On 25 February 2014 15:43, Andy Bennett <andyjpb@ashurst.eu.org> wrote:
Hi,

> Do people think that's a sensible way to proceed?

This is very rarely, if ever a sensible way to proceed.
How do you know you won't end up right back where you are only without
anything to analyse?

I don't. How do I know that if I try another method of fixing the problem it will work? I don't either.

Life is full of uncertainty.
 
> (On a wider note, how to avoid these sorts of problems in the future? Is
> docker the way to go? Heroku?)

Get to the root cause of the problem and solve it from there.

I think you misunderstand me; I was discussing ways of avoiding similar problems in future, rather than fixing this one. 

Throwing
extra layers at the problem won't help you understand it any better.


It sounds like you have some kind of apache module issue. Maybe try
starting from the provided http config file (save your copy and get apt
to reinstall the provided one) and then build it back up piece by piece
until you work out the directives that are broken.

I know exactly which ones are broken, the WSGIScriptAlias ones, as i have already mentioned.

 
Then consult the
Apache documentation to work out which modules they are from. Then work
out whether those modules are installed and enabled and at the correct
version. Go from there.

I have tried using apt-get to uninstall and then re-install apache, without any success.

I strongly suspect the issue is to do with a mismatch between the versions of python used by different parts of the system.
 
--
Phil Hunt, <cabalamat@gmail.com>