Re: [Gnash-dev] FLTK2 (was: Removal candidates)

[Rob Savoye]

On Mon, Mar 21, 2011 at 01:24:03PM +0100, address@hidden wrote:

> Having said that, I don't see much point in keeping with such a niche
> toolkit, as there are much better supported choices for portable C++
> toolkits -- such as Qt...
 
  Obviously not an embedded developer. :-) FLTK2 was for small embedded
platforms to replace SDL. QT is much larger, and not always appropriate.
Yes, I know Nokia uses QT on phones, but I'm talking about devices that
have even less resources. But anyway, if it'll never get released, the FLTK2
code useless. Too bad, FLTK2 is much nicer than FLTK1.

> Most projects suffer from nobody *wanting* to do refactoring. Here we
> have people who volunteer for this -- and you try to discourage them?

  We have the problem of doing more refactoring than moving Gnash forward.
While I do agree refactoring is a good thing much of the time, so is not
letting Gnash stagnate.

> That's probably true -- but it doesn't work like that in a volunteer
> project. People work on what they are interested in themselfs. Trying to
> tell volunteers what they should work on, will just make sure they loose
> interest in working on the project at all.
 
  I think I know about volunteer projects, having worked on them for several
decades. As the Gnash maintainer my job is to attempt to coordinate
things. Telling any Gnash dev to do anything usually gets me called a
dictator though... I'm not telling anyone to do or not do anything,
but I am making a strong suggestion. A true dictator *would* tell people what
to do with no discussion allowed. Instead I do let developers mostly do what
they want, which is partially why Gnash has stagnated.

  If people aren't interested in moving Gnash forward, they should find
another project.. Instead some people would rather endless refactor
the same code till it achieves some preverse type of "perfection". As
I said, if one can't handle large code bases, they need a new career,
like flipping burgers... 

> But I'm glad to hear that you want to work on AVM2 next... ;-)
 
  It may be that after OpenVG is done AVM2 support is probably a good
task. It won't happen without funding though, I gotta pay bills like
everyone else... Without AVM2 support, Gnash will become useless. When
we have little resources we should focus on keepin Gnash relevant, not
making changes nobody will notice but a few developers. If we don't start
on AVM2 support soon, we might as well give up. Less and less sites work
with Gnash everyday...

> I'm not sure why either is necessary... I'm pretty sure GTK and Qt both
> cover everything that FLTK would; and together with FB, these make SDL
> redundant as well for all I can tell.
 
 GTK is large with many dependencies, which make it completely inappropriate
for small embedded systems. QT is large as well. Think small...

> Admittedly, Gnash is not in the same boat as X; which has tons of
> hilariously crappy vendor code that has been added without question back

  Well, Gnash has crappy code leftover from the GameSWF days too...

        - rob -