[postgis-devel] Docbook ID's and <a name> links

[Kevin Neufeld]

Obe, Regina wrote:
> I just moved over ST_MakeLine.  I realized that since we are going with
> meaningful names, probably makes sense to have ST_MakeLine variants
> appear on the same page.  So I made as one entry instead of 2.  Does
> that look right?
> 
> I figure that's the most complicated case we would need to deal with.

It looks good to me.  I like your examples you've been interjecting 
throughout.  It adds to clarify to the methods which I think is the 
whole point of this exercise.  After we migrate all the functions over, 
I would like to additionally add a few pictures to help visual the 
examples.  One thought - I don't think either prototype though uses an 
automatic index call, however.  We would probably remove that note.

> 
> I'm also wondering if there is a way in the TOC to pull out the things
> that take 
> 
> Geometry set  
> 
> as argument.
> 
> I've always found it useful to see a listing of all aggregate functions
> together, but its also useful to see them in the specific category area.
> 
> Thanks,
> Regina  
> 

Do you mean separating the ST_MakeLine prototypes you just merged 
together or having a second TOC set linking to the same function pages?

I believe DocBook does allow us to *manually* create an additional TOC 
section that's not automatically generated so we can group methods 
together as we see fit, (like aggregates, geometry returning, boolean 
returning, etc).  Of course, this will again duplicate the function 
names on the main TOC page and will require a higher degree of manual 
intervention.

Cheers,
-- Kevin