On 6/28/16 7:31 PM, Vikas Rawal wrote:
Imagine the date is followed by 00:00:00. They are instants in time not entire
days.
If they are meant to be instants in time, and the only possible time is
00:00:00, why should it still be exclusive of the terminal instant?
Ledger’s treatment of time has to be consistent between the way time is
interpreted when transactions are recorded (in the data file) and the way time
is treated in the queries.
If we are assuming 00.00.00 time of the day if no time is (or can be)
specified, then we have to do the same with transactions. This would imply that
all transactions that took place on 2016/03/31 should be assumed to have taken
place at 2016/03/31 00:00:00. That by itself is not a sufficient reason for
excluding that instant from this query.
The current semantics are that a transaction recorded with date
2016/03/31 took place at some unspecified time on that day in your
current timezone, later than any similarly-dated transactions already
parsed.
What we have now is that if I want transactions from April 1, 2015 to March 31,
2016, I have to say, show me transactions from April 1, 2015 to April 1, 2016.
Forget ledger and programming for a second. Does this not sound odd from the
perspective of an accountant/lay person? That is not how most people would
think of “from” and “to”.
Ledger users are used to it, but yes, of course you're right. That's why
John suggested maybe adding support for specifying/showing inclusive end
dates as well. This has some costs, eg more complexity. It's awaiting
someone motivated to do the work. The first step is to come up with a
specification I think.
--
---
You received this message because you are subscribed to the Google Groups "Ledger" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to ledger-cli+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
