DSSS, Dsource, and cpan

[janderson]

Bill Baxter wrote:
> janderson wrote:
>> Walter Bright wrote:
>>> Anders F Björklund wrote:
>>>> Walter Bright wrote:
>>>>> You're right in the case of non-libraries. But I think the ddoc 
>>>>> requirement should apply to libraries.
>>>>
>>>> Not all projects use ddoc, some use for instance Doxygen instead...
>>>
>>> Rather than get into an argument about what makes a project certified 
>>> or not, a 'digg' style voting system might be better:
>>>
>>> 1) each registered user can vote +1, 0, or -1 on each project. The 
>>> project then gets a score, which is a simple sum of these.
>>>
>>> 2) a user can change their vote on a project at any time.
>>>
>>> 3) votes that are over 1 year old get expired, and set to 0.
>>>
>>> One potential difficulty with this is someone registering a boatload 
>>> of accounts in order to shill a particular project. This can be 
>>> mitigated if people can, sliced by project and user, see who voted 
>>> for what.
>>
>> I'm not sure about the whole negative re-enforcement.  You may find it 
>> scare contributions away.
> 
> Matlab has a big user contribution site.  Each project or code snippet 
> gets its own page, people can rate the projects, and people can leave 
> comments.  The ratings are "stars" so it doesn't really seem so 
> negative.  Even if you have the worst rating you still have a "1 star", 
> which sounds better than a minus anything.  In practice lame projects 
> just don't get reviewed because no one cares.  Probably the only reason 
> you'd get a bad review is if you promised the world and in the end your 
> program just simply didn't work at all and thereby ended up wasting 
> people's time.
> 
> Here's one page from there.
> http://www.mathworks.com/matlabcentral/fileexchange/loadCategory.do?objectType=category&objectId=131&objectName=Filtering 
> 
> 
> 
> If dsource added the *'s rating, and a way to search for projects based 
> on keywords, I think it would be pretty much there.  The download count 
> on the matlab pages are kind of helpful as a guide too, but I don't know 
> how you easy that would be to do given open access svn servers.
> 
> --bb

Sounds great!

-Joel