Re: backport 3.4 <- 3.5 after today's freeze, but prior to ssamerge?

[Andreas Jaeger <aj at suse dot de>]

Paul Schlie <schlie@comcast.net> writes:

> Sorry, although I may misunderstand; the basis of my suggestion was that it
> may be beneficial in both the long and short runs to consider adopting 3.5
> prior to it's merge with ssa to considered the basis of 3.4, as although it
> does contain "new features", it also seems to be as, if not arguably more
> robust than 3.4 presently is. (not a big deal, just thought it may be worthy
> of consideration, if it hadn't already been given serious thought.)

Please read how we do GCC development, especially the different
phases.  Declaring 3.5 as 3.4 would start the clock again for 3.4
throwing 3.4 two months back.

3.5 might look as good to you as 3.4 but look at the patches that went
into 3.5, there is new code that can destabilize the compiler and just
nobody tested it.  We also have removed code (or currently work on
removing code) that we promised would be in 3.4 but will be removed in
3.5, e.g. -fwrite-strings.

I personally do not like to have 3.4 delayed by another two months and
your proposal would be just that.  We have made a plan and should
stick to it, I see no technical reason to do a change.

Andreas
-- 
 Andreas Jaeger, aj@suse.de, http://www.suse.de/~aj
  SuSE Linux AG, Maxfeldstr. 5, 90409 Nürnberg, Germany
   GPG fingerprint = 93A3 365E CE47 B889 DF7F  FED1 389A 563C C272 A126

Attachment: pgp00000.pgp
Description: PGP signature