[pygeoapi] [PSC vote] AI/LLM contribution statement
Tom Kralidis
tomkralidis at gmail.com
Sun Oct 18 10:34:24 PDT 2026
On Sun, Oct 18, 2026 at 12:54 PM Francesco Bartoli <xbartolone at gmail.com> wrote:
>
> thanks for the agree, my only flag is that the message might be read only in the direction of humans while integration and interoperability are foundational for system to system communication where AI already applies.
>
> We know HTML support is just recommended if I’m not wrong and given mostly for web search engines.
>
OGC APIs do not mandate any encoding per se. GeoJSON, XML, etc. are
recommended just the same as HTML. Providing HTML is not only for
search engines, but useful user interfaces can be constructed just the
same to address UI/UX and user requirements.
> Wrt human factor of the development: I like an human-centric usage of the AI aiming at improving the quality of the software I can produce.
> I know this is an hot and debatable subject (intentionally used I) but I feel to be humble that I can learn faster to write better code.
> That’s how I can read the first half and the most relevant for me, By Humans. The second half is less indeed IMHO.
>
Fair enough. I would echo Joana's remarks [1], but fully respect your
opinion just the same my friend.
..Tom
[1] https://lists.osgeo.org/pipermail/pygeoapi/2026-October/000527.html
> My 2 cents
>
> Francesco
> On 18 Oct 2026 at 17:42 +0200, Tom Kralidis <tomkralidis at gmail.com>, wrote:
>
> Thanks Francesco for your vote. Agreed, an API can be used by a
> machine, but it can also be used by a human, especially in the context
> of the current OGC API efforts with more focus of being "of the Web"
> (read HTML support). More importantly, how a given software is
> developed is also a factor. I like that BHFH clearly states "humans
> in the driver seat" :)
>
> Thanks
>
> ..Tom
>
>
> On Sat, Oct 17, 2026 at 5:23 PM Francesco Bartoli via pygeoapi
> <pygeoapi at lists.osgeo.org> wrote:
>
>
> I like the message so I leave my +1.
>
> However, I feel the statement for pygeoapi is a bit weird since an API in principle is more for machine than humans
>
> Francesco
> On 14 Oct 2026 at 16:00 +0200, Tom Kralidis via pygeoapi <pygeoapi at lists.osgeo.org>, wrote:
>
>
> Dear pygeoapi PSC: per subject, a PR has been put forth in [1] to
> update/clarify the project's position on AI/LLM based contributions.
>
> Note that "By Humans, For Humans" is also being put forth for other
> Geopython projects (pycsw, OWSLib, GeoHealthCheck, pygeometa, etc.).
>
> Please feel free to vote either here on the mailing list or in [1].
>
> Starting with my +1.
>
> Thanks
>
> ..Tom
>
> [1] https://github.com/geopython/pygeoapi/pull/2441
> _______________________________________________
> pygeoapi mailing list
> pygeoapi at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/pygeoapi
>
>
> _______________________________________________
> pygeoapi mailing list
> pygeoapi at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/pygeoapi
More information about the pygeoapi
mailing list