FreeTDS Development Group

[ZIGLIO Frediano <Frediano.Ziglio AT vodafoneomnitel.it>]

> 
> Question to you, Mr. Portability:  Do you know of a compiler out there
> without 64-bit integer support?  Even VC++ has it (but it's 
> not "long long",
> it's __int64, don't get me started).  Anything without it would be
> constrained to a 4 GB file length, which seems pretty 
> unlikely these days,
> for anything we care about.  

Some other use _int64, other use long (like alpha)...
I think we should define a new type like TDS_INT8 and TDS_UINT8... or
whatever you want...

> 
> There are four things we can do when we run out of integer bits: 
> 
> 1.    Punt.  My favored option for now.
> 2.    Use long, punt if the incoming value is too big.  
> 3.    Use float & sprintf, taking care w/rounding.  One day.  
> 4.    Write "atoll" (cf. atoi) w/gcc, offer it in 
> binary/assembler form to
> broken systems.  
> 5.    Use a 64-bit library.
> 6.    Write a 64-bit library.  
> 

Mmmm... a multiple precision library or only a 64 bit library?
Numeric are coded in 128 bit, perhaps we could use this code too or share
something...
I think the only usefull routine we need are 64bit-to-string and
string-to-64bit and some test for convert to/from 32bit.

freddy77

=================================
"STRICTLY PERSONAL AND CONFIDENTIAL

This message may contain confidential and proprietary material for the sole
use of the intended recipient. Any review or distribution by others is
strictly prohibited. If you are not the intended recipient please contact
the sender and delete all copies.
The contents of this message that do not relate to the official business of
our company shall be understood as neither given nor endorsed by it."

=================================