Show HN: we open sourced our startup's help desk as django-knowledge.

[bryanh]

We had a lot of reasons we wanted to build our own help desk, the least of which (and easiest to mention) is that a three seat license to the top tier desk systems are well over $100/mo. Plus they were slowing us down quite a bit.

Right now its only dependency is Django, it supports email alert for updates, has bundled style, a permission like public/private/internal status, optional anonymous questions/responses, categories, searching, and a few more I've missed.

What I'd like to do is: extend tests, class based views, parse incoming emails as questions & response (this is important for help desks), better mod tools and internationalization.

Anyways, see the GitHub repo: https://github.com/zapier/django-knowledge, the docs: http://django-knowledge.readthedocs.org/en/latest/index.html or check out a live demo at Zapier: https://zapier.com/support/

We've also registered django-knowledge.org. If anyone has thoughts on a good, cheap lightweight host, let me know.

[bryanh]

Some clickables. GitHub repo: https://github.com/zapier/django-knowledge, the docs: http://django-knowledge.readthedocs.org/en/latest/index.html or check out a live demo at Zapier: https://zapier.com/support/

[ABS]

looks pretty nice, will test it out as soon as I need a new one.

Your premise begs the question though: how much time of how many people was spent developing this? what was the opportunity cost of working on this rather than your stuartup core product? if you divide that amount by $100 how many months of the top tier desk systems could you have bought? :-)

[bryanh → ABS]

This is a valid question which is why I was kind of hesitant to mention the price... I knew someone would mention this. That said, we definitely talked it over and there were a lot of reasons for keeping it internal. Call it an investment.

1. Mainly, our service has all kinds of crazy nooks and crannies for errors, and being able to tie straight into the underlying model and storage layer between questions, users and errors will be vital. We currently spend hours everyday trying to find out which this or that created the error because we're using Desk.com (formerly Assistly.com).

2. We have a few strategies around marketing. Anyone familiar with patio11 will understand. This is a bit of a gamble.

3. We'd like to create something awesome in the eyes of developers. They're gonna be a key piece in our plans going forward. It makes sense to help them where we can.

There are more, but they are smaller reasons, like the $100's of dollars a month for a similar service. Plus, its fun to OS stuff. ;-)

[ABS → bryanh]

fair enough, thanks for clarifying. In fact you did write " the least of which" and I was curious to know the main ones.

ALl the best

[bryanh]

A few comments over at my x-post on Reddit's Python subreddit: http://www.reddit.com/r/Python/comments/q13gs/show_rpython_w...