[Mailman-Developers] Searchable archives for MM

[Barry Warsaw]

On Aug 9, 2009, at 9:31 PM, Jordan Hayes wrote:

> Um, you need Phython in this equation exactly why?
>
> The archives are web pages.  Who cares what language was used to  
> implement the search?
>
> You don't demand a Python web server to deliver your archives, do you?

Heh, well Mailman 3 will include a couple of Python web servers.   
There will definitely be a web server for the REST interface and the  
web ui will likely be vended from a Python web server.  They're super  
easy to write!

Note that we've adopted WSGI for all web protocols so the ultimate  
port 80 listener can be almost anything.  This will make it easy to  
integrate with frameworks such as Zope, Django, Turbogears, etc. or  
even Apache through mod_wsgi.

I'm going to need really strong arguments to bundle anything not  
written in Python.

>> I was using the debian archive search only as an example for
>> what can be achieved on big archives.
>
> For me, I think the people who Think Hard about searching text can  
> work in whatever language they want; SWISH-E has done a good job of  
> providing a platform for search, and it's easily integrated with  
> Mailman.  FWIW, I have lots of "big archives" ...

The way to think about this then is modularization and pluggability.   
The archiver itself is already pluggable (with more well defined  
interfaces in Mailman 3).  The question is whether for Pipermail, the  
search itself should be pluggable.  Think about how that would work.

-Barry

-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 832 bytes
Desc: This is a digitally signed message part
URL: <http://mail.python.org/pipermail/mailman-developers/attachments/20090809/11d85dc9/attachment-0001.pgp>