4 ms·
It's not even really important that it's a date bug from a testing perspective; it's just a bug and a signal of an insufficient testing pipeline. You make a pro
by jrootabega 5y ago
It's not even really important that it's a date bug from a testing perspective; it's just a bug and a signal of an insufficient testing pipeline. You make a product (including a definition file), you test it before publishing it. You have canaries. If testing it happens to involve moving the date forward, you do that.
- rbanffy 5y agoHow hard is it to write a test that checks every date in the next 5 years to make sure we know the function will not break without giving us sufficient time to publish a fix in the regular version pipeline?
- jrootabega 5y agoGood point. It feels like what you're describing is like time travel to erase the bug from the timeline completely, vs. getting a one-time advanced warning. I guess for date-related bugs where you have no other way to avoid them, you would need to set the clocks forward at least as far as you have the capacity to get a fix in, regression tested, and get your users enough time to do the same.
- aldebran 5y agoWe’re talking about 1 specific instance. Every single problem looks small until you start looking at the big picture It’s hard!