Closed bago2k4 closed 8 years ago
I added a PR with a failing test for this. Here's the correct result from clj-time:
(def start (t/date-time 2015 10))
start
#object[org.joda.time.DateTime 0x53d93610 "2015-10-01T00:00:00.000Z"]
(t/minus start (t/months 21))
#object[org.joda.time.DateTime 0x79943cc6 "2014-01-01T00:00:00.000Z"]
Hi,
This is probably a bug, though I do remember fixing something like this previously. Maybe it's a regression.
Thanks for adding the tests @belucid !
Awesome @andrewmcveigh !
Hi, I'm working with quarters for a project of my company and i found something that may be a bug. Here are the steps to reproduce:
As you can see if you do 2015/10/01 minus 18 months (6 * 3) you get 2014/04/01 that is correct but if you subtract 21 months (7 * 3) you get 2013/01/01 that is incorrect, it should be 2014/01/01 or 2013/12/31 (or something like that).
Do you think this is a bug or I'm using it the wrong way?