Re: Pango branched

[Michael Meeks <michael ximian com>]

Hi Daniel,

On Mon, 2002-04-15 at 12:10, Daniel Veillard wrote:
>  If you can come up with a clear an concrete case where/why this is
> needed, why not.

	Ok; well - I don't have any great insights here, but I imagine that we
need to be able to do quite rigid source control on the libxml that we
call the "ultra stable Gnome 2.0 version", whereas inside libxml I
imagine you want the freedom to re-hash your algorithms, etc. to your
hearts content ( while keeping the same ABI ) ?

	Also, in many cases it's extremely useful to be able to see what
actually changed between versions to be able to binary chop bugs out,
and that can be considerably obscured by lots of other improvements in
the tree.

> But unless absolutely necessary I prefer not having
> to handle this, I honnestly don't see the point, though I see the
> associated cost.

	As above really. To get any vestige of stability, and make hard core
testing worthwhile we need a branch with fascist commit rules; patch
review, a 'minimal changes' regimen etc. and this is typically quite
antithetical to 'getting it right' and producing a maintainable codebase
for the future. 

	But it depends how stable, and perfect libxml2 is :-) clearly if it's
not going to need any work / internal re-organisation / re-writing, then
there's simply no problem, and need to branch.

	I think the most important thing is then working out what regimen we
think is reasonable to ensure ABI / API compatibility, minimal changes
etc. during the lifetime of Gnome 2.0 - and getting people to sign up to
it - rather than trying to make maintainers branch when they don't want
to.

	Regards,

		Michael.

-- 
 mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot