Jörn Reder  wrote:
> Phil Ehrens wrote:
> 
> > Perl and Python are NOT platform independent at the same level.
> > Have you coded in Perl, Python, and Tcl for Windows, OSX, Linux,
> > and Solaris? I have. You will spend at LEAST three times as much
> > time doing this with Perl or Python as you would with Tcl/Tk.
> 
> I do this since 12 years with Perl on Solaris, Irix, Linux and Windows.
> Never did it with Tcl/Tk, because I'am happy how good it works with Perl
> ;)

Because of where I work, I'm not free to become attached to any
one scripting language. I'm *supposed* to consider long-term
support over the next 30 years for whatever I develop. 12 years
ago I thought Perl was the only way to go ;^)

> > It should have no priority at all, but I thought that Transcode
> > supported Windows. My bad.
> 
> You really thought transcode supports Windows? I am stunned.

Sorry, even though I maintain the wiki, I only ever really cared
about Transcode's ability to help me make dvd's, and never paid
much attention to it otherwise. Transcode has done SO much for me
that I feel obliged to maintain the wiki for as long as it's
considered useful.

> > I'm not kidding. I have no idea why you find this so unbelievable.
> 
> Because _Internet technology_ for video editing / transcoding purposes 
> is just not the best technology you can choose (of course anything is 
> _possible_ ...). I did many HTML/JavaScript/Ajax coding in my life and I
> never would try to cover a video editing solution with that technology.
> It's simply not appropriate for that and bloats things up tremendously 
> (if it's not just a simple HTML form, which just generates one simple 
> command line and just spits out transcode's output... but I think that's
> not the idea of an "offical GUI" we're originally talking about 
> here...).

Video EDITING?! Does Transcode support editing?! Jeez, I really am out
of the loop then.

> > All the gui needs to be able to do is take user input, build
> > command lines, and hook output from Transcode, right?
> 
> That would be the minimum. More than that for example would be realtime
> previewing which needs some more effort than "just" wrapping command 
> lines. As well it sounds very easy: user input -> build cmdline -> grab
> transcode output. Building an intuitive GUI which supports the user in 
> his _goals_ (not just enumerating command line options), is much more 
> difficult. This challenge is of course mostly independent from the 
> toolkit used.

Hence my suggestions, which were made in the spirit of minimising the
amount of toolkit related hacking that the GUI author would have to
deal with.

Reply via email to