[postgis-devel] Clarifying support of standards and specs

Mateusz Loskot mateusz at loskot.net
Wed Dec 11 05:54:43 PST 2013

On 10 December 2013 22:41, Paul Ramsey <pramsey at cleverelephant.ca> wrote:
> The docs are just really old and haven’t be substantially touched in this
> area for ages, certainly not much since 2.0 when the wkb/wkt code got
> entirely replaced. So what you’re seeing is a need for a rewrite.

Got it, I have opened a ticket to keep the task alive


> (1) “superset” means more than what’s in the early SFSQL standard, so the
> things in the ISO spec, and then things not even in that spec yet, like
> TRIANGLE and TIN and POLYHEDRAL SURFACE (though I guess now the SFSQL
> standard has mostly caught us up, object-wise).
> (2) the old SFSQL version that had no 3d (1.1 from 1999)
> (3) sentence probably dates from 1.5 or earlier, when the Z/M tags were not
> supported on WKT
> (4) yep, Q&A probably post-dates the other sentence.

Thanks for clarifying this, hopefully it will make it into docs.

> Since the standards are themselves a hodgepodge, I think we should not worry
> so much about explicitly referencing chapter and verse about what standard
> each function or object or representation derives from, but rather be more
> clear about what the actual Current State of representations is, and scrub
> the docs a little more carefully for old information.

I agree.

If the team attempts such clean-up, I will try to help, hence the ticket above.

Sandro's old comment [1] resound in my head:

"We can't be compatible to MULTIPLE standards at the same time."

[1] http://blog.cleverelephant.ca/2010/03/nothing-nada-zip-bupkus.html

Best regards,
Mateusz  Łoskot, http://mateusz.loskot.net

More information about the postgis-devel mailing list