Re: More problems with `make -n' in automake-generated rules.

2010-11-10 Thread Ralf Wildenhues
* Ralf Wildenhues wrote on Mon, Nov 01, 2010 at 10:18:55PM CET: > Fix and document rules to not touch the tree with `make -n'. > > * doc/automake.texi (Multiple Outputs): Document the problem of > modifications during dry-run execution, propose solution. > * NEWS: Update. >

Re: [PATCH 2/7] Tests defs: prefer "$curdir" over "`pwd`".

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:16:53PM CET: > * tests/defs.in: We already save the value of `pwd` in $curdir > early in the file, so there no need to recalculate it later, when > the current working directory is not changed. OK. Thanks, Ralf

Re: [PATCH 4/7] Tests defs: do not print message "Running test $0" anymore.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:18:53PM CET: > * tests/defs.in: Printing the message "=== Running test $0" at > the beginning of each tests made sense when Automake used the old > test-driver, which sent all the output directly to stdout/stderr. > Now that the parallel test-dri

Re: [PATCH 3/7] Tests defs: rename $curdir -> $testbuilddir

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:17:48PM CET: > * tests/defs.in: Rename $curdir to $testbuildir, for clarity and > consistency with $testsrcdir and $top_testsrcdir. OK. Thanks, Ralf

Re: [PATCH 5/7] Tests defs: $testsbuilddir is now AC_SUBST'ed.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:19:55PM CET: > * tests/defs.in ($testsbuilddir): Initialize "statically" > with the value AC_SUBST'ed from @abs_builddir@, rather than > "dinamically" with the value returned by `pwd`. dynamically You could also just write instead: * tests/d

Re: [PATCH 6/7] Tests defs: new AC_SUBST'ed variable $top_testsbuilddir.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:21:09PM CET: > * tests/defs.in ($top_testsbuilddir): New variable, initialized > with the value AC_SUBST'ed from @abs_top_build...@. Mostly for > completeness and consistency with $testsrcdir and $top_testsrcdir. Let's add this only when we nee

Re: [PATCH 1/7] Tests defs: various reorderings, some improvements.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:00:47PM CET: > This is basically just a reordering patch, which should organize the > code in `tests/defs.in' in a clearer and more rational way. > > * tests/defs.in: Reordered various snippets of code in a clearer > way. Improved a couple of e

Re: [PATCH 7/7] Tests defs: move static definitions in a new file `minidefs'.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:29:27PM CET: > Tests defs: move static definitions in a new file `minidefs'. minidefs is not a good name, as it is bound to not stay "mini". How about defs-static or defs-init? (And yes, I'm shying away from renaming defs, that's just too big a

Re: [PATCH 5/5] Tests defs: improve messages for skipped tests.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Wed, Sep 08, 2010 at 08:57:19PM CEST: > * tests/defs.in: Give meaningful messages about the reasons of a > test skip; this is especially useful as this file is run without > verbose xtraces on. Related reorderings in the code and new > comments. What happened to this

Re: [PATCH 5/7] Tests defs: $testsbuilddir is now AC_SUBST'ed.

2010-11-10 Thread Stefano Lattarini
On Wednesday 10 November 2010, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:19:55PM CET: > > * tests/defs.in ($testsbuilddir): Initialize "statically" > > with the value AC_SUBST'ed from @abs_builddir@, rather than > > "dinamically" with the value returned by `pwd`.

Re: [PATCH 6/7] Tests defs: new AC_SUBST'ed variable $top_testsbuilddir.

2010-11-10 Thread Stefano Lattarini
On Wednesday 10 November 2010, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:21:09PM CET: > > * tests/defs.in ($top_testsbuilddir): New variable, initialized > > with the value AC_SUBST'ed from @abs_top_build...@. Mostly for > > completeness and consistency with $te

Re: [PATCH 7/7] Tests defs: move static definitions in a new file `minidefs'.

2010-11-10 Thread Stefano Lattarini
On Wednesday 10 November 2010, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:29:27PM CET: > > Tests defs: move static definitions in a new file `minidefs'. > > minidefs is not a good name, True, but I couldn't think of a better one right away. Luckily, you could ;-

Re: [PATCH 1/7] Tests defs: various reorderings, some improvements.

2010-11-10 Thread Stefano Lattarini
On Wednesday 10 November 2010, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:00:47PM CET: > > This is basically just a reordering patch, which should organize the > > code in `tests/defs.in' in a clearer and more rational way. > > > > * tests/defs.in: Reordered vari

Re: [PATCH 5/5] Tests defs: improve messages for skipped tests.

2010-11-10 Thread Stefano Lattarini
On Wednesday 10 November 2010, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Wed, Sep 08, 2010 at 08:57:19PM CEST: > > * tests/defs.in: Give meaningful messages about the reasons of a > > test skip; this is especially useful as this file is run without > > verbose xtraces on. Related reord

Re: [PATCH 7/7] Tests defs: move static definitions in a new file `minidefs'.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Thu, Nov 11, 2010 at 12:16:27AM CET: > On Wednesday 10 November 2010, Ralf Wildenhues wrote: > > * Stefano Lattarini wrote on Sun, Nov 07, 2010 at 04:29:27PM CET: > > > Tests defs: move static definitions in a new file `minidefs'. [...] > > > * tests/defs: ... this new

Re: [PATCH 1/7] Tests defs: various reorderings, some improvements.

2010-11-10 Thread Ralf Wildenhues
* Stefano Lattarini wrote on Thu, Nov 11, 2010 at 12:37:23AM CET: > Well, the patch is "almost" just code reordering, with the very minor > exceptions that: > > 1. in the older version, some error messages could be printed before > `$me' was defined, so they couldn't use it; now `$me' is de