$3800 for the winner

[Dietmar]

What about changing the rules and offering the bounty to an individual porting the KHTML rendering engine (in form of a shared library)? That's doable and would have much nicer effects for Amiga and Pegasos than porting a monster application.

[Dietmar]

>Fact is, the gecko engine, separated out from the mozilla app...which btw, mozilla is eol already....but the gecko engine is quite portable and a very competent rendering engine...

From what I've heard (and I don't claim to be an expert), gecko is better as in "compatible with more pages" but otherwise a real mess of code, even mocked by its own maintainers. Apple is reported to have choosen khtml for that reason. I've had a quick look at the kthml source code.  It's surprisingly small: download webcore from Apple and you will find khtml nicely contained in a 3.5MB folder ([edit:] plus 1 MB to detach from qt and use whatever Apple's OS is using).  I have not bothered to download gecko but found a much larger figure via google (7 MB?).

[Dietmar]

>Apple has chosen it, and I looked at webcore, and its great for Mac OS X, I think I'd start with the AtheOS port, rather than apples.

I'm not so sure. a) the AtheOS/syllable port is an older port b) I haven't heard about AtheOS before you mentioned it but after looking at its web page, it appears to be a Linux clone with a lot of Linux stuff already ported, including a minimal qt. Since qt is the gadget toolkit used by khtml, the AtheOS khtml port might still be based on that (just a guess) which would have made porting significantly easier. Apple, on the other hand,  already has done the untie-khtml-from-qt-and-linux job. One would basically be following their tracks, look at the changes they made, and substitute Amiga APIs. Apple's changes are commented in the source code.

[Dietmar]

>Two reasons, A: Mozilla uses the Gecko rendering engine, its not like you can just plug in another rendering engine and recompile the source.

Once a html engine has been ported, developing a browser around it is comparatively simple. In fact I think it would be better to reward the port of an engine rather than a browser because that would make sure that the engine is ported with re-use in mind (as a "component") and not embedded into/tied to a larger project. That would kickstart browser development.