Re: Frustrated trying to build beecrypt & rpm on a Debian system whereI'm not root

[Mike Meyer <mwm-keyword-rpm 43b295 mired org>]

On Sun, 25 Nov 2007 15:21:25 -0600 "Paul Johnson" <pauljohn32 gmail com> wrote:

> On Nov 25, 2007 11:41 AM, Mike Meyer <mwm-keyword-rpm 43b295 mired org> wrote:
> > On Sun, 25 Nov 2007 06:31:58 -0500 "Johri, Mayank (GTI)" <mayank_johri ml com> wrote:
> >
> > > Paul,
> > >
> > > It should be
> > > from os.path import curdir,... Instead of from os.path import (curdir,
> >
> > Of course, that this is in their sources means they haven't tested
> > against 2.4, so you may encounter more such problems. In this case,
> > the "What's new in 2.5" (http://docs.python.org/whatsnew/whatsnew25.html)
> > is probably worth reading.
> Wait a minute. The error is not in my python code. It is in the python
> library itself.   This one:
> 
> >From /usr/lib/python2.4/os.py
> 
> sys.modules['os.path'] = path
> from os.path import (curdir, pardir, sep, pathsep, defpath, extsep, altsep,
>     devnull)

Arg. In that case, Both Mayank and I are off base.  The change must
have been in 2.4, not 2.5. And the syntax error he identified
shouldn't be a syntax error. In fact, that code loads fine on my
RHEL4u4 using the system python (2.4.3).

Can you make sure that createrepo is using the Python you think it
should? That's the only explanation I can think of for the error you
quoted earlier.

Where can I find a copy of creatrepo (assuming it's python)?

> If RPM did compile against python2.4, then why is this cropping up now?

Catching errors at run time instead of compile time is a common
behavior for dynamic languages.

     <mike
-- 
Mike Meyer <mwm mired org>		http://www.mired.org/consulting.html
Independent Network/Unix/Perforce consultant, email for more information.