Re: [PATCH] gnulib-tool: List modules separately, explicit vs dependencies.

2010-04-02 Thread Bruno Haible
Hi Thien-Thi, > I saw the huge list of modules > pulled in and wondered where to start cutting (via "--avoid"). > This change to gnulib-tool helps me to do the job more quickly > by letting me ignore (for the purposes of experimental cutting) > the explicitly specified modules. If you just cut ra

[PATCH] maint.mk: fix some fallout

2010-04-02 Thread Eric Blake
Also rearrange ChangeLog to match commit order. * NEWS: Document the incompatible change, and its effect on cfg.mk. * top/maint.mk (sc_prohibit_test_minus_ao): Update. Signed-off-by: Eric Blake --- > I've folded those patches plus one more (here) > into your 3/3 and pushed the result: > I'm pu

gnulib-tool: Provide the value of --m4-base to modules

2010-04-02 Thread Bruno Haible
For the 'havelib' tests, it is necessary for the module to know the location of the macros in the tarball. This provides it, through a shell variable at configure time. Its use is generally not recommended, but in cases like this one, I don't see how to avoid it. 2010-04-02 Bruno Haible

gnulib-tool: Obey user's environment variables

2010-04-02 Thread Bruno Haible
When creating built sources, gnulib-tool should obey the environment variables that the user has set (those that are listed at the top of gnulib-tool). 2010-04-02 Bruno Haible gnulib-tool: Obey user's environment variables. * gnulib-tool (func_create_testdir): When creating bu

gnulib-tool: Create distributed built sources also for the tests

2010-04-02 Thread Bruno Haible
So far, "gnulib-tool --create-testdir" creates distributed built sources (e.g. getdate.c from getdate.y) only in the lib directory. This is also needed in the tests directory: 1) because getdate might occur as a test-related module, or 2) because some tests can also have distributed built sourc

Re: havelib tests

2010-04-02 Thread Bruno Haible
Jim Meyering wrote: > > Any objections to that? > > That sounds fine. > Go for it. Ralf Wildenhues wrote: > I think having these tests in gnulib is TRT. OK, I've moved the tests, after including a couple of necessary tweaks to gnulib-tool.

gnulib-tool: Ensure the tests in the main directory are executed first

2010-04-02 Thread Bruno Haible
This ensures that tests in subdirectories (likely long-running ones) are done last. 2010-04-02 Bruno Haible gnulib-tool: Ensure the tests in the main directory are executed first. * gnulib-tool (func_emit_tests_Makefile_am): Initialize SUBDIRS to start with the current

gnulib-tool: Ensure that long-running tests are executed last

2010-04-02 Thread Bruno Haible
This ensures that long-running tests are executed last. So that the user gets the maximum of information as quickly as possible. 2010-04-02 Bruno Haible gnulib-tool: Ensure that long-running tests are executed last. * gnulib-tool (func_emit_tests_Makefile_am): Emit the code fo

Re: gnulib-tool: Ensure that long-running tests are executed last

2010-04-02 Thread Eric Blake
On 04/02/2010 12:08 PM, Bruno Haible wrote: > This ensures that long-running tests are executed last. So that the user > gets the maximum of information as quickly as possible. That goes counter to the idea of parallel execution, where you want the long-running tests front-loaded so that the short

Re: gnulib-tool: Ensure that long-running tests are executed last

2010-04-02 Thread Jim Meyering
Eric Blake wrote: > On 04/02/2010 12:08 PM, Bruno Haible wrote: >> This ensures that long-running tests are executed last. So that the user >> gets the maximum of information as quickly as possible. > > That goes counter to the idea of parallel execution, where you want the > long-running tests fro

Re: gnulib-tool: Ensure that long-running tests are executed last

2010-04-02 Thread Bruno Haible
Hi Eric, Jim, > >> This ensures that long-running tests are executed last. So that the user > >> gets the maximum of information as quickly as possible. > > > > That goes counter to the idea of parallel execution, where you want the > > long-running tests front-loaded so that the shorter tests can

Re: [PATCH 09/19] Remove some obsolete source files from POTFILES.in

2010-04-02 Thread Eric Blake
[adding bug-gnulib] On 04/02/2010 04:47 PM, James Youngman wrote: > --- > po/POTFILES.in | 15 --- > 1 files changed, 0 insertions(+), 15 deletions(-) > > diff --git a/po/POTFILES.in b/po/POTFILES.in > index 3ab3107..ef3485a 100644 > --- a/po/POTFILES.in > +++ b/po/POTFILES.in > @@