RFC 11: Curved Labels

[Steve Lime]

>>>> Daniel Morissette <dmorissette at DMSOLUTIONS.CA> 02/16/06 1:12 PM >>>
> Steve Lime wrote:
>> Hi all: You may or may not know that a patch for curved label support just 
>> mysteriously appeared one morning a bit ago in bugzilla. Output is pretty 
>> impressive. I worked with the author on a few integration points (collision 
>> avoidance being the biggie) and an patched patch was submitted in the last 
>> week. (Thanks Benj!)
>> 
>> Anyway, we wrote up a quick RFC for your consideration. See:
>> 
>>   http://mapserver.gis.umn.edu/development/rfc/ms-rfc-11
>> 

> I'm ready to say +1, but just a quick comment about "ANGLE FOLLOW" in 
> the mapfile... follow what? Should we use "ANGLE FOLLOWPATH" instead?

I still like FOLLOW for a couple of reasons:

  1) what else could it mean? (PATH is redundant)
  2) I like to keep the keyword set small and FOLLOW might have uses else where in the future

> Also, what if we ever implement more advanced or more expensive 
> algorithms later, for instance options to try different label positions 
> in case of collisions, that would be more computationally expensive and 
> we might not want to enable that by default so we'd need more mapfile 
> options to control that. As we define a name for this new feature in the 
> mapfile, should we plan for a way to name multiple algorithms or options 
> in the future?

It's hard to envision what those might be. I think we need to be able to 
play with this new code and then tweak things if necessary. It's only the
development version so there is time. 

I think for positioning we have enough parameters at the moment.

> I guess I'm just thinking outloud... "ANGLE FOLLOW" (or FOLLOWPATH) 
> could be fine for now and we can deprecate it later if we add more 
> advanced options. At that point we will better know which kind of 
> options we would need anyway.

> Daniel

Steve