Mailing List Archive

Re: svn commit: r1893284 - /httpd/test/framework/trunk/README
Hi,

I've also been looking at this README file today and I have a few questions:
  - some of the explicitly named modules looks redundant with what is
already in Bundle::ApacheTest
  - Crypt::SSLeay looks spurious. It doesn't compile on my machine
because of function name conflicts with my openssl
    there is also a big warning that says that (recent)
LWP::Protocol::https should be enough
  - this list is not in line with what is configured in .travis.yml

What puzzles me the most is the 3rd point.

CJ


Le 12/09/2021 à 19:12, covener@apache.org a écrit :
> Author: covener
> Date: Sun Sep 12 17:12:55 2021
> New Revision: 1893284
>
> URL: http://svn.apache.org/viewvc?rev=1893284&view=rev
> Log:
> add FCGI and show cpanminus syntax for installation of all at onace
>
>
> Modified:
> httpd/test/framework/trunk/README
>
> Modified: httpd/test/framework/trunk/README
> URL: http://svn.apache.org/viewvc/httpd/test/framework/trunk/README?rev=1893284&r1=1893283&r2=1893284&view=diff
> ==============================================================================
> --- httpd/test/framework/trunk/README (original)
> +++ httpd/test/framework/trunk/README Sun Sep 12 17:12:55 2021
> @@ -29,19 +29,15 @@ http://svn.apache.org/viewvc/perl/modper
> To run the tests for all Apache web server modules, some additional
> CPAN modules will be required. If the tests don't work, make sure
> that you have up to date versions of each of these perl modules:
> -- HTTP::DAV (DAV tests)
> -- DateTime (mod_include tests)
> -- Time::HiRes
> -- Protocol::HTTP2::Client and AnyEvent (mod_http2 tests)
> -- Test
> -- Test::Harness
> -- Crypt::SSLeay
> -- Net::SSLeay
> -- IO::Socket::SSL
> -- IO::Socket::IP
> -- IO::Select
> -- LWP::Protocol::https
> -- AnyEvent::WebSocket::Client;
> +
> +```
> +cpan App::cpanminus
> +cpanm Bundle::ApacheTest \
> + HTTP::DAV DateTime Time::HiRes Protocol::HTTP2::Client Test \
> + Test::Harness Crypt::SSLeay Net::SSLeay IO::Socket::SSL \
> + IO::Socket::IP IO::Select LWP::Protocol::https AnyEvent \
> + AnyEvent::WebSocket::Client FCGI
> +```
>
>
> Quick Start
>
>
Re: svn commit: r1893284 - /httpd/test/framework/trunk/README [ In reply to ]
On Sun, Sep 12, 2021 at 1:36 PM Christophe JAILLET
<christophe.jaillet@wanadoo.fr> wrote:
>
> Hi,
>

Caveat: I don't know much of anything on this stuff


> I've also been looking at this README file today and I have a few questions:
> - some of the explicitly named modules looks redundant with what is
> already in Bundle::ApacheTest
> - Crypt::SSLeay looks spurious. It doesn't compile on my machine
> because of function name conflicts with my openssl
> there is also a big warning that says that (recent)
> LWP::Protocol::https should be enough

Seems consistent with this rev:

$ svn log -c r1237461
------------------------------------------------------------------------
r1237461 | rjung | 2012-01-29 15:39:23 -0500 (Sun, 29 Jan 2012) | 14 lines

Add LWP::Protocol::https to Bundle::ApacheTest
and update version number of the bundle.
Otherwise https support will not be installed.

The docs of the added module tell us:

"This module used to be bundled with the libwww-perl,
but it was unbundled in v6.02 in order to be able to
declare its dependencies properly for the CPAN
tool-chain. Applications that need https support can
just declare their dependency on LWP::Protocol::https
and will no longer need to know what underlying
modules to install."

I guess the way we run it we can just casually change it in SVN and
not worry about a release?

> - this list is not in line with what is configured in .travis.yml

From eyeballing they don't seem too far off

> What puzzles me the most is the 3rd point.

Unfortunately I spoiled my env a little after reporting my results by
playing with the cpanm one-liner, so when I just went in to remove
Crypt::SSLeay I actually got more successful tests than what I
reported for 2.4.49.

I think we could synch them up, using the more modern list in travis
as a basis, and cross-link them with comments so they have a chance of
not drifting much further