Re: Ann: CherryPy-2.0-beta released
I'm a great believer that avoiding query strings in URL's is good practise ( http://www.holloway.co.nz/book/9 for good arguments why). How tricky is it to remap URL's to query strings in cherryPy? Also how much does it complicate matters to run cherryPy under an existing webserver? Is any functionality lost? -- http://mail.python.org/mailman/listinfo/python-list
Re: Integration with java
How about this? http://jpype.sourceforge.net/ (I haven't used it myself) -- http://mail.python.org/mailman/listinfo/python-list
Re: Overriding methods in classes you don't control
I've also come across several situations where modifying existing
classes seems the simplest and most intuitive thing to do.
Obviously it is fine for quick hacks but people tend to go 'whoo'
and wave their fingers around when people suggest it as a general
technique.
Has anyone got any experience of doing this on medium-size projects and
how quickly things descend into an unmaintainable mess?
What are the specific problems and can they be avoided/worked around?
I suppose the main thing is the problems it causes for other people
reading your source code i.e. how are they supposed to know that Class
X now has some of your code injected into it. And thinking about it.
How are you supposed to remember 3 months down the line!
Maybe a 'best practises' system where any such modifications were made
or mentioned right at the start of the source tree or in a distinctly
named module could alleviate this?
runsun pan wrote:
> Some untested idea:
>
> 1. find what is the base class of wxPython
> 2. x= subclass from that base class
> 3. make wxPython.__base__ = x
>
> Don't know if this works, but this is what I would try if I were u
> ('cos its ease).
--
http://mail.python.org/mailman/listinfo/python-list
setup.py when you can't write to site-packages?
There are several gaps in my Python knowledge, one of which is the what exactly setuptools does and how it works. I'm on a shared host so can't write to site-packages therefore most setup.py's fail. My strategy for pure python packages is to run setup.py locally and copy anything that gets put in site-packages across to the shared host somewhere in my Python path. I'm sure there is a better solution! What's the best approach for situations when you can't tamper with the Python install? -- http://mail.python.org/mailman/listinfo/python-list
Re: setup.py when you can't write to site-packages?
Thanks Diez. For one thing I was getting setuptools mixed up with distutils. Then it occurred to me that this might be covered in the standard distutils docs (obvious I know but before I was thinking of it as a general Python problem and therefore wasn't sure where it might be documented) The solution was therefore right in plain view: http://docs.python.org/inst/alt-install-windows.html (despite the misleading url this page applies to Unix) (Must. Check. Docs. Before. Posting.) Diez B. Roggisch wrote: > andybak wrote: > > > There are several gaps in my Python knowledge, one of which is the what > > exactly setuptools does and how it works. > > > > I'm on a shared host so can't write to site-packages therefore most > > setup.py's fail. > > > > My strategy for pure python packages is to run setup.py locally and > > copy anything that gets put in site-packages across to the shared host > > somewhere in my Python path. I'm sure there is a better solution! > > > > What's the best approach for situations when you can't tamper with the > > Python install? > > If you've got setuptools installed, you can use the development mode + a > staging dir that you put in your PYTHONPATH. Then the copying is done for > you. > > Diez -- http://mail.python.org/mailman/listinfo/python-list
Jason Orendorff's path module
Hi, I know this has been consigned to oblivion by the BDFL but I'm rather addicted to it's use. I notice the original site is no longer there. Has anyone got a cached copy of the page that I can host for posterity? Which rather makes me wonder - lot's of people liked this module or rather disliked os.path. Is there anything salvageable from the BDFL's objections worthy of a PEP? Andy -- http://mail.python.org/mailman/listinfo/python-list
