freetds AT lists.ibiblio.org
Subject: FreeTDS Development Group
List archive
- From: "Bill Thompson" <thompbil AT exchange.uk.ml.com>
- To: freetds AT franklin.oit.unc.edu
- Subject: Re: NUMERIC seems broken again
- Date: Tue, 9 Jul 2002 07:09:13 -0400
> Confirmed. My Alpha box (little endian) fails in exactly the same way:
>
> 1> select cast(123 as numeric(10,0))
> 2> go
>
> -------------
> -886308406666513612
>
> 1> select cast(1 as numeric(10,0)), cast(-1 as numeric(10,0))
> 2> go
>
> ------------- -------------
> -7205759403792793 7205759403792793
>
> regardless of TDSVER. Maybe we are swapping when we shouldn't, instead of
> not swaping when we should? Bytes, that is.
>
James,
I've diagnosed what I think the fault is with (my) code.
But - before I fix it, my diagnosis suggests that in 0.53 numeric
datatypes would have been broken for big-endian clients talking to SQL
server 7 or 2K. This may well be the reason I decided to change this code
(as that is exactly my configuration). Can anyone confirm this ?
Bill
-
NUMERIC seems broken again,
Lowden, James K, 07/08/2002
- <Possible follow-up(s)>
- Re: NUMERIC seems broken again, Bill Thompson, 07/09/2002
- Re: NUMERIC seems broken again, Bill Thompson, 07/09/2002
Archive powered by MHonArc 2.6.24.