Re: Gtk# in 2.16

[Murray Cumming <murrayc murrayc com>]

On Tue, 2006-04-25 at 07:16 +0200, Murray Cumming wrote:
> > Technically, gnomeprint is a show-stopper for us.  We expose its API in
> > gnome-sharp.dll and therefore could not split it out and still maintain
> > our API stability guarantees.

I think, if there's absolutely no way to fix this, then it should be
allowed to pass, as long as the level of API stability of that part of
Gtk# is very clearly documented (I don't mean just in a README
somewhere). Just saying that it's OK because GtkHtml has been API stable
enough for you so far isn't enough, and for many developers it won't be
enough to know that Gtk#'s GtkHtml API was stable so far.

For instance, it's easy to know what API stability the GNOME Platform
APIs offer because 
a) They are in the GNOME Platform.
b) The API/ABI rules for the GNOME Platform are online.

But, for instance, if you can even find any documentation about the
API/ABI promise of GtkHtml then I don't think it's anywhere obvious. I
don't think anybody wanted to make that promise, so we shouldn't pretend
that they have. 


Of course I am annoyed that you've known about this requirement for so
long and have let it become a problem instead of fixing it when you
could.

If you think this documentation requirement is excessively demanding,
take a look at Brian Cameron's emails about interface stability.

-- 
Murray Cumming
murrayc murrayc com
www.murrayc.com
www.openismus.com