[gdal-dev] CSharp bindings queued for removal (was Re: GDAL CSharp bindings maintainers/contributors listening... ?)

Fri Jan 31 04:09:39 PST 2025

The problem about "moving" the bindings is going to be that all of the SWIG
definitions for all bindings are intimately wound together. It would be a
lot of work for little gain.

The danger of the "surveys and feedback" driven approach, which I felt like
saying at the time of the last survey but in the end thought it would just
be ignored, is that the approach, although good and important, is
intrinsically biased towards certain channels.

The people who use the CLI are going to be massively over represented -
since I am sure they all know what GDAL is and use the website.

However, there are vast swathes of end users of GDAL who have absolutely no
idea what GDAL is.

- the vast majority of QGIS users, the archeologists, the data scientists
etc etc have no idea what GDAL is,
- I am absolutely sure that none of the users of Google maps knows that (I
am sure I heard anecdotally) GDAL is important to the data ingestion for
maps,
- I know that very few of the users in R are aware of GDAL - even when they
refer to rgdal in their scripts (it is just a name).

And just so, since the Java and C# bindings (and to a lesser extent the
Python bindings in Fiona and Rasterio) are bundled to packages (e.g Nuget,
npm and UPM) that are (in some cases) bundled into other packages that are
bundled in to applications that are distributed - I am absolutely sure that
no one does or can know how many end users are using GDAL through these
channels but I would be willing to bet that it is orders of magnitude
higher than you think it is.

The danger is, of course, that if you based strategy purely on the latest
survey then you would say that GDAL should ONLY be a CLI.

If you based strategy on what we know about end user numbers, you would
probably say that GDAL should be a dedicated module of QGIS and the CLI
should be deprecated!

Or possibly a dedicated batch back end for Google maps.

NONE of these possibilities would be a valid response to our lack of
knowledge.

On Thu, 30 Jan 2025 at 18:00, Michael Smith via gdal-dev <
gdal-dev at lists.osgeo.org> wrote:

> Maybe the bindings should be moved out if gdal to a separate repo so those
> interested in the bindings can monitor that specific repo and not gdal in
> general.
>
>
>
> Mike
>
>
>
>
>
> --
>
> Michael Smith
>
> Remote Sensing/GIS Center
>
> US Army Corps of Engineers
>
>
>
>
>
> *From: *gdal-dev <gdal-dev-bounces at lists.osgeo.org> on behalf of Tamas
> Szekeres via gdal-dev <gdal-dev at lists.osgeo.org>
> *Reply-To: *Tamas Szekeres <szekerest at gmail.com>
> *Date: *Thursday, January 30, 2025 at 12:41 PM
> *To: *Even Rouault <even.rouault at spatialys.com>
> *Cc: *"gdal-dev at lists.osgeo.org" <gdal-dev at lists.osgeo.org>
> *Subject: *Re: [gdal-dev] CSharp bindings queued for removal (was Re:
> GDAL CSharp bindings maintainers/contributors listening... ?)
>
>
>
> Even,
>
>
>
> The silence is not necessary because we neglect the C# bindings, just
> because we are doing our daily work (using the bindings).
>
> I'm personally maintaining the GDAL <https://www.nuget.org/packages/GDAL>
> and it's related nuget packages based on these bindings for several years
> and it has 658.1K downloads overall and 147 downloads per day in average.
> The packages are used by projects such as ArduPilot/MissionPlanner,
> DotSpatial, SharpMap which are also widely used projects, so we can
> consider the importance sigificant, and such change (removing the bindings
> entirely) would involve general dissatisfaction against GDAL I think.
>
> In my opinion, we should also distinguish whether an issue reported
> against the binding is a blocker or just a nice to have change. In my
> understanding the implementation of the UTF-8 to unicode conversion at all
> relevant methods would be beneficial, but not something which would block
> the overall usage (it has already been done for the most important parts I
> think). If the required functionality is really important to someone, I
> think he/she will spend some time to contribute it or pay someone to make
> it happen.
>
>
>
> I also think that the gdal-dev list is not proactively monitored by the
> GDAL users (or by the indirect users) well enough, so that we can make such
> decisions to remove a binding or a driver nased on that. Whe should
> probably find some alternative ways (like surveys or feedback forms at the
> site or gather some other statistics like doc page loads and downloads) by
> which the community can express their interest against a specific component
> and would be sufficient to detect disinterests.
>
>
>
> In any case, I'll review and will be trying to to sort out some open
> issues against csharp in the next few weeks.
>
>
>
> Best regards,
>
>
>
> Tamas
>
>
>
>
>
> Even Rouault via gdal-dev <gdal-dev at lists.osgeo.org> ezt írta (időpont:
> 2025. jan. 29., Sze, 5:33):
>
> Hearing silence, the only logical conclusion is that there is no interest
> ==> https://github.com/OSGeo/gdal/pull/11746
>
>
>
> Le 27/09/2024 à 20:15, Even Rouault via gdal-dev a écrit :
>
> Hi,
>
> This is your regular remainder that nobody in the core GDAL maintainer
> team is a CSharp aficionado (we don't have a personal grief against it,
> just that we are blatantly ignorant, at least speaking for myself !), so
> related tickets about it will definitely result in no action.
>
> See
> https://github.com/OSGeo/gdal/issues?q=is%3Aissue+is%3Aopen+label%3A%22csharp+bindings%22
>
> The current trend is that people seem to be annoyed by UTF-8 related
> issues. The issue is likely that our methods that accept or take a const
> char* in the SWIG bindings should use a specific typemap to map to CSharp
> Unicode strings instead of the "C" one or whatever those concepts are
> called in CSharp. There is an existing "utf8_path" typemap that is used in
> method that accept filenames, that should probably be renamed to
> utf8_string and be used more extensively. And probably with a version of
> the methods to also return a raw C string / bytearray in the cases where
> drivers don't know the encoding and might return "random" stuff. Cf pull
> request #10652 where I did something for Java that suffered from that later
> issue.
>
> Happy pull requests!
>
> Even
>
> --
>
> http://www.spatialys.com
>
> My software is free, but my time generally not.
>
>
>
> _______________________________________________
>
> gdal-dev mailing list
>
> gdal-dev at lists.osgeo.org
>
> https://lists.osgeo.org/mailman/listinfo/gdal-dev
>
> --
>
> http://www.spatialys.com
>
> My software is free, but my time generally not.
>
> Grumpy maintainer.
>
> "De l'égo à l'égoût, il n'y a qu'une bouche mal refermée", André Isaac
>
> _______________________________________________
> gdal-dev mailing list
> gdal-dev at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/gdal-dev
>
> _______________________________________________ gdal-dev mailing list
> gdal-dev at lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/gdal-dev
> _______________________________________________
> gdal-dev mailing list
> gdal-dev at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/gdal-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20250131/eff549eb/attachment-0001.htm>