Marcus Boerger kirjoitti:
Hello Jani,
Monday, July 21, 2008, 6:08:15 PM, you wrote:
http://cvs.php.net/viewvc.cgi/php-src/run-tests.php?r1=1.226.2.37.2.35.2.34&r2=1.226.2.37.2.35.2.35&diff_format=u
Index: php-src/run-tests.php
diff -u php-src/run-tests.php:1.226.2.37.2.35.2.34
php-src/run-tests.php:1.226.2.37.2.35.2.35
--- php-src/run-tests.php:1.226.2.37.2.35.2.34 Mon Jul 21 11:05:23 2008
+++ php-src/run-tests.php Mon Jul 21 16:08:15 2008
@@ -857,16 +863,16 @@
$failed_tests_data .= "Bundled Libtool:\n$libtool\n";
$failed_tests_data .= "System Libtool:\n$sys_libtool\n";
$failed_tests_data .= "Compiler:\n$compiler\n";
- $failed_tests_data .= "Bison:\n". @shell_exec('bison --version
2>/dev/null'). "\n";
+ $failed_tests_data .= "Bison:\n". shell_exec('bison --version
2>/dev/null') . "\n";
Seeing this again, is it really a good idea to list those? Shouldn't we
instead check whether we have the Makefile for the php we test and show the
versions from the respective programs specified in it instead?
I only tried to sync HEAD and PHP_5_3. Nothing else. Seeing that part..well..how
useful info is that anyway? You actually can only trust what is in the binary.
And we don't currently save compiler, bison, re2c etc. versions anywhere, but
maybe they should be? Steph already added some COMPILER / ARCHITECTURE stuff to
phpinfo() (only set with win32 builds now). Libtool version is irrelevant. That
has been bundled for a long time.
Other than that, thanks for cleaning this up. I still thing we should
rewrite larger chunks of it though :-)
Well, the whole thing is quite complex as is. Kinda grew up to be like that? :)
If I had time I might rewrite some if not all of it..but perhaps it's better
left as a project for next year's GSoC. ;)
--Jani
--
PHP CVS Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
