Fails to properly decode slim tzdata

[Ken Murchison]

On 3/15/21 8:15 AM, Milan Crha wrote:
> On Sun, 2021-03-14 at 16:19 -0400, Ken Murchison wrote:
>> Are you in a position to see if
>> https://github.com/libical/libical/pull/484
>> <https://github.com/libical/libical/pull/484> fixes the problem?
>
> 	Hi,
> sure, it works for me. I tried both slim and fat tzdata. There is
> increased failure rate of the src/tests/timezone test with the slim
> format (1248 days failed for the fat format, while 1486 days failed for
> the slim format), but that can be due to the changed recurrence rule
> generation. Interestingly, reverting the change shows 2708 days failed
> for the fat format and 22943 days failed for the slim format here.
> Meaning there is an improvement for both formats.


Do you mean src/test/timezones.c or some other test?


> Just a note, I tried it on the 3.0 branch, which required the
> icalrecurrencetype_encode_day() from the master branch. Not a big deal,
> just in case you'd want to backport to the 3.0 branch as well.


Noted.  Thanks.

-- 
Kenneth Murchison
Senior Software Developer
Fastmail US LLC