Re: GCC 3.0.4 Released.

Re: GCC 3.0.4 Released.

[Toon Moene]

Mark wrote:

> The next GCC release will be GCC 3.1.  This release will contain 
> many new features and improvements.  The anticipated release date 
> is April 15th, 2001.
                 ^^^^
I know it's hard to predict the future, but this smells like
revisionism.

--
Toon Moene, KNMI, PO Box 201, 3730 AE De Bilt, The Netherlands.
Tel. +31302206443, Fax +31302210407,  e-mail moene@knmi.nl
URL: http://www.knmi.nl/hirlam

Re: GCC 3.0.4 Released

[the_sage2000]

Mr. Mitchell?  Do you teach at LHS?

Oz

On Thu, 21 Feb 2002 16:26:11 -0800 Mark Mitchell <mark@codesourcery.com>
writes:
> GCC 3.0.4 is now available from sites listed at these URLs:
> 
>   http://www.gnu.org/order/ftp.html
>   http://gcc.gnu.org/mirrors.html
> 
> Not all of the mirrors have the release yet.  Therefore, it would
> be polite to wait for a day or two before rushing to download the
> new release.  [I would have delayed making this announcement, 
> except
> that people are starting to ask lots of questions about the release
> already.]
> 
> This release is a bug-fix release in the GCC 3.0 series.  There are
> no major new features; however, there are many bugs fixed.  People
> too numerous to mention contributed to fixing these bugs.
> 
> A summary of some of the improvements in this release is available
> at:
> 
>   http://gcc.gnu.org/gcc-3.0/features.html
> 
> Joe Buck was kind enough to prepare a list of PRs fixed in this
> release:
> 
>   http://gcc.gnu.org/ml/gcc/2002-02/msg01198.html
> 
> The next GCC release will be GCC 3.1.  This release will contain 
> many
> new features and improvements.  The anticipated release date is
> April 15th, 2001.
> 
> The MD5 sum files on mirrors of gcc.gnu.org are wrong; apparently
> some automated job created an md5.sum file there in the midst of
> my uploading of the files.  I have corrected the problem on
> gcc.gnu.org, mirrors will get the new md5.sum shortly.  The correct
> contents are:
> 
> 1e7ecec7e616d85cdec2d627bac3bf9b  gcc-3.0.4.tar.bz2
> 0d1bdcb6fb737a6cfa03b3f21e7c9a5d  gcc-3.0.4.tar.gz
> 91b2eb00067defd75b6ab35e25e18a3e  gcc-core-3.0.4.tar.bz2
> 9d251180255a83e43e38efd8a4adf276  gcc-core-3.0.4.tar.gz
> 3640e2dc5d06257ed612da6da402a0c7  gcc-g++-3.0.4.tar.bz2
> 6f5d491632f19610c180ddda06c2203d  gcc-g++-3.0.4.tar.gz
> 22c9def796cd7e01753cde3e01c125af  gcc-g77-3.0.4.tar.bz2
> 9b7c94554a488fa1a705c6dfeee7cb95  gcc-g77-3.0.4.tar.gz
> 5a921afda0a64aae2d933241201ee15c  gcc-java-3.0.4.tar.bz2
> e799c7c895ce82d90b1a0ffa798e  gcc-java-3.0.4.tar.gz
> af54f3eb7ea363728397fe9a445773e6  gcc-objc-3.0.4.tar.bz2
> 1d54eeb4da67297c67226df7f71d7b87  gcc-objc-3.0.4.tar.gz
> 3fa966d121e0fd69bbf833739241cc80  gcc-testsuite-3.0.4.tar.bz2
> 1c71fd57959467ea9008eb72ba545549  gcc-testsuite-3.0.4.tar.gz
> 
> -- 
> Mark Mitchell                mark@codesourcery.com
> CodeSourcery, LLC            http://www.codesourcery.com
> 

GCC 3.0.4 Released

[Mark Mitchell]

GCC 3.0.4 is now available from sites listed at these URLs:

  http://www.gnu.org/order/ftp.html
  http://gcc.gnu.org/mirrors.html

Not all of the mirrors have the release yet.  Therefore, it would
be polite to wait for a day or two before rushing to download the
new release.  [I would have delayed making this announcement, except
that people are starting to ask lots of questions about the release
already.]

This release is a bug-fix release in the GCC 3.0 series.  There are
no major new features; however, there are many bugs fixed.  People
too numerous to mention contributed to fixing these bugs.

A summary of some of the improvements in this release is available
at:

  http://gcc.gnu.org/gcc-3.0/features.html

Joe Buck was kind enough to prepare a list of PRs fixed in this
release:

  http://gcc.gnu.org/ml/gcc/2002-02/msg01198.html

The next GCC release will be GCC 3.1.  This release will contain many
new features and improvements.  The anticipated release date is
April 15th, 2001.

The MD5 sum files on mirrors of gcc.gnu.org are wrong; apparently
some automated job created an md5.sum file there in the midst of
my uploading of the files.  I have corrected the problem on
gcc.gnu.org, mirrors will get the new md5.sum shortly.  The correct
contents are:

1e7ecec7e616d85cdec2d627bac3bf9b  gcc-3.0.4.tar.bz2
0d1bdcb6fb737a6cfa03b3f21e7c9a5d  gcc-3.0.4.tar.gz
91b2eb00067defd75b6ab35e25e18a3e  gcc-core-3.0.4.tar.bz2
9d251180255a83e43e38efd8a4adf276  gcc-core-3.0.4.tar.gz
3640e2dc5d06257ed612da6da402a0c7  gcc-g++-3.0.4.tar.bz2
6f5d491632f19610c180ddda06c2203d  gcc-g++-3.0.4.tar.gz
22c9def796cd7e01753cde3e01c125af  gcc-g77-3.0.4.tar.bz2
9b7c94554a488fa1a705c6dfeee7cb95  gcc-g77-3.0.4.tar.gz
5a921afda0a64aae2d933241201ee15c  gcc-java-3.0.4.tar.bz2
e4f9e799c7c895ce82d90b1a0ffa798e  gcc-java-3.0.4.tar.gz
af54f3eb7ea363728397fe9a445773e6  gcc-objc-3.0.4.tar.bz2
1d54eeb4da67297c67226df7f71d7b87  gcc-objc-3.0.4.tar.gz
3fa966d121e0fd69bbf833739241cc80  gcc-testsuite-3.0.4.tar.bz2
1c71fd57959467ea9008eb72ba545549  gcc-testsuite-3.0.4.tar.gz

-- 
Mark Mitchell                mark@codesourcery.com
CodeSourcery, LLC            http://www.codesourcery.com

Re: GCC 3.0.4 Released

[Joe Buck]

> GCC 3.0.4 is now available from sites listed at these URLs:
> 
>   http://www.gnu.org/order/ftp.html
>   http://gcc.gnu.org/mirrors.html

Cool.  Could someone (Gerald?) update the gcc.gnu.org main web page
News/Announcements?

Re: GCC 3.0.4 Released

[Mark Mitchell]

--On Thursday, February 21, 2002 04:47:57 PM -0800 Joe Buck 
<jbuck@synopsys.COM> wrote:

>
>> GCC 3.0.4 is now available from sites listed at these URLs:
>>
>>   http://www.gnu.org/order/ftp.html
>>   http://gcc.gnu.org/mirrors.html
>
> Cool.  Could someone (Gerald?) update the gcc.gnu.org main web page
> News/Announcements?

I just did it.

Thanks,

-- 
Mark Mitchell                mark@codesourcery.com
CodeSourcery, LLC            http://www.codesourcery.com

Re: GCC 3.0.4 Released

[Joe Buck]

> > Cool.  Could someone (Gerald?) update the gcc.gnu.org main web page
> > News/Announcements?
> 
> I just did it.

You added the announcement, but the page still says, just above, that
3.0.3 is the latest release.

Re: GCC 3.0.4 Released

[Joseph S. Myers]

On Thu, 21 Feb 2002, Mark Mitchell wrote:

> GCC 3.0.4 is now available from sites listed at these URLs:

There should of course be diffs as well.  Does releasing.html need to
cover the actual process of running gcc_release, with a specific note that
the -p option (pointing to a previous release .tar.gz file, in a directory
with all the .tar.gz files of the previous release) should be used for
minor releases?

-- 
Joseph S. Myers
jsm28@cam.ac.uk

Re: GCC 3.0.4 Released

[Mark Mitchell]

--On Friday, February 22, 2002 01:13:22 AM +0000 "Joseph S. Myers" 
<jsm28@cam.ac.uk> wrote:

> On Thu, 21 Feb 2002, Mark Mitchell wrote:
>
>> GCC 3.0.4 is now available from sites listed at these URLs:
>
> There should of course be diffs as well.  Does releasing.html need to
> cover the actual process of running gcc_release, with a specific note that
> the -p option (pointing to a previous release .tar.gz file, in a directory
> with all the .tar.gz files of the previous release) should be used for
> minor releases?

Yes.  It needs to say everything you need to do to make a release; there
are already about twenty steps and the chances of one getting missed are
very high unless they're all written down or automated.

Perhaps -f without -p should just result in gcc_release issuing an error
message and stopping.

-- 
Mark Mitchell                mark@codesourcery.com
CodeSourcery, LLC            http://www.codesourcery.com

Re: GCC 3.0.4 Released

[Joseph S. Myers]

On Thu, 21 Feb 2002, Mark Mitchell wrote:

> Perhaps -f without -p should just result in gcc_release issuing an error
> message and stopping.

The trouble with that is that when releasing 3.1, you probably don't want
diffs from a previous (3.0.4) release.  (Diffs from a 3.1 prerelease or a
snapshot would be useful for some people - but it isn't so clear that they
should be mandated.)

-- 
Joseph S. Myers
jsm28@cam.ac.uk

Re: GCC 3.0.4 Released

[Joe Buck]

> On Thu, 21 Feb 2002, Mark Mitchell wrote:
> 
> > Perhaps -f without -p should just result in gcc_release issuing an error
> > message and stopping.

Joseph Myers writes:
> The trouble with that is that when releasing 3.1, you probably don't want
> diffs from a previous (3.0.4) release.  (Diffs from a 3.1 prerelease or a
> snapshot would be useful for some people - but it isn't so clear that they
> should be mandated.)

You're thinking of developers as opposed to end users, who will only be
using FSF releases.  For them, diffs from 3.0.4 to 3.1 will be useful if
the size is substantially smaller than just downloading 3.1.

Re: GCC 3.0.4 Released

[Joseph S. Myers]

On Thu, 21 Feb 2002, Joe Buck wrote:

> You're thinking of developers as opposed to end users, who will only be
> using FSF releases.  For them, diffs from 3.0.4 to 3.1 will be useful if
> the size is substantially smaller than just downloading 3.1.

They're also quite likely to run into limitations of gcc_update - even if
they do run it after patching to adjust timestamps, by design it touches
files in CVS only, so the timestamps of generated files that aren't in CVS
may well be wrong after patching.  (This will happen with info files
non-critically even with minor releases, and sometimes with
Bison-generated files in minor releases, but is much more likely to happen
for major releases.)

We may well also find during the 3.1 series (or for diffs 3.0.4 to 3.1 if
we create those) that I was overly optimistic to include -a in the diff
options in gcc_release, and that where binary files (compiled message
catalogs) change, diffs with -a cause problems for too many users, in
which case we'd need to go back to incomplete diffs (excluding changes to
compiled message catalogs) instead.

-- 
Joseph S. Myers
jsm28@cam.ac.uk

Re: GCC 3.0.4 Released

[Joe Buck]

> On Thu, 21 Feb 2002, Joe Buck wrote:
> 
> > You're thinking of developers as opposed to end users, who will only be
> > using FSF releases.  For them, diffs from 3.0.4 to 3.1 will be useful if
> > the size is substantially smaller than just downloading 3.1.
> 
> They're also quite likely to run into limitations of gcc_update - even if
> they do run it after patching to adjust timestamps, by design it touches
> files in CVS only, so the timestamps of generated files that aren't in CVS
> may well be wrong after patching.

If the patch is ordered in such a way that diffs to derived files appear after
diffs to the base files, then gcc_update is not needed.

Re: GCC 3.0.4 Released

[Joseph S. Myers]

On Thu, 21 Feb 2002, Joe Buck wrote:

> If the patch is ordered in such a way that diffs to derived files appear after
> diffs to the base files, then gcc_update is not needed.

I don't think GNU diff has an option to provide any sort of partial 
ordering information to get the diffs ordered that way.

-- 
Joseph S. Myers
jsm28@cam.ac.uk

Re: GCC 3.0.4 Released

[Jakub Jelinek]

On Fri, Feb 22, 2002 at 09:49:37AM +0000, Joseph S. Myers wrote:
> On Thu, 21 Feb 2002, Joe Buck wrote:
> 
> > If the patch is ordered in such a way that diffs to derived files appear after
> > diffs to the base files, then gcc_update is not needed.
> 
> I don't think GNU diff has an option to provide any sort of partial 
> ordering information to get the diffs ordered that way.

Yes, but it shouldn't be too hard to write an awk or perl script which would
do the necessary postprocessing of the diff.

	Jakub

Re: GCC 3.0.4 Released

[Mark Mitchell]

--On Friday, February 22, 2002 01:38:00 AM +0000 "Joseph S. Myers" 
<jsm28@cam.ac.uk> wrote:

> On Thu, 21 Feb 2002, Mark Mitchell wrote:
>
>> Perhaps -f without -p should just result in gcc_release issuing an error
>> message and stopping.
>
> The trouble with that is that when releasing 3.1, you probably don't want
> diffs from a previous (3.0.4) release.  (Diffs from a 3.1 prerelease or a
> snapshot would be useful for some people - but it isn't so clear that they
> should be mandated.)

Well, gcc_release knows what's a major release and what's a minor release.

-- 
Mark Mitchell                mark@codesourcery.com
CodeSourcery, LLC            http://www.codesourcery.com