Re: RFC: date parser rewrite
From: Julian Foad <julianfoad_at_btopenworld.com>
Date: 2003-12-16 23:51:22 CET
Greg Hudson wrote:
I don't have that particular misconception, but maybe I failed to explain my meaning clearly. Let me try again.
When parsing a string of the form "hh:mm[:ss]", MBK was suggesting that we interpret it as the specified time on the current date, but he agreed with me that it would be useful to interpret it instead as the specified time within the last 24 hours or so.
These two interpretations conflict for times that are greater than the time of day "now". In the former proposal, such a time means "later today" and is only useful for overcoming clock skew, and times up to a few minutes greater than "now" might be encountered. In my latter proposal, a time much greater than "now" is interpreted as the specified time yesterday; a time just a little bit greater than "now" is interpreted as "later" to allow for clock skew.
I was then trying to say that, in order to allow future choice of interpretation, we should for now only accept a time in this format if it would be interpreted the same way in both schemes. I said that the set of times for which this is true is the times between 00:00 and "now", but in fact that is not restrictive enough.
I would now say that enforcing such a restriction is unnecessary, because people will not get into the habit of using times of "later today", because Subversion does not deal with times in the future.
- Julian
---------------------------------------------------------------------
|
This is an archived mail posted to the Subversion Dev mailing list.
This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.