Re: dist-xz compression level

Hello Pavel, * Pavel Sanda wrote on Wed, Apr 07, 2010 at 01:22:06PM CEST: > the newly added dist-xz target produce worse compressed archives > than lzma-dist. The reason is that automake call lzma with > best compression while it won't use -9 level for xz. > Is this intention or bug? Bug, I guess

Re: [PATCH] Check that symlinks are resolved by `make dist'.

Hello Stefano, * Stefano Lattarini wrote on Sun, Apr 04, 2010 at 08:57:19PM CEST: > At Sunday 04 April 2010, Ralf Wildenhues wrote: > > * Stefano Lattarini wrote on Fri, Apr 02, 2010 at 12:29:49AM CEST: > > > > By the way, how is the Automake-generated Makefile expected to > > > > behave if a fil

Re: [PATCH 0/3] Enable `errexit' a.k.a. `set -e' shell flag in all test scripts.

* Stefano Lattarini wrote on Thu, Apr 08, 2010 at 04:42:35PM CEST: > Oops, I forgot to bump the copyright years in the modified test. > An updated patch is attached, which also slighty improves the > considerations about the "safety" of the introduction of `set -e'. I've squashed these patches a

Re: [PATCH] Avoid possible false negatives in cond46.test.

* Stefano Lattarini wrote on Fri, Apr 09, 2010 at 02:19:32AM CEST: > Avoid possible false negatives in cond46.test. > > * tests/cond46.test: Enable shell `errexit' flag (and bumped > copyright years). Due to this change, the testcase should now > fail on unexpected failures in calls to $ACLOCAL/$

Re: [PATCH 2/2] Bugfix in confh5.test w.r.t. Solaris/Heirloom Sh.

* Stefano Lattarini wrote on Fri, Apr 09, 2010 at 05:11:44PM CEST: > * tests/confh5.test: In the generated Makefile.am: do not use > `test ! -e FILE' to check for the non-existence of a file, since > that is not supported by Solarish/Heirloom Sh. Applied to maint. Thanks! Ralf

Re: [PATCH 1/2] Use `set -e' in confsub.test (avoid false negatives).

Hello Stefano, * Stefano Lattarini wrote on Fri, Apr 09, 2010 at 07:25:31PM CEST: > -$ACLOCAL || Exit 1 > -$AUTOMAKE || Exit 1 > +$ACLOCAL > +$AUTOMAKE there is no need to actually remove these. They may not be strictly required any more with 'set -e', but they may still serve as visual reminder

Fix per-Makefile.am setting of -Werror.

Setting -Werror in a Makefile.am affects the handling of Makefile.am files treated later by the same automake process (more precisely, the same thread): sub/Makefile.am: AUTOMAKE_OPTIONS = -Werror # no warnings happen in this file Makefile.am: AUTOMAKE_OPTIONS = -Wall # redefined variable warning

Re: [PATCH 1/2] Use `set -e' in confsub.test (avoid false negatives).

At Sunday 11 April 2010, Ralf Wildenhues wrote: > Hello Stefano, > > * Stefano Lattarini wrote on Fri, Apr 09, 2010 at 07:25:31PM CEST: > > -$ACLOCAL || Exit 1 > > -$AUTOMAKE || Exit 1 > > +$ACLOCAL > > +$AUTOMAKE > > there is no need to actually remove these. Strictly speaking, no; but their r

Re: [PATCH] Avoid possible false negatives in cond46.test.

At Sunday 11 April 2010, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Fri, Apr 09, 2010 at 02:19:32AM CEST: > > Avoid possible false negatives in cond46.test. > > > > * tests/cond46.test: Enable shell `errexit' flag (and bumped > > copyright years). Due to this change, the testcase shou

Re: [PATCH] Avoid possible false negatives in cond46.test.

* Stefano Lattarini wrote on Sun, Apr 11, 2010 at 09:04:55PM CEST: > At Sunday 11 April 2010, Ralf Wildenhues wrote: > > * Stefano Lattarini wrote on Fri, Apr 09, 2010 at 02:19:32AM CEST: > > > Avoid possible false negatives in cond46.test. > > > > > > * tests/cond46.test: Enable shell `errexit' fl

Re: dist-xz compression level

Hello Andreas, * Andreas Jellinghaus wrote on Sun, Apr 11, 2010 at 09:37:13PM CEST: > isn't xz extremely slw with -9? > maybe it wasn't a bug, bit intentionally not used, > as that huge extra amount of time doesn't result in > that many bytes saved. Well, does somebody have numbers (memory, t

[PATCH 0/5] Prepend 'error: ' or 'warning: ' to messages.

Hello Bruno, all, I'm sure it was you who requested this feature, but I can't find the reference to this issue now. Anyway, here's a small patch series to let automake and aclocal consistenly prefix warning and error messages, applied to git master. I wasn't completely sure about the case of war

Re: [PATCH] Avoid possible false negatives in cond46.test.

At Sunday 11 April 2010, Ralf Wildenhues wrote: > > How'd you suggest to shorten the log entry above, while keeping > > it meaningful? > > By cutting the part which I did not quote. :-) Oops, I missed that. > > Being verbose is fine for things that are unobvious, but you may > assume that dev

[PATCH 1/5] Fix connected warnings about obsolete exeext override.

* lib/Automake/Rule.pm (define): Merge two warnings that belong together, by setting the 'partial' flag for the first one. Signed-off-by: Ralf Wildenhues --- This is just a preparation patch, so we don't prepend the second part of the message. Cheers, Ralf ChangeLog|4 li

[PATCH 2/5] Fix capitalization of error messages, reword one message.

* lib/Automake/Variable.pm (define): Do not capitalize the first word in the error message. * automake.in (require_file_internal): Likewise. Also, reword and line-wrap for better readability. * tests/distcom7.test, tests/pluseq5.test, tests/pluseq9.test: Adjust tests. Signed-off-by: Ralf Wildenhu

[PATCH 3/5] Prepend type to warning, error, and fatal messages.

For the first part of messages of types `error' or `fatal', prepend `error: ' to the message. Prepend `warning: ' to warning messages, whatever the setting of -Werror. * lib/Automake/Channels.pm (partial): Move up definition. (_format_message): Emit `header' and `footer' strings only with the fir

[PATCH 4/5] Ensure we don't print 'warning:' or 'error:' twice.

* Makefile.am (sc_ensure_testsuite_has_run): New helper rule. (sc_tests_logs_duplicate_prefixes): New rule, to check for wrong-looking warning and error messages in the test suite logs. * aclocal.in (scan_configure_dep, scan_file): Remove duplicate `warning: ' prefix. * tests/vars3.test: Rename var

[PATCH 5/5] Print 'warnings are treated as errors' note if needed.

* lib/Automake/Channels.pm: Fix typo in comment. (_print_message): If -Werror is enabled, print a 'warnings are treated as errors' note before the first such warning. * tests/warnopts.test: Adjust comment to mention expected warning. Adjust code to ignore 'warnings are treated as errors' note. * te

[PATCH] Fix typos in comments in test confh5.test

I've just found a couple of stupid typos in the last version of the patch. They are in a comment, so no big deal, but a fix is always welcome, right? Thanks, and sorry for the noise, Stefano From 8005809d6e64af309d2e4b5499ac93317f1f1d3b Mon Sep 17 00:00:00 2001 From: Stefano Lattarini Date:

Re: [PATCH 0/5] Prepend 'error: ' or 'warning: ' to messages.

Hello Ralf, > I'm sure it was you who requested this feature, but I can't find the > reference to this issue now. Yes, I raised it, but I'm sure others have wanted the same feature before me. > I wasn't completely sure about the case of warnings in the presence > of -Werror. ... we print a note

PATCH 1/2] Avoid possible false negatives in canon-name.test.

Avoid possible false negatives in canon-name.test (by enabling `set -e'). * tests/canon-name.test: Enable shell `errexit' flag (and bumped copyright years). Due to this change, the testcase should now fail on unexpected failures of aclocal, and on failures in grepping the expected diagnostic in a

[PATCH 2/2] Improve description of test canon-name.test.

Improve description of test canon-name.test. * tests/canon-name.test: Improve the brief description of this test that is given in heading comments. I thought a better description was needed, since at a first glance it was definitely unclear to me what the purpose of the test was, and only a searc

Re: [PATCH] Avoid possible false negatives in cond46.test.

* Stefano Lattarini wrote on Sun, Apr 11, 2010 at 10:30:19PM CEST: > At Sunday 11 April 2010, Ralf Wildenhues wrote: > > Being verbose is fine for things that are unobvious, but you may > > assume that developers know autoconf.info. > Quite right. However, as you said when discussing my proposed