Re: [Chicken-users] bizarre srfi-19 behaviour (maybe DST?)

[Kon Lovett]

On Mar 12, 2007, at 11:51 AM, Graham Fawcett wrote:

> Hi folks,
>
> I'm puzzled by this behaviour, which I'm only seeing on one of two
> Linux machines:
>
> (use srfi-1 srfi-19)
> (delete-duplicates
> (fold (lambda (n acc) (cons (date-minute (current-date)) acc))
>       '() (iota 2000)))
>
> On one machine I get a list of one element, the current minute, eg:  
> => (32)
> But on the other machine I'm getting this: => (40 39 38 37 36 35)
> Of course, the code is running in subsecond time, so I should see no
> more than two values for (date-minute ..) in a given list, and most
> likely only one.
>
> The system clock is ok:
>
> (delete-duplicates (map (lambda (n) (current-seconds)) (iota 2000)))
> => (1173724927.0)
>
> The good machine is
>  Version 2, Build 41 - linux-unix-gnu-x86 - [ dload ptables  
> applyhook ]
> The bad one is
>  Version 2.5 - linux-unix-gnu-x86 - [ dload ptables applyhook ]
>
> I'm in Canada, and we just moved to daylight-savings time; I can't
> tell whether this is a post-DST bug or something else. Thoughts?

Well, your machine is not at fault. At least there was a real bug in  
'(current-date)'. It was performing dst arithmetic when no tz-locale  
was supplied. So, yes, caused by the switch to daylight saving time.  
Thank you for reporting it.

Please get the current srfi-19 egg, version 2.6.1.

Best Wishes,
Kon

> Thanks,
> Graham
>
>
> _______________________________________________
> Chicken-users mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/chicken-users