[qmtest] running qmtest on bash in cygwin...

[Stefan Seefeld]

Mark Mitchell wrote:

>>>The default installation would put the modules under python's site-packages,
>>>and the rest into /usr/local/[share, bin], which is what I would expect
>>>from a typical debian / rpm package.
> 
> 
> Yes, that sounds right.
> 
> 
>>>And for those who want everything in a single isolated place, an installation
>>>using the '--home' parameter would probably do the trick. In any case, the
> 
> 
> "./configure --prefix" should still have that effect.

hmm, I thought to eliminate configure, too.

> That's very important because one mode in which QMTest is distributed is
> as a tarball containing binaries for both QMTest and Python.  That's
> important for users on systems where Python isn't installed by default.

could you elaborate a bit on that ? Usually when distributing binary
packages, there isn't any configure script (involved any more).

> The high-level concept is that while we'd like QMTest to play better
> with Python, we want to keep it usable as an independent application as
> well.

independent in the sense that it is bundled together with a python distro
such that there are no external dependencies ?

How do you package python then ? It sounds all you need for this case
is an installer that first unpacks python at a specific place and then
calls the python binary with the 'setup.py' script that is part of the
QMTest distribution.

>>>main executable could be a python script, and some little experimentation
>>>suggests that the qm/test/qmtest.py file is almost ready to serve, i.e.
>>>the qmtest.[bat, sh] files aren't really needed then.
> 
> 
> That is also a desirable goal.  
> 
> However, the logic where by qmtest.sh finds a version of Python
> installed in /path/to/qmtest/bin is required for the set-up above. 

the 'python setup.py build' command usually replaces the
'#! /usr/bin/env python' line in all installed scripts with the
real path to the python interpreter you use for the installation.
Sounds like that would meet your requirement.


> Therefore, something has to be done to preserve that behavior, or
> something equivalent.

see above.

Regards,
		Stefan