[mod_python] PYC download prompt

Jim Gallacher wrote:

> Sean Jamieson wrote:
>
>> Jim Gallacher wrote:
>>
>>> Sean Jamieson wrote:
>>>
>>>> Graham Dumpleton wrote:
>>>>
>>>>> Sean Jamieson wrote ..
>>>>>  
>>>>>
>>>>>> Has anyone else run into the problem, that when there is some 
>>>>>> unknown error with mod_python it spits out the .pyc version of 
>>>>>> your source code?
>>>>>> either that or gives a File Not Found (rather than a useful error)?
>>>>>>
>>>>>> I have a source that is working fine on my laptop:
>>>>>> Apache/2.0.54 (Ubuntu) mod_python/3.1.3 Python/2.4.2
>>>>>> and my dev server has the same setup...
>>>>>>
>>>>>> What happend when I copied the source to my dev server was:
>>>>>> 1) I got a Not Found error
>>>>>> 2) I refreshed
>>>>>> 3) I was given the .pyc for the file requested
>>>>>>
>>>>>> again, this works fine on my laptop which is using the same 
>>>>>> OS/apache/python/mod_python
>>>>>> there maybe some minor differences in apache config, but nothing 
>>>>>> that I
>>>>>> know of which would affect mod_python.
>>>>>>
>>>>>> Can anyone give me a clue as to where I should try looking?
>>>>>>   
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> One can get wierd things like this happening if you have MultiViews
>>>>> enabled and are using URLs which have no extensions.
>>>>>
>>>>> So for starters try using:
>>>>>
>>>>>  Options -MultiViews
>>>>>
>>>>> in your Apache configuration and see if things change.
>>>>>
>>>>> Also, what handler are you using, your own, mod_python.publisher
>>>>> or something else? It is possible that you might be hitting one of
>>>>> the issues since fixed in mod_python 3.2.
>>>>>
>>>>> Graham
>>>>>
>>>>>  
>>>>>
>>>> Puiblisher, and yes, I realized that the .pyc being sent was a 
>>>> result of MultiViews later on. My main issue now, is that my 
>>>> codebase which runs fine on my laptop, does not run on my 
>>>> development server; both are running the same software.
>>>>
>>>> I've been trying to find some key difference in apache's 
>>>> configuration that may cause it, but alas, no matter how much I 
>>>> copy it doesn't work.
>>>>
>>>> This is rather serious, as I have a client waiting to see this 
>>>> thing, this afternoon.
>>>
>>>
>>>
>>>
>>> Is the server publicly accessible? If so I can see what I get from 
>>> here. If you don't want the URL on the mailing list you can email me 
>>> privately.
>>>
>>> Jim
>>
>>
>>
>> currently there is nothing to get, now, when one tries to go to the 
>> URL, the browser does nothing, literally. i.e. if you're on the 
>> directory index page, and you click one of the .py files, or a 
>> directory with an index.py, it does nothing, if I open a new tab, and 
>> paste the URL, hit enter or click 'go', it does nothing, nothing 
>> loads, it still says (Untitled) as if it went nowhere.
>>
>> and this seems to only happen when I try to import something from my 
>> codebase,
>> a plain .py file that simply has the following defined, works:
>>
>> def index( req ):
>>    return req.filename
>>
>> I just don't get what's going on here, it doesn't make the least bit 
>> of sense.
>
>
> I'll agree with you there.
>
> I'm sure this is stating the obvious but is there anything in the 
> error log? Have you tried logging a message to the error log? 
> (req.log_error("It's alive!"))
>
> Have you taken a look at the page source for the page returned? I've 
> been tricked by this before where I haven't set the content_type and 
> returned some string representation of an object which the browser 
> interprets as an html tag. eg <thingy.main.thing object at 0xb7d2670c>.
>
> Jim

no. I don't think you get what I mean, the page does not transition to 
any other page.
i.e. 1) I and viewing the Directory Index
      2) I click the test.py that imports something from my project
      3) firefox's loading icon spins for a second
      4) I am still seeing the Directory Index

apache2/error.log is showing more than a few segfaults...