[Q&A] Change or improve: who, how, what ?
I would like to brig up to this great post Rethinking the Q&A forum concept 2.0 by @StevenPuttemans adding my considerations (If you think that this should be continued on GitHub I can delete my question).
Knowledge Management is an open issue. Improve this Q&A is open too.
I think that a KM system should have
- A "scope of work" to define needed functions
- A team of people available for implementation, administration and maintenance
- A set of ICT tools selected to achieve the scope of work and based on Admin resource
In addiction a KM requires ongoing maintenance. If this is not done it will become quickly inadequate.
Consequently, above steps should consider size and availability of the Admin team, in special case of community KM.
Recent 2 weeks down of Q&A shows that the Admin team is already stretched, may be they can't do additional effort.
So I think that a back step would be useful
The "scope of work"
From FAQ
...basically a forum which is created to help people with their OpenCV implementation problems.
I would try to transpose points exposed by Steven and many other user into a "behind scene" scope of work:
Explicate tacit knowledge about OpenCV
This would define this forum as one of the tools of OpenCV Knowledge Management. A nice goal would be the convergence of information produced here into the OpenCV Doc/Tutorial, better if with a 2 side way.
The Admin team
How many people have rights to do things ? but much more relevant, have they any time slot free ?
The tools
Because it's easy to ask and easy to answer, users ask and often receive an answer with a right solution. Than AskBot achieves the task "to help people with their OpenCV implementation problems".
But AskBot is boring supporters because they often have to answer on same or similar subject. This is because, user are lazy, answers are not organized and search tools are few and poor. In a word Is wasteful.
To create some custom functions for AskBot could be an option if some Python/Django developers are available here.
Points exposed by Steven are great start, but, is the Admin team sized/timed enough ?
At the end...
May be a central place for suggestions should be defined. It could be as answers this post (or some elsewhere ) with:
- features request for this Q&A
- OR a new engine for this Q&A
People could vote each request such as a pool
After a bit of time the Admin team, based on their available time, would select and implement a subset of the suggestions.
@pklab, thanks for providing this, but I would love to move the discussion to github, update it there and work closely together through the provided chat link!
I will add your comments to the list! Lets keep this open and let others post replys if they want to!
I agree that the forum needs a good reorganization.
First problem with OpenCV is that the official docs are really bad. The 2.x docs were barely useable, but the 3.x are totally useless. Just try to implement video stabilisation based on this page: http://docs.opencv.org/3.1.0/d5/d50/g... The samples are also over-complicated sometimes (videostab.cpp has 562 lines, when it can be explained in 15 lines).
The forum is also too big, with hundreds of questions every month, and the search function is also very basic (searching "videostab" gives only 3 connected questions on the first page, but no real solution). There are also a lot of unanswered questions which won't help.
I agree about official docs usability regression and I'm not alone see this and this. also referring from this Q&A to docs isn't really well designed see here but may be this is out of our scope now.