[commit/branch] Version set to 6.7.90

[commit/branch] Version set to 6.7.90

[Joel Brobecker]

Hello,

In preparation for the upcoming pre-prelease, the version number has been
set to 6.7.90.

2008-03-13  Joel Brobecker  <brobecker@adacore.com>

        * version.in: Set version to 6.7.90.

-- 
Joel

PS: I'm thinking, for the next release, I'll do the README and version.in
    update together in one change. Need to update my scripts...

Index: version.in
===================================================================
RCS file: /cvs/src/src/gdb/version.in,v
retrieving revision 1.2428.2.15
diff -u -p -r1.2428.2.15 version.in
--- version.in  13 Mar 2008 00:00:35 -0000      1.2428.2.15
+++ version.in  13 Mar 2008 17:50:01 -0000
@@ -1 +1 @@
-6.7.90.20080313-cvs
+6.7.90

Re: [commit/branch] Version set to 6.7.90

[Andreas Schwab]

Joel Brobecker <brobecker@adacore.com> writes:

> Hello,
>
> In preparation for the upcoming pre-prelease, the version number has been
> set to 6.7.90.
>
> 2008-03-13  Joel Brobecker  <brobecker@adacore.com>
>
>         * version.in: Set version to 6.7.90.

This is unfortunate, since 6.7.90 is smaller than 6.7.90.20080313-cvs.
It would have been better to bump the version to 6.7.91.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [commit/branch] Version set to 6.7.90

[Joel Brobecker]

> This is unfortunate, since 6.7.90 is smaller than 6.7.90.20080313-cvs.
> It would have been better to bump the version to 6.7.91.

I'm just following the guidelines that are documented at
http://www.sourceware.org/gdb/current/onlinedocs/gdbint_16.html#SEC129.
The reason why it sounds strange is because I think the guidelines
are assuming that I should be making the first pre-release off the
branchpoint, so there should be no 6.7.90.something before 6.7.90.

I don't mind changing the procedures to do that, or changing the
numbering scheme, if it makes more sense.

-- 
Joel

Re: [commit/branch] Version set to 6.7.90

[Daniel Jacobowitz]

On Fri, Mar 14, 2008 at 10:29:00AM -0700, Joel Brobecker wrote:
> > This is unfortunate, since 6.7.90 is smaller than 6.7.90.20080313-cvs.
> > It would have been better to bump the version to 6.7.91.
> 
> I'm just following the guidelines that are documented at
> http://www.sourceware.org/gdb/current/onlinedocs/gdbint_16.html#SEC129.
> The reason why it sounds strange is because I think the guidelines
> are assuming that I should be making the first pre-release off the
> branchpoint, so there should be no 6.7.90.something before 6.7.90.

Right.  But if you don't make that snapshot, the first prerelease
should be 6.7.91; version numbers going backwards is bad mojo :-)

-- 
Daniel Jacobowitz
CodeSourcery

Re: [commit/branch] Version set to 6.7.90

[Andreas Schwab]

Joel Brobecker <brobecker@adacore.com> writes:

>> This is unfortunate, since 6.7.90 is smaller than 6.7.90.20080313-cvs.
>> It would have been better to bump the version to 6.7.91.
>
> I'm just following the guidelines that are documented at
> http://www.sourceware.org/gdb/current/onlinedocs/gdbint_16.html#SEC129.

It says: "As draft releases are drawn from the branch, the
@var{patchlevel} is incremented", which I think supports my position.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [commit/branch] Version set to 6.7.90

[Joel Brobecker]

> Right.  But if you don't make that snapshot, the first prerelease
> should be 6.7.91; version numbers going backwards is bad mojo :-)

Roger that - the first pre-release version number will be .90 if
it is made immediately after cutting the branch, or .91 if made later.

I'm still unsure whether making a pre-release right after cutting
the branch brings value or not. On the one hand, it's easier to
test, and in many ways more meaningful, a pre-release than something
downloaded off a branch. On the other hand, sometimes we know the
branch is missing some changes that will come later. Maybe this
needs to be a case-by-case judgement call. In the case of the 6.8
release cycle, we could have made a first pre-release immediately.

All in all, I'm thinking that we should make it a policy of creating
the pre-release immediately. The only case where this would provide
no benefit is when we know a major piece is missing, but in this case
the branch shouldn't have been cut in the first place, no?

-- 
Joel

Re: [commit/branch] Version set to 6.7.90

[Daniel Jacobowitz]

On Fri, Mar 14, 2008 at 12:44:37PM -0700, Joel Brobecker wrote:
> All in all, I'm thinking that we should make it a policy of creating
> the pre-release immediately. The only case where this would provide
> no benefit is when we know a major piece is missing, but in this case
> the branch shouldn't have been cut in the first place, no?

Makes sense to me.

-- 
Daniel Jacobowitz
CodeSourcery