[postgis-devel] PSC Vote: Make spatial_ref_sys as a view

[Paul Ramsey]

> On Feb 8, 2022, at 9:58 AM, Regina Obe <lr at pcorp.us> wrote:
> 
>> Anyways, I'm a -0 shading into a negative one unless you trot out some
> users
>> who will worship this change.
>> 
>> P.
>> 
> [Regina Obe] 
> Paul,
> 
> If you do decide to give a -1 for this, please come up with an alternative
> plan for fixing this bug moving forward
> and making sure it stays fixed.
> 
> My feeling is spatial projections are more influx these days, so this will
> become more of an issue moving forward.
> 
> https://trac.osgeo.org/postgis/ticket/5024

I'm less sure of that, and I feel like more knobs and whistles in this area of the project will just result in more issues and confusion. I have sketched out an alternate approach that addresses the issue raised on the mailing list in a straight-ahead and blunt-force manner.

https://trac.osgeo.org/postgis/ticket/5024#comment:22

Does it do everything the piles of view machinery and so on does? no, it does not, but that's really the point. It's very easy to reason about the state of things and explain what user behaviour will result in desired results and what will result in undesired results.


P


> 
> What I liked about strk's solution is that it provides a clear audit trail
> of
> 
> 1) Is the user's entries they are complaining about different from what we
> published.
> 2) Have we changed entries from one upgrade to another.
> 3) It lets the user know they have entries different from what we shipped by
> checking 
> and a path for a user to ignore or purge their custom ones to be inline.  
> 
> SELECT postgis_full_version();
> 
> 4) It provides a path forward to what Darafei and Raul have been complaining
> about that spatial_ref_sys table is not a flexible enough way of dealing
> with this.
> 
> 5) In some cases if a user is using old entries and indexes based on that,
> they might not want to use the "more correct ones" as it screws up there
> workflow.
> I have some clients that fit into that boat.
> 
> Thanks,
> Regina
> 
> 
> 
> 
> 
> 
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/postgis-devel