Folks,

regarding the `datetime(now)` issue, I need more information from you 
folks, to make sure I'm not hunting a phantom.


As far as I can tell, there may be as many as three variants of behavior 
out there, with `datetime(now)` returning the current time as either...:

(A) UTC time,

(B) LOCAL time, properly respecting daylight savings time, or

(C) LOCAL time, IGNORING daylight savings time,

with all variants appending a literal "Z" (intended to indicate UTC, but 
being bonkers in cases (B) and (C)).


There are also two variants of the source code out there... well, 
actually three by now:

(1) Some old boost-based source code;

(2) some newer C++11-based code; and

(3) a variant of (1) with a fix to enforce behavior (A).


These variants were used in the following POV-Ray versions:

v3.7.0.0 - v3.8.0-alpha.10013324 (2019-01-14):
Variant (1)

v3.8.0-alpha.10064268 (2019-02-19) - v3.8.0-alpha.11272893 (last alpha):
Variant (2)

v3.8.0-beta.1:
Variant (1) again

v3.8.0-beta.2:
Variant (3)


I need your help to test the different variants of the code (especially 
(1) and (2)) on your machines, and let me know:

- What behavior you are observing for code variant (1);
- what behavior you are observing for code variant (2);
- what operating system you are using; and
- what version of boost POV-Ray has been compiled with.

Bonus questions:

- Using genuine POV-Ray, have you ever, and beyond doubt, observed 
behavior (B) during times when DST was in effect in your time zone?


(Please disregard derivatives such as rpov in this test/poll. They have 
no impact whatsoever on the question I'm eager to get answers to.)


I have a hunch that I may have been looking in a completely wrong 
direction in this matter.

Post a reply to this message