Implementation of topology-network approach

[Pierre HARDY]

In addition to my previous email (My apologies for double post)

Point 2 : An alternative would be to bring a small addon to the current
topology-geometry implemtentation and provide a way for the user to disable
node creation when 2 lines intersects or 1 line self-intersects... This
could be an answer to to most uses cases I know, but as long as l know this
is not promoted by any ISO Standard.

I do not know if such implementation would be accepted as part of a postgis
topology extension. What do you think ?

Le mer. 20 nov. 2024, 22:02, Pierre HARDY <pierre.hardy5 at gmail.com> a
écrit :

> Hello Felipe,
>
> I am highly interested in implementing such feature for an application in
> hydraulics network modelling (check EPANET) some implementations of the
> topology-network logic in postgis contrib packages would be nice to amswer
> this need.
>
> To answer your points:
>
> 1. I agree, but PgRouting does not provide editing functions, this is a
> high limitation and make PgRouting not suitable for easy and reliable
> topology edition from a desktop client such as QGIS  without implementing
> some additional logic outside.
>
> 2. After reading the standard I agree with you, and suggest that ST_
> function described by the standard must be implemented as a first step. In
> a second step we should plan to combine them in more "smart"
> functions/logic that can handle real user needs for edition from a desktop
> client.
>
> 3. For a first implementation, a 2D support is sufficient to the
> application I knows in water supply modelling.
>
> Le mer. 20 nov. 2024, 02:24, Felipe Matas <felipematas at yahoo.com> a
> écrit :
>
>> Ufff, I also wanted this time ago, and still want it! is just there is
>> some things to take care.
>>
>> 1. PGRouting already has Topologies, they do not follow the SQL/MM
>> standard, also do not have the functions for a user be able to use it
>> 2. The standard is well defined, is just actually do not define anything
>> else, functions that take advantage of a Network Topology to make them
>> useful, would be great be able to at least list functions for that, the
>> standard is like SF without any spatial operation.
>> 3. Personally, I think in this case 2D Points are a very hard constrains,
>> one very useful case for a Network Topology are roads, but in a lot of
>> places int he world, to have a valid Network Topology you need 3D Points,
>> this causes to one of the most useful uses would be impossible.
>>
>> From my perspective the ideal is have a Standard v2, and be implemented
>> on PGRouting instead of Postgis.
>>
>> Thx!
>>
>> El martes, 19 de noviembre de 2024, 11:23:04 a. m. GMT-3, Pierre HARDY <
>> pierre.hardy5 at gmail.com> escribió:
>>
>>
>> Hello PostGIS devs,
>>
>> It seems that "postgis_topology" extension does not support
>> topology-network approach.
>> Here are my sources:
>>
>>    1. What is topology-network:
>>    https://www.gaia-gis.it/fossil/libspatialite/wiki?name=topo-intro
>>    2. Source no1 indicating that Postgis does not support
>>    topology-network approach:
>>    https://gis.stackexchange.com/questions/122636/prevent-postgis-topology-from-splitting-edges-on-line-insert
>>    3. Source no2 indicating that Postgis does not support
>>    topology-network approach:
>>    https://terragis.net/docs/other/MASTERING_POSTGIS.pdf => p253
>>
>> "Before we start working with topology, we need to discuss the key
>> concepts behind the PostGIS topological data model. PostGIS topology
>> implementation is based on ISO standard 13249 - Information technology -
>> Database languages - SQL multimedia and application packages - Part 3:
>> Spatial. The standard name is often abbreviated as ISO SQL/MM. This
>> standard defines two data models for topology: TopoGeometry and
>> TopoNetwork; PostGIS implements only the former."
>>
>>
>> I could verify this by testing on a basic local setup with QGIS and *Postgis
>> version 3.3*.
>> I guess this feature was not added since.
>>
>> I did not found any issues regarding this missing feature.
>>
>> Is an issue already openned to request this new feature or should I
>> create one ?
>>
>> Best regards,
>> --
>> Pierre HARDY
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-devel/attachments/20241120/7be80bc9/attachment.htm>