[R-sig-Geo] rgdal release candidate 1.5-9 rev. 1000 ready for testing

[Edzer Pebesma]

On 6/6/20 9:15 PM, Patrick Schratz wrote:
> Since I use and contribute to R (6 years for now) I was always on the
> side of "help everyone", "things can get better/easier", "don't create
> your own island solutions". However, I was hitting some (in my view, not
> understandable) "walls" recently and I am not sure if I want to continue
> with this attitude.

Dear Patrick,

We should keep in mind that the system requirements of the R spatial
packages is quite likely the most complicated there is on CRAN, that it
has a long legacy, and that GDAL/PROJ came very dynamic recently, and
that many people (though not enough) are involved.

I think that you're learning that working with people is much harder
than working with technology.

I highly appreciate the contributions you made to CI for R packages [*],
and the way you helped making it work for sf; this helps robust
development and finding bugs early. However, the moment something breaks
in the CI setup, I have no clue what to do and have to ask your help. If
I wouldn't get that help a couple of times, I would drop the whole thing
and throw it out of the window. For people using sf this is the same:
they have no clue how it works; if it doesn't work a couple of times and
help doesn't come quickly, they throw it out of the window and look for
something else that does work.

What you have to realize is that what looks simple for you (e.g. CI)
looks incredibly complicated for other people, for the simple reason
that they have been doing very different things for the last 6 years,
and have neither time nor ambition to make up for that. Convincing
others to change something by arguing the change is "simple and makes
things better" often does not work because the receiver has a different
perception of "simple" and is not convinced that the status quo is not
good enough, or is convinced the change brings a lot of work and/or
added complexity. I don't think it helps to get emotional publicly in
such a case and move to twitter [&], nor to become disappointed in the R
project. If you'd move to another project, you'll find yourself again
confronted with people, and discover that communication is always difficult.

[*] see e.g. https://github.com/ropensci/tic
[&] https://twitter.com/pjs_228/status/1269301044481339392
-- 
Edzer Pebesma
Institute for Geoinformatics
Heisenbergstrasse 2, 48149 Muenster, Germany
Phone: +49 251 8333081

-------------- next part --------------
A non-text attachment was scrubbed...
Name: pEpkey.asc
Type: application/pgp-keys
Size: 3110 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20200606/d1ee363b/attachment.bin>