[PROJ] The state of PROJ C++ 2020

[Howard Butler]

> On May 7, 2020, at 9:57 AM, Søren Holm <sgh at sgh.dk> wrote:

> 
> I know that PROJ 6.3.0 contains more code (approx x3) and the tests take a 
> long time to compile because of google-test but subtracting those 20 test-
> files (2 seconds each) still leaves 6.3.0 with a compiletime 165 longer that 
> 5.1.0.

Compile times matter to you? If you are recompiling PROJ so much, where are all your pull requests? 

> It might just be that I am old or something (apparently 39 is old these days), 
> but it seems to me that taking a working software library and increasing 
> linecount x3 while increasing compiletime x10 without adding new files is just 
> crazy.

You are welcome to keep using PROJ 5.1.0 indefinitely. That's the beauty of open source. You are not forced to upgrade.

> Is it realy worth it?
> Is it easier to maintain?
> Do we get better performance (I doubt it)?
> How about portability?

It's worth it to me, because PROJ 5.1 produced wrong answers for any sufficiently interesting geodetic transformations. As geodetic agencies move toward space-based geodesy and time dependent coordinate systems, that problem was going to get much worse. Without the updates that PROJ received over the past two+ years, software dependent upon PROJ would have been producing incorrect values. Those incorrect answers would have had consequences in the form of PROJ-based software being disallowed in situations where geodetic accuracy matters. YOU might not need that, but the active contributors and sponsors that supported gdalbarn.com <http://gdalbarn.com/> did. As a result, PROJ is now one of the most advanced capability software libraries available commercial or otherwise in this regard.

PROJ has more active contributors now than it has in its 35+ year lifespan. Indeed there is more to maintain because more is being asked of the software. There's more docs and more tests than at any point in the project's lifespan too.

The project would be very interested in performance metrics, numbers, and examples. The PostGIS and QGIS projects have provided a ton of performance feedback that has resulted in many significant improvements. Disappointingly, commercial users of PROJ did not provide much feedback on the topic at all to date.

As for portability, that costs. Is your organization offering to support portability through contributed time or money? 

Howard


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20200507/193ba598/attachment.html>