[geeklog-devel] Spam-X project (was: GSoC 2010 is on)

[Dirk Haun]

Coming back to this:

>I think the main problem with the Spam-X ideas is that they are not very
>attractive as a project. It's a collection of separate ideas with "has
>something to do with spam" as the only common theme.

We should also not fall into the trap hoping that GSoC will somehow
magically solve all our problems. While a newcomer (e.g. a student) can
provide some fresh and welcome input, there are some issues that we
simply have to address ourselves. And most of the things listed on the
"Spam-X overhaul" ideas page fall into that category, I think.

So I'm going to move some of the items from that page over to the
bugtracker. In the meantime, how about this mix for a Spam-X GSoC project?

- implement SWOT
- implement some new modules for external services

There is a bunch of services similar to SLV out there that we could also
use. Some require registration, some don't. So we could give our users
some alternatives to SLV.

I'm counting 4 or 5 such services. Implementing a module should require
a week max./each. Throw in some tweaks to the admin UI to allow the site
admin to switch between services easily (since you probably don't want
to run them all in parallel) and that could roughly cover the first half
of GSoC.

Part 2 would then be to implement SWOT. Which is pretty much yet another
anti-spam service, if you think about it.

Implementing the modules for the other services would give a student an
idea how things work in Geeklog. It would expose them to some ideas how
such a service works, which may (or may not) come in handy when
implementing SWOT later on. And they would easily get something working,
which would give them a sense of achievement and - hopefully - some
motivation for the more challenging half of the project.

Overall, it's probably still an "easy" project. But it's a little more
coherent than last year's project, I would hope.

Does that make sense?

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/