Site Recommendation: Move the Forum to an Askbot Powered Q&A Site?

We won't be doing this. This would mean everyone having to create yet another login on one of those sites and having all of that information owned by the corporate entity behind those sites. There's no advantage and quite a few disadvantages to that.

Another alternative is GetSatisfaction.com. I know that Songbird and even PayPal operate their Q&A forums there. Are PortableApps or John Haller on Twitter or FB? People could participate there too -- you'd have a much BIGGER forum to work with, one billion strong and growing!

The status of Askbot is somewhat unclear; I've asked them for a clarification. It seems to be open source (http://pypi.python.org/pypi/askbot/, https://github.com/ASKBOT/askbot-devel/), but there is very little about it on their site. They seem recently to have focused on commercial developments (such that the root of askbot.org now redirects to askbot.com).

In that vein, there is also OSQA, which is properly open source. I've also seen it being successfully used by WebFaction for their community site and Atlassian for their answers site (and I have a very high regard for Atlassian in particular, for they make very good tools—if they select such a tool for much of their customer support, it's good enough to use, or they would have written their own), in each case initially as an experiment and after a while replacing their existing forum systems. OSQA is not particularly actively developed, however, with DZone working mainly on their own commercial development originally based on it, AnswerHub (formerly Qato). The last commit to their SVN repository was on December 6, a month ago. Still, Atlassian using it is a pretty forceful argument for me, based on the tools of theirs that I have used and how drastically superior they are to much of their competition. They've customised it a lot more than most OSQA deployers have, however, and have quite a bit more customisation coming. (Some of the most interesting is the design; take a look at their mockups to see how a Q&A site does not need to tightly resemble Stack Overflow.)

For a little bit of background, both Askbot and OSQA are originally based on an unmaintained Python/Django system named CNPROG. OSQA has been almost entirely rewritten, however, while Askbot holds more of its heritage.

In both of these, and any other options around, really, we see the risk of migrating to such a platform; none of them have a long history, and most of the options that have been developed have stalled; I think OSQA is in danger of stagnating (if it has not already) owing to not matching DZone's requirements (why they built AnswerHub); Askbot seems to be being taken commercial in its current form which, if successful, increases the probability of its long survival.

With either of these, I think it will be wise to wait a little and see how they react to Django 1.5 (RC1, due for release in a couple of weeks); its configurable user model in particular may change how each of these projects work in user management, quite a bit. In such a site, I would picture integrating it into the current user accounts system, which would avoid the need to create new user accounts. (This is quite possible in Django but the manner in which it is done may have interesting repercussions in the new user model. Being open source software, however, modifications can be made if necessary and should be generally beneficial to the projects concerned. I'm not sure how easy it would be to get it working in Askbot or OSQA, not having looked at their code.)

However, I don't think this Q&A model suits for certain things which take place on our forums, most notably the Beta Testing forum. Soliciting translations would be interesting; it doesn't really fit either model well (those who have tried to manage or participate in the process will know the difficulties in using the forums for it). I'm also afraid we could end up having more noise in a Q&A site; it would require very careful moderation early on to bend it in the right direction or else. At the least, I don't think replacing just the forums would be the right thing to do; rewriting more of the site in Django (especially automating many tedious administrative things in the management of apps) would be well worth while undertaking at the same time. A special interface for Beta Testing could then be made incidentally to the management of the app system (there would be little difference between them, as I view them, and that would provide much a smoother transition from beta testing to pre-release to release, and it would make a release team more feasible).

But then, you see, this is a much more complex project and would require quite some time to develop.

In favour of these solutions, I would expect a Python/Django-based system to be quite a bit more efficient than PHP/Drupal, especially if deployed with PyPy and careful caching, potentially reducing costs. (Quote me not on that, however. It's a dangerous opinion!)

Finally, remember: the Q&A model is no silver bullet, nor a panacea which will rapidly fix things. Any changes must be made very carefully and deliberately, evaluating benefits and costs. It is no small change which is being suggested.

(For the record, I am in favour of the Q&A model and have considered it for a year or so and am especially in favour of various other more drastic architectural changes for the site, as mentioned in an earlier paragraph of this post. But I myself cannot at present justify being much involved in such a project owing to the time required.)