Documentation of "experimental" features was: Re: [mapserver-users] Rendering bridges - SOLVED

[Stephen Woodbridge]

Jan Hartmann wrote:
> 
> 
> On 4-2-2010 12:04, tommasodb wrote:
>> The trick was SYMBOL SIMPLE, which is not really documented on the 
>> official docu (http://mapserver.org/mapfile/symbol.html):
>> /lists.osgeo.org/mailman/listinfo/mapserver-users
> 
> Neither is DATARANGE and friends, at least not in the official MapFile 
> reference. You can find it it http://mapserver.org/input/raster.html, 
> but shouldn't these keywords be included in the MapFile section too?

Hi Jan,

Regarding the DATARANGE documentation ...

I do not remember the full discussion that we had on thess keywords at 
the time, but I believe that they were considered "experimental" and 
subject to change or removal.

The issue for the PSC is that sometimes code is in the trunk and gets 
released, but we have not agreed on the RFC and the design of the code 
and features. We do not want a bunch of users jumping on some feature 
that is not ready for prime-time only to cause them pain later when we 
have to change it.

I think that there might be a case for documenting stuff like this and 
labeling it as "experimental" and therefore subject to change or 
removal. I think that this would have value in:

1) giving it more visibility which would lead to more input from users
2) providing better overall documentation
3) alerting users to its state as "experimental" which does not happen now

I've cc'd the dev list on this to start this particular discussion there.

Best regards,
   -Steve W