Re: Thoughts on mm-common and gmmproc

[Murray Cumming <murrayc murrayc com>]

On Sat, 2010-01-09 at 21:04 +0100, Krzesimir Nowak wrote:
> On Thu, 2010-01-07 at 09:39 +0100, Murray Cumming wrote:
> > gmmproc's use of perl and m4 (and how those parts are interdependent)
> > makes it very hard to add new features or to fix problems.
> > 
> 
> How much extensibility is needed? I mean - should it be easy just to add
> new feature to gmmproc itself or should it be easy to make it extensible
> by maintainers of modules (by writing sort of plugins doing specific
> things these modules need)?

It should be simpler and easier to maintain than the current system.
There's no great need for any particular plugin system. A simple rewrite
is enough work - there's no need for feature creep.

> > A rewrite would be useful if it was easier to maintain. We have plenty
> > of existing .hg/.ccg code to check that a rewrite would still process
> > existing files correctly.
> > 
> 
> Files using m4 voodoo would need modification. Or not, if rewrite should
> strive for strict backwards compatibility.

The convert*.m4 files are not complicated. We could probably support
most of them. But I'd be OK with requiring a python syntax for them.

The main aim of supporting some backwards compatibility is to provide
structure to the rewrite process.

> > Anyone should feel free to play with ideas in a git branch, even on
> > gitorious or github if they don't have git.gnome.org write access.
> > 
> 
> Good to know, but something official could also be helpful - something
> like a place on wiki somewhere in gnome.org for brainstorming or to keep
> ideas from irc or this mailing list.

I don't have time for it myself. Feel free to add something to a gtkmm
sub-page in live.gnome.org


-- 
murrayc murrayc com
www.murrayc.com
www.openismus.com