One observation I will make, which will probably not be news to many of you, is that most of the breakage I see happens in the test phase of the install.
EG, module Foo doesn't depend directly on module Bar, but one of the tests dies uglily ("died before could run tests" or something along those lines), and then you get, won't install without force. When you then install Bar so the tests can run wtihout the ugly die, and the rest of the install goes smooth. The difficulty is that that die happened deep inside pages of output that you have to manually browse over to figure out where things went awry.
If cpan (or cpanp) had detected Bar as a dependency during the install, and added it to the install queue, presumably things would have install cleanly.
I'm wondering if CPANPLUS has a mechanism for that, and if that could account for some of the difference. Or if not that, what.
Thomas.
2006/8/21, Chisel Wright <[EMAIL PROTECTED]>:
On Mon, Aug 21, 2006 at 11:47:26AM +0200, Thomas Hartman wrote:
> Having trouble with just cpan, and was suggested to use cpan plus.
I find that some modules install better with cpan, and others with
cpanplus
I think I'm primarily using cpan, but in an ideal world I'd be creating
deb/ubuntu packages for everything and installing those.
(the "official" ones are terribly out of date)
--
Chisel Wright
e: [EMAIL PROTECTED]
w: http://www.herlpacker.co.uk/
This is not an automated signature. I type this in to the bottom of every
message.
_______________________________________________
List: Catalyst@lists.rawmode.org
Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst
Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/
Dev site: http://dev.catalyst.perl.org/
_______________________________________________ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/
