Re: Will tree-ssa be GCC 3.5?

[Gabriel Dos Reis <gdr at integrable-solutions dot net>]

"Zack Weinberg" <zack@codesourcery.com> writes:

[...]

| supported.  That would be sufficient disruption to the way our users
| do things, to justify calling it 4.0.
| 
| But tree-ssa isn't that.  tree-ssa is an internal implementation
| change. 

Still, it is an huge instructure reorginazation.  I go not think that
only external behaviour calls for major version bumping.

There are many developers out there who base their toolchains on
official FSF trees.  Those people do see and are affected by major
reorganization.  I don't think we could just hand wave, under the name
that they are developers. I believe that a major reorganization
indicated by a major version bumping is A Good Thing to do.  We've
done that in the past.  I see nearly no reason not to do that again
for tree-ssa merge.

[...]

| 4 is of interest only to developers; that's not a reason to change
| the major version number, which is an advertisement to end users.

A major version number is *not* just an advertisement to end users.
And in a sense, outside developers that base thei products on official
FSF GCC sources are also end users.  I do not believe that anything
that affect "only" developers is not important.  They are just as
important as things that affect end users.

[...]

| > If we maintain gcc3 in parallel with gcc4 for a while, then this
| > gives us a chance to add in the tree-ssa stuff now even if it isn't
| > perfect yet.  People can continue to use gcc3 if necessary until
| > gcc4 stabilizes.
| 
| How is that different from continuing to maintain the 3.4.x patch
| series until 3.5 stabilizes?  It's just a matter of labeling. 

Yet, it matters.  Doesn't it?

-- Gaby