LiveDocs Comments - coldfusion - 6 - CFML_Reference - functions-pt158.htm

coldfusion/6/CFML_Reference/functions-pt158.htm

2005-08-26T10:09:53

OK - further research has found that apparently datediff('d'....) counts "whole days" and not date boundaries. This is stupid if you ask me - if I wanted to count the notion of "whole days" - I'd do 24 hours.

Anyway, a fix I found is to format the days to drop the time. For instance, if I'm counting the number of days a certain date is from today:
DateDiff('d', DateFormat(variables.mydate, 'mm/dd/yyyy'), now()) would yield 0 for today, 1 for yesterday, etc. Notice I only do the format on the first date since it basically makes it count as 12am on that day - so anytime today would count as at least a whole day.

Macromedia: could you please fix this or add a new date part that would count day boundaries?

coldfusion/6/CFML_Reference/functions-pt158.htm

2005-08-26T09:59:23

I'm running CFMX7 - it seems that DateDiff('d', start, end) doesn't count the number of day boundries crossed (as it does in SQL, etc). For instance:

DateDiff('d', '8/26/2005 8:00pm', '8/27/2005 9am') returns 0 - instead of 1 (as I would expect). Are my expectations wrong?

coldfusion/6/CFML_Reference/functions-pt158.htm

2005-08-09T12:13:58

It seems to me that DateDiff returns the larger of 2 options when asked to calculate over the fall daylight savings. The following code returns 75, for the number of minutes between 00:45 and 01:00 on Oct 30, 2005. But 15 would also be a correct answer since 01:00 occurs twice on this day. And my code is expecting the 15. Is there anyway to specifiy this difference?

DateDiff ("n",CreateDateTime("2005","10","30","0","45","0"),CreateDateTime("2005","10","30","1","0","0"))

always returns 75. What about the other correct answer, 15?

coldfusion/6/CFML_Reference/functions-pt158.htm

jrunrandy — 2005-03-15T13:40:47

captainplanet: Are you ColdFusion MX or ColdFusion MX 6.1? My understanding is that the DateDiff problem is fixed in the CFMX 6.1 updater: http://www.macromedia.com/support/documentation/en/coldfusion/mx61updater/releasenotes_cfmx61_updater02.html

coldfusion/6/CFML_Reference/functions-pt158.htm

captainplanet — 2005-03-08T06:45:48

Also, we are fully updated with all the patches.

coldfusion/6/CFML_Reference/functions-pt158.htm

captainplanet — 2005-03-08T06:44:48

returns 29

returns 29

returns 29

returns 30

I think this is a problem. We're writing finacial software & the interest calculation is off because of this problem. I need some help.......

coldfusion/6/CFML_Reference/functions-pt158.htm

godardk — 2004-09-02T12:03:09

the patch worked for me, CFMX Standard 6.1 on win2000

coldfusion/6/CFML_Reference/functions-pt158.htm

2004-08-03T03:53:20

this hotfix didn't work on my mx 6.1

coldfusion/6/CFML_Reference/functions-pt158.htm

jrunrandy — 2004-04-13T07:49:38

Are you running ColdFusion MX or ColdFusion MX 6.1? AFAIK, DateDiff works OK in CFMX. For 6.1, there is a hot fix to the DateDiff function: http://www.macromedia.com/support/coldfusion/ts/documents/duplicate_hotfix.htm

coldfusion/6/CFML_Reference/functions-pt158.htm

WatVduck — 2004-04-13T03:25:35

I am trying to use datedif to calc age.
DATEDIF(yyyy,14-Apr-1996,13-Apr-2004) = 8
whereas in every other language i use it gives 7 years.
Why is this?

coldfusion/6/CFML_Reference/functions-pt158.htm

Hutz — 2004-02-09T07:56:18

DateDiff applies daylight savings differences to Day Of Year ("y") and Day ("d") calculations, and this leads to non-intuitive results. Viz:

(The first date of the datediff is during daylight savings time, the second date is not; The first datediff uses 10:30pm and 0:00, the second datediff uses 11:30pm and 0:00)
#DateDiff("d", CreateDate(2003, 10, 25), CreateDateTime(2004, 2, 6, 22, 24, 0))#
#DateDiff("d", CreateDate(2003, 10, 25), CreateDateTime(2004, 2, 6, 23, 30, 0))#

The first one returns 104, the second 105.

Move the date so BOTH are NOT during daylight savings:

#DateDiff("d", CreateDate(2003, 10, 27), CreateDateTime(2004, 2, 6, 22, 24, 0))#
#DateDiff("d", CreateDate(2003, 10, 27), CreateDateTime(2004, 2, 6, 23, 30, 0))#

And they both return the same 102.

To further perplex the issue, the intuitive way to make the above first result happen would be to convert the local datetime to UTC (11:30 daylight savings is really 12:30 the next day, so when converting to UTC, it would make sense for there to be a difference between 10:30pm and 11:30 pm). HOWEVER, when you do this, instead what happens is that there is no longer any difference in the output between daylight-savings days and non-daylight-savings days:

#DateDiff("y", DateConvert("Local2utc",CreateDate(2003, 10, 25)), DateConvert("Local2utc",CreateDateTime(2004, 2, 6, 22, 24, 0)))#
#DateDiff("y", DateConvert("Local2utc",CreateDate(2003, 10, 25)), DateConvert("Local2utc",CreateDateTime(2004, 2, 6, 23, 30, 0)))#

returns 105 for both, and:

#DateDiff("y", DateConvert("Local2utc",CreateDate(2003, 10, 27)), DateConvert("Local2utc",CreateDateTime(2004, 2, 6, 22, 24, 0)))#
#DateDiff("y", DateConvert("Local2utc",CreateDate(2003, 10, 27)), DateConvert("Local2utc",CreateDateTime(2004, 2, 6, 23, 30, 0)))#

returns 102 for both.

coldfusion/6/CFML_Reference/functions-pt158.htm

clu@oncallinteractiv — 2003-01-10T18:28:00

Just FYI: This bug is fixed in update pack 1.

coldfusion/6/CFML_Reference/functions-pt158.htm

clu@oncallinteractiv — 2003-01-07T19:42:00

I think it's buggin in CFMX. Try this on a MX machine:

#DateDiff('m','1/1/1999','3/1/1999')# = 1
#DateDiff('m','1/1/1999','3/2/1999')# = 1
#DateDiff('m','1/1/1999','3/3/1999')# = 2
#DateDiff('m','1/1/1999','3/31/1999')# = 2

coldfusion/6/CFML_Reference/functions-pt158.htm

daveline — 2002-12-12T20:58:00

this function has changed since coldfusion 5
the result of
#dateDiff("m","1/1/2002","12/1/2002")# was 11
in MX it is 10