Re: Module semi-proposal: gnome-shell

[Emmanuele Bassi <ebassi gmail com>]

On Thu, 2009-11-05 at 06:29 -0800, Sandy Armstrong wrote:
> On Thu, Nov 5, 2009 at 4:45 AM, Emmanuele Bassi <ebassi gmail com> wrote:
> > all in all, this whole digression on the scripting language smells an
> > awful lot of bikeshedding to me. if the Shell had been written in Python
> > we would have had the same arguments on d-d-l - emails of pure vapor and
> > a Shell that trawled the disk loading hundreds of unused modules on
> > startup, taking 60 seconds just to load.
> >
> > the Shell has *already* been written in JS. unless you (and by "you" I
> > mean anyone, not specifically you Jamie) are willing to port it all to
> > Python or C# or Genie or COBOL (to which I suggest to JFDI) discussing
> > the language of choice is pure bikeshedding.
> 
> Well I don't think that's necessarily true.  If gnome-shell is
> accepted, it implies that extending the shell (probably a more common
> activity than extending other GNOME desktop components) will require
> writing javascript, which is kind of new for this developer community.

extending the panel through applets in Python, Perl, C#, C++,
${WHATEVER} was equally true when the bindings for libpanel-applets were
written, and this enabled a lot of people contributing to the project.

so, though not entirely tangential, the language used is a non-issue;
JavaScript is already (through Seed) part of the blessed language
bindings for GNOME; this makes JavaScript a blessed language for writing
GNOME applications and proposing them for inclusion in the Desktop
suite.

I would have understood the objections or the questions raised in this
thread if JavaScript had never been used; or if it hadn't been already
present in the approved language bindings; or if the Shell had been
written in an obscure dialect of LOLCODE on a fork of the Parrot Virtual
Machine ported to brainfuck.

> I actually think introducing javascript is a great idea (I agreed with
> a lot of the logic behind creating Pyro), and that the choice of
> language should be taken as an argument in favor of accepting
> gnome-shell (based on all the reasons you, Owen, and others have
> brought up).

more than the actual language or VM chosen, by and large, is the use of
the GNOME platform through introspection that makes the Shell as it is
today a great project, in my opinion.

could have been Lua, as far as I'm concerned - or Io. hell, even Scheme
(which would have closed the circle with Miguel's first GNOME
announcement). :-)

ciao,
 Emmanuele.