Got this reply from Michael:Hi there!
> I'd like to ask how the Haiku Gnash port is proceeding? I'ts been discussed
> here:
http://www.haikuware.com/forum/bounty-discussion/
284-gnash-portWell. In fact I have completely given up on Gnash for some reasons. First of
all their source is a mess. It combines numerous different libraries/sources
that were slapped together so they more or less work. This makes it hard to
get into Gnash. Then the sources are in flux pretty much all the time, which
makes it hard to keep any porting effort reasonably up to date.
Then the people involved in Gnash did apparently (can't really understand why)
set their roadmap to subsequently use the boost templates and libraries
(http://www.boost.org/) to replace as much as possible of the stuff that was
previously done internally or using more portable means. While this might be
for the better in some cases, it brings a _huge_ dependency. A dependency that
BeOS cannot and probably will never be able to meet. This is simply due to
being GCC2 only and missing numerous POSIX APIs. The boost libraries are
pretty complex and it would be a hopeless effort to try to get an up to date
BeOS port done with GCC2. Haiku will be able to "host" boost at some point in
the future, but it will certainly be GCC4 only and it will require quite a lot
of work probably. I have in fact reimplemented quite a bit of the boost
functionality just to keep the port working, but it is a loosing battle and
not very satisfactory. Then at some point I just gave up on it after another
major source change in favor of boost.
> I'd consider starting a bounty if that would help and speed things up.
So you see, there are enough problems (starting from coding style / design
philosophy up to dependency problems) which are big enough to completely stall
the port. A bounty would therefore not really help. In fact I'd suggest not
betting on Gnash, but on another project instead. One that does not come with
such a hurried together design. Maybe looking at swfdec would be the better
option, but I have not looked into it more than checking out its dependencies
(there are a few of course, so it might be just as difficult).
If someone would really want to get going with Gnash, he/she should first set
up the dependencies. If this works out, I will gladly provide my native BeOS
audio/video/graphics backend (which has an interface that is now probably
completely outdated and therefore unusable as is anyway). But before getting
to that point, I don't really see the point in digging out the sources.
I'm sorry to say it that way, but this is pretty much how I feel about it.
Hope you (and others) can understand that. Feel free to share this info as
necessary.
Regards
Michael