Re: reply>> my first newbie question

On Tue, Nov 18, 2003 at 03:14:25PM +0100, Tomastiq wrote:
> Hi,
> 2. I'm a little bit confused...
> > Skunk is supposed to look at the mtime of the file you've changed
> > and recompile it if it is older than the cached version.
> older? Is this a misprint or I misunderstanding something?
> I mean the compiling is needed if the changed file is newer than the cached, isn't it?

Did I write that?  Sorry.  It recompiles when the cached version is
older than the file in the docroot.

> > What platform are you using, first of all?
> Ok, I forgot it, although I thought it's important.
> I'm a poor win user, so my platform is:
> nt4+cygwin+python2.3.2(cygwin)+mx-base2.0.5+skunk3.4b5

Aha!  

> One more interesting thing. In the var/log/access.log every entry
> has wrong time (real time - 1 hour).

Could you compare what Python gives for
time.time()/time.asctime()/etc.  with the system clock?  And also
verify that cygwin is reporting the correct mtimes *and* ctimes for
files?  Look for what values you get in Python from os.stat(), too.

What is odd, though, is that if cygwin is off by an hour for some
reason, you'd expect both the mtimes of the docroot files and of the
cache files to be off by the same amount.  hmmm.

> note1: maybe running skunk under cygwin is not supported, but I'd like
> to start testing until I have the linux pc. It would be just a big help.

It is worth getting it to work well enough on cygwin so that
development can be done there, but it won't be a production target!

> note2: to brave cygwin/skunk testers: modify first line of the swmgr shell
> script from #!/bin/sh to #!/bin/bash is a good idea. Just because in this
> form it will be run ;-)

This is fixed in CVS (bash features are removed from that script).

Cheers,

Jacob S.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQE/umGAuqamFyFXXLIRAhUcAJ9OD+8OLw/4/jcYKykiyOGoPDjS3ACgqUi0
V4nlvmhZqk8Mz6RNkFrGOg8=
=O6wm
-----END PGP SIGNATURE-----