David Wheeler wrote:
On Jan 21, 2004, at 3:15 PM, Stas Bekman wrote:
So did you get the interactive prompt for the first time? (which of
course shouldn't happen if you have passed the -httpd/-apxs args/or if
it was against mp2, which it then has used and will remember to use in
the future if you
On Jan 21, 2004, at 3:15 PM, Stas Bekman wrote:
So did you get the interactive prompt for the first time? (which of
course shouldn't happen if you have passed the -httpd/-apxs args/or if
it was against mp2, which it then has used and will remember to use in
the future if you don't pass any args)
David Wheeler wrote:
On Jan 21, 2004, at 2:03 PM, Stas Bekman wrote:
So does it all just work or you haven't had a chance to test my rework
of the sticky configuration? Once you are happy with it, I'd like to
invite more testers from the outside (probably post to modperl users
list) before we ma
On Jan 21, 2004, at 2:00 PM, Stas Bekman wrote:
I think, yes.
It's actually a goodness to make tests relocatable, in case someone
moves things around.
Nice.
I'll wait for a blessing from Geoff and then try to fix the mp2 and
httpd-test/perl-framework tests first. If successful, then it'll be in
On Jan 21, 2004, at 2:03 PM, Stas Bekman wrote:
So does it all just work or you haven't had a chance to test my rework
of the sticky configuration? Once you are happy with it, I'd like to
invite more testers from the outside (probably post to modperl users
list) before we make a new A-T release.
On Wed, Jan 21, 2004 at 01:51:05PM -0800, Stas Bekman wrote:
> >>LogLeveldebug
> >
> >That's the one I needed to set.
>
> You can override it later in your extra conf files. But it's a good idea to
> keep it at the debug level.
I was able to override but will follow your advice. The info abo
William, please take a look at the current cvs, I've updated it with an
extended section on tracing.
here is the document explaining how to get the latest cvs of the docs
http://perl.apache.org/download/docs.html
Please check that it's all clear now and send patches if something is still
missing
So does it all just work or you haven't had a chance to test my rework of the
sticky configuration? Once you are happy with it, I'd like to invite more
testers from the outside (probably post to modperl users list) before we make
a new A-T release.
__
David Wheeler wrote:
And of course all tests that rely on the previous behavior need to be
fixed too. There quite a few tests to fix in mp2, I haven't tested
other repositories, yet.
It's a nice, straight-forward patch, so my guess is that the tests would
be fairly easy to fix, yes?
I think, yes
William McKee wrote:
On Wed, Jan 21, 2004 at 12:35:07PM -0800, Stas Bekman wrote:
William McKee wrote:
The testing documentation at perl.apache.org says that the current trace
level defaults to info. However, my system is outputting debug messages
to the t/logs/error_log file. Is this the correct b
On Jan 21, 2004, at 1:11 PM, Stas Bekman wrote:
It just doesn't do any chdirs, so you can really run:
/foo/bar/tar/mar/t/TEST and it will run from /
Cool, I suspect that that's more or less how Test::Harness works, too
-- it just runs the test from whatever Cwd you're in when you call
runtests().
On Wed, Jan 21, 2004 at 04:31:50PM -0500, William McKee wrote:
> > Please suggest a clarification to the docs now that you know all the
> > answers to this issue, William ;)
>
> Attached :).
Please ignore the previous diff. I've revised the changes to include how
to output messages.
William
--
On Wed, Jan 21, 2004 at 12:35:07PM -0800, Stas Bekman wrote:
> William McKee wrote:
> >The testing documentation at perl.apache.org says that the current trace
> >level defaults to info. However, my system is outputting debug messages
> >to the t/logs/error_log file. Is this the correct behavior in
David Wheeler wrote:
On Jan 21, 2004, at 12:39 PM, Stas Bekman wrote:
Sorry, David, I guess I was testing with the wrong version. Here is a
new patch. It breaks quite a few tests in modperl-2.0 but it's because
they rely on that chdir_t, and can be fixed to be relocatable. It's
possible that the
On Jan 21, 2004, at 12:39 PM, Stas Bekman wrote:
Sorry, David, I guess I was testing with the wrong version. Here is a
new patch. It breaks quite a few tests in modperl-2.0 but it's because
they rely on that chdir_t, and can be fixed to be relocatable. It's
possible that there are other implicat
David Wheeler wrote:
On Jan 20, 2004, at 11:20 PM, Stas Bekman wrote:
Does the patch below solve the problem?
Maybe, but it's not quite there yet:
Can't locate object method "chdir_t" via package "Apache::TestHarness"
at lib/Apache/TestRun.pm line 648.
Sorry, David, I guess I was testing with th
William McKee wrote:
The testing documentation at perl.apache.org says that the current trace
level defaults to info. However, my system is outputting debug messages
to the t/logs/error_log file. Is this the correct behavior in which case
the docs are wrong? Or have I somehow configured my system t
On 22/01/2004, at 3:23 AM, William McKee wrote:
Hi Stas, Geoff, Nick, et al,
A quick question for you folks regarding response handlers. Now that
I'm
understanding how to use these tests, I'm finding that I can replace my
tests that used the Apache::FakeRequest module for doing testing of my
serv
The testing documentation at perl.apache.org says that the current trace
level defaults to info. However, my system is outputting debug messages
to the t/logs/error_log file. Is this the correct behavior in which case
the docs are wrong? Or have I somehow configured my system to default to
debug mo
On Jan 20, 2004, at 11:20 PM, Stas Bekman wrote:
Does the patch below solve the problem?
Maybe, but it's not quite there yet:
Can't locate object method "chdir_t" via package "Apache::TestHarness"
at lib/Apache/TestRun.pm line 648.
Not yet. It should be trivial for someone who knows Module::Buil
Hi Stas, Geoff, Nick, et al,
A quick question for you folks regarding response handlers. Now that I'm
understanding how to use these tests, I'm finding that I can replace my
tests that used the Apache::FakeRequest module for doing testing of my
server-side code. I've also found that A::T will run
David Wheeler wrote:
Hi All,
A while ago, I enquired about the Cwd that Apache::Test uses, and asked
if something could be added to make it different. This is because a lot
of existing test modules (such as Test::POD) expect tests to be run from
the distribution package root. Apache::Test is fai
Hi All,
A while ago, I enquired about the Cwd that Apache::Test uses, and asked
if something could be added to make it different. This is because a lot
of existing test modules (such as Test::POD) expect tests to be run
from the distribution package root. Apache::Test is fairly unique
(outside
23 matches
Mail list logo