Linux version at that time). Therefore, we decided to wait until a version came out that would at least render correctly (since that is basically what we would have to implement ourselves). When beta 7 came out, the layout seemed correct, so we started with the Firefox 4.0b7 source code and applied our code changes to it.
This weekend, I got 4.0b7 to fully compile, although the implementation of the rendering code is still a stub. Since the final release version of Firefox 4.0 came out just recently, we decided to try to update from 4.0b7 to 4.0 final. The update worked quite well, although a few kinks need to be ironed out.
This means we now have all the code that will make up the final release of Timberwolf. In the coming days (time permitting) we will start the implementation of the actual rendering engine. We're not able to give time estimates, but I would guess that we will be able to come up with the next preview version sometime in early summer.
For more information about Project Timberwolf (including screenshots and updates on the progress), visit the page on AmigaBounty.
I'm aware that we have been quiet lately, but don't worry, there's been work going on, and Timberwolf is making progress.
The 4.0 code base of Firefox now fully builds and a prototype implementation of the new rendering code was implemented, based on the old rendering code.
Instead of "fixing" the old code, though, we opted for a complete rewrite, for a couple of reasons. First of all, the old code became really bad with time (the usual thing that happens with experimental code). Secondly, reimplementing it all would get rid of any artifacts in the code that were still necessary in the 3.5 source tree. Basically, it's a clean start.
With the information gathered in the previous effort, it only took me some few hours to rewrite the skeleton for the widgets module. This does not include the rendering code yet. This will be up next, right after fixing a few outstanding bugs that cause crashes on exit.
In the process of porting the 4.0 base, we also ported a new version of sqlite. We'll release that in the near future (when I find the time, if you need it up front, drop me a mail).
I'm usually quite bad at predicting how long things will take (Actually, precisely "when it's done", of course), but I hope to have the rendering code in in the course of the next two to three weeks (depending on how much time I have for working on the project). We will release a new pre-release version as soon as we have something that can be remotely used.
We also wish to take the opportunity to thank all of those that supported us, and apologize for the long stretches of silence. We're still at work, and although progress is slow, we'll eventually get there.
andres ha scritto:A posteriori (ovvero dopo circa 2 anni?) forse questo porting si sta rivelando troppo complicato
andres ha scritto:probabilmente se avessero sviluppato un browser totalmente nuovo da zero, utilizzando il motore webkit, sarebbe pronto; certo non sarebbe ai livelli di FF (che solo per i plugins, è su un altro piano), però...
andres ha scritto:Oltretutto lo sviluppo di FF negli ultimi tempi va a una velocità decisamente maggiore del passato, si parla già di versioni 6 e 7, di una interfaccia completamente rivista...
samo79 ha scritto:
Cmq se fanno bene la 4.0 (stabile, veloce e amiga-like) ci campiamo secoli con questa
Visitano il forum: Nessuno e 7 ospiti