[gdal-dev] Extended tab/mrr driver questions

Nyall Dawson nyall.dawson at gmail.com
Tue Dec 10 15:28:39 PST 2019


On Tue, 10 Dec 2019 at 22:13, Even Rouault <even.rouault at spatialys.com> wrote:

> For everyone understanding, I had preliminary contacts with them beginning of
> 2018 about the MapInfo Extended part, when they tried to know how to get it
> into GDAL. At that time, I did suggest to them that extending MITAB would be
> the most natural thing to do, which would make MapInfo Extended "immediately"
> available to everyone using GDAL. At that time, they already floating around
> the idea of using parts of their code base. I told them this would be far less
> practical for their users, even if their SDK was open sourced (since it would
> have packaging implication). At the end, they followed the route of using
> their closed source SDK and submitted this PR mixing 2 things, and using a
> unusual model for drivers relying on SDK.
> We had some email discussions with them one month ago within the GDAL PSC to
> tell them that their PR couldn't be merged as it and they should at the very
> least rework it as 2 PRs and follow the usual model, but that it would still
> require finding a reviewer keen to review them & press the merge button. That
> part is going to be problematic.

Thanks for the clarification Even! I misread the comment as meaning
"they had GDAL PSC blessing to move ahead with the plugin approach",
not "they've been warned, and continue to waste everyone's time" ;)

I really think it would be a mistake for the community to budge here
and merge either of the new PRs. Pitney Bowes are clearly motivated to
get their code into GDAL, and we should use this motivation to force
them to abide by what's best for the community instead of what's best
for their middle management..! So I'm very glad to hear that I
misinterpreted the situation.

Nyall


More information about the gdal-dev mailing list