[postgis-devel] [WKT Raster] Documentation on the WKT Rastertype

[Mateusz Loskot]

Pierre Racine wrote:
 > >In my personal opinion, that's the purpose of Wiki, to brainstorm,
 > >discuss and collaborate in editing.
 >
 > But if I put it in the wiki then you won't be able to pin-point
 > sentences with comments. Only changes. I was thinking you could first
 > comment, then I would made some changes myself and then put them in
 > the wiki.

OK, I got your point.

 > The wiki will be the official documentation so it is not a place to
 > comment. Only to make changes. But I can put it in the wiki if you
 > prefer. (Unless you are religiously reluctant to Word documents.)

I do have access to Microsoft Office.

 > >By the way, after the recent re-(loss-of)-structure of
 > >specifications, I've clearly got lost in between N versions of the
 > >same documents dangling on the Wiki.
 > >I don't have time to copy and paste, for instance gdal2wktraster
 > >usage, between specification 01, 02, 03, etc.
 >
 > There is only two documents now: Working spec and final spec. You
 > should copy the text in the final spec only once when everybody is
 > happy with the working spec. You certainly do not have to copy
 > everything to the next version specs.

In practice, there are as many documents as 3 for 01 version with IMO 
same or very similar content

 > We agreed on all this when we moved the wiki. I don't see why we are
 > not happy today.

Sometimes you can observe a program failure when you run it.

 > I understand that this is not the way GDAL used to work but there are
 > other ways to work. I think this way makes specs available to
 > everybody (not only to those who have the code and can commit) and it
 > ensure that we keep track (history) of every addition and change.

If the docs structure work for you and others, that's perfectly OK and 
let's leave it as is.
It doesn't work just for me, so I'll stay focused on coding.

 > But we can change all this if you wish.

I don't express any wish, but proposal to IMO improve things.

Best regards,
-- 
Mateusz Loskot, http://mateusz.loskot.net