Hacker News Reader

Comment by ridruejo | Hacker News Reader

OSQA is based on Django, which makes it somewhat harder to deploy. I know because we package it as a BitNami stack http://bitnami.org/stack/osqa and also use it for our Q/A site http://answers.bitnami.org Without actually having tested Coordino yet, I believe the fact that it is based on PHP should make it easier to adopt and extend. It will be interesting to see if it gets traction, there are plenty of StackOverflow clones that have been abandoned over time. An important factor in choosing OSQA was the fact that even though it was open source, it was backed by a company. A Q/A site, like forum software, is not something that you can move away from easily once you have content in it.

View on HN

Having just installed Coordino in under 5 minutes (from git clone to login) on my Mac workstation at the office, I can attest to the ease of deployment.

We've been looking for an in-office knowledge base solution but we're a bit lazy (more accurately swamped with client work) so this was a no-brainer to try when I saw it was PHP.

I find it's installation method, like most PHP programs, to be somewhat off-putting. Yeah, it's great for a development box--place in front of a web server and hit the browser to finish up--but it's a nightmare for automated deployments: source code and configuration are two different assets that _should_ exist in two different places on the filesystem. When they do it's reliable to update one or the other without issue, else it's a bit chancy for my tastes.

Yes I can attest to the ease of OSQA/Bitnami. It's what I used to create http://chicagofounders.com. Thanks for including OSQA on Bitnami, as it was a relatively simple deployment!

A Q/A site, like forum software, is not something that you can move away from easily once you have content in it

Why not? If it has a decent API it should be easy.

The problem is that you need to move the content, user accounts, ratings, etc. from one system to another. The existing system has to had an API to export that information in a format that the new system can understand. If you have that, then it is easy, the problem is most systems out there do not have that :) Situation is a bit better with extremely popular projects for blogs or forums, like Wordpress or phpBB, which have an ecosystem of exporter/importer tools.

in a format that the new system can understand

Or you can write your own converter - it probably takes ~30m to whip up something in Python.

depends on the system - 30 minutes is not enough to convert users, groups, permissions, threaded conversations, voting, keywords, etc, especially if both systems do not use the same data model.

Sorry, that's nonsense. Each one of the systems is going to have a different schema and constraints for users, posts, comments, votes, that most likely will not have a one-to-one mapping.

Something - yes. Something which implements everything required and is tested - not likely.

There's a huge difference between converting data from your little hobby and converting the vital knowledge of your company. It's different because it's more mission critical - you don't want to screw it up.

It's also different because the bigger dataset is likely to have some customizations and other "small" stuff which tends to matter in big perspective. You might be okay for losing e.g. votes if it's a small data set, but you really want it in with a big one.

Reply on news.ycombinator.com