https://bz.apache.org/bugzilla/show_bug.cgi?id=65217
Mariusz_W <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW --- Comment #10 from Mariusz_W <[email protected]> --- (In reply to Felix Schumacher from comment #8) > On the question about the default value of 'value to shift', if should be > mandatory. I tend to document the current default, that is no shift will be > done on an empty value. Some thoughts: Don't you find this function is too flexible and some cases are too implicite? Maybe it should be more explicit regarding parameters and errors. Maybe all (or almost all) parameters should be mandatory (format especially). Exceptions shouldn't be catched. The function should distinguish whether the pattern is with or without a zone and take the appropriate class ZonedDateTime/LocalDateTime (now ZonedDateTime is always used with the system zone identifier). I wonder if in this case it would not be better to create a new function, e.g. timeShift2, because then the api (mandatory parameters) will change. Additionally, there is a matter of distinguishing the timeToShift parameter - now it is always treated as Duration class. Maybe new additional function should be created eg. timeShift3 (which will treat input parameter "value to shift" as Period class)? -- You are receiving this mail because: You are the assignee for the bug.
