Originally Posted by sjgadsby

I've been thinking on this, and isn't this the purpose of karma weighting?

[snip]
[*]the existing karma weighting system is broken and should be reworked prior to the time when the committee begins making sponsorship decisions

Originally Posted by dneary

Hi,



I don't mind a community engagement talk, but it would have to be framed right.

I don't want a presentation that can be summarised as:

• Here's all the stuff Nokia's doing wrong
• Here's all the stuff the community should be doing
• Here's how to make everything better

A good community engagement presentation for me grows out of action rather than reflection. Allow me to explain.

My 'Documentation co-creation" session will talk about the successes we've had with installing MediaWiki and using it pretty effectively, integrating tmo content into wiki.maemo.org, proof-reading official documentation like the HIG, and bringing Nokia produced documentation to a point where it will be in the wiki for Fremantle, and has been available as early pre-releases for quite some time.

I won't be talking about how life is all rosy, but I will be saying "we have the tools to make a difference, we have the people to make a difference, we have been making a difference, now let's hit the home run".

If there were an area where, through your work, you have perceptively improved the Maemo community, and you believe that there are (a) lessons to be learned from that and (b) more work to be done that you want to "sell" to people, that would be an interesting talk.

Cheers,
Dave.

Originally Posted by Texrat

I think the karma system itself is broken. The "thanks" count is not accurate in my case, unless there's a qualifier I'm missing...

Originally Posted by sjgadsby

I've been thinking on this, and isn't this the purpose of karma weighting?

Given that:

• karma exists specifically for assisting in summit sponsorship and device program decisions

-and-

• the karma weighting system exists so that karma reflects the relative value of various contributions to community

then either:

• karma scores should be used as-is by the sponsorship committee

-or-

• the existing karma weighting system is broken and should be reworked prior to the time when the committee begins making sponsorship decisions

Originally Posted by VDVsx

What I meant to say is that karma is very easy to obtain...

As a example you can jump from zero karma to 1904, only adding comments to the existent applications in downloads.

Originally Posted by Texrat

Other than discarding the download comment karma calc outright, fraud could be detected by checking time stamps (one method). If people are running down the list in an reasonably short time frame that would indicate abuse. One solution could be to add another condition into the algorithm that adds value to the time delta between comments, but I see that as very difficult to implement.

Originally Posted by VDVsx

And can be also unfair, people are free to comment on 3 or more apps in the same day.