Re: Criteria for GCC 4.0

["Ian S. Nelson" <nelsonis at earthlink dot net>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

|- Some people consider x.0 releases as unstable and not usable.
|- Other consider that it is the first "usable" release (not
|  yet extremely stable, but enough to work for most people). Usually,
|  this perception happens with products that had a long x.99 or beta
|  history (but note that despite several test-release candidates,
|  no one trusted the 2.6.0 linux kernel to base a distribution upon
|  it).
|
I'm just a lurker here mostly... but I want to chime in.

This is a perception perception issue. 2.6.0 was a very good release,
there weren't any distributions based on it because it takes time to
rev a kernel in a distribution, I don't know of any particular issues
with the early 2.6 kernels that caused them to not go that direction.
Just the normal quality control issues of changing the kernel for your
product and timing issues (new GNOME and ` were on the horizon and
that's a big thing to include also.)  Now I've not been privy to all
the discussions within SuSE, Redhat and the others but they've got
other business constraints and I think this was a far more complex
timing issue than x.0 trustworthyness. By the time they were ready to
roll with 2.6 kernels, the kernel was up a few revs and the fixes made
sense to incorporate.  Maybe I'm full of it but I think I remember
most of the major players having other schedules that they made prior
to the 2.6 release.

I think if the tree-ssa stuff doesn't deserve a 4.0 then I'm not sure
what's in the pipeline that does.  Language support is changing, at
least for a while, that's huge.  Regardless of whether it's 3.5 or 4.0
there will be an x.y.1 release within a few months also, not because
of quality problems exactly but there will be bugs that won't be found
until a lot more people start using it.  I'm not advocating the
lowering of standards or saying bad things about people and their code
or anything like that, I'm just being pragmatic, this isn't the flight
controller for the space shuttle and it won't be developed that way,
there will be bugs andthere will be minor releases to fix them.  So
I'd say it should be 4.0 to reflect the depth of the changes that are
being made.  The perception that I'd want to present is that this is a
big upgrade,  it might not be perfect, it could affect you
dramatically in some way (like your Ada won't compile) so buyer
beware.  If they are depending on a tool and don't weigh and put
serious thought in to changing it then it they've got other problems
and the versioning won't fix that.  I'm not sure why you wouldn't bump
it to 4 unless there are some fear issues or political issues but the
crazy stuff has already happened, tree-ssa is trunk, that's the course
of action, the train is moving in that direction.  Personally, I'd
jump to 3.5 from 3.4 with a little less trepidation and that's the
wrong impression to give.


- --
signature


*Ian S. Nelson
PGP/GPG email preferred.
Public Key: 00D3D983
<http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x00D3D983>
Fingerprint: 3EFD7B86B888D7E229B69E97576F1B9700D3D983 *

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAweaZV28blwDT2YMRAvSfAJ4s7lj4a6MPTe5krrCy/Amc//aJeACdFUfD
Gt2sCanR87t3uhC6C/04Nkc=
=zNHC
-----END PGP SIGNATURE-----