Vlad,
Thanks for the comments, you do have a valid point there. There is no
need to take the extra step of computing the intermediate result.
Especially when it may lead to ambiguous results. I will change my
implementation accordingly :)
Btw, that was a fast review on a Saturday!
/Lars-Erik
On Nov 22, 2008, at 4:24 PM, Vladislav Vaintroub wrote:
> Lars-Erik, I like the comments and test-case very impressive. Very
> good:)
>
> There is one subtlety though. Unlike in math, relying on difference
> between
> values when making comparisons does not always work, specifically in
> corner
> cases, where difference would overflow the range for the data type.
>
> Illustration is the program below. It will incorrectly think that
> INT_MAX(==2147483647) < INT_MIN(-2147483648) and assert ,because
> INT_MAX -
> INT_MIN is -1.
>
> #include <limits.h>
> #include <assert.h>
>
> int compare(int a, int b)
> {
> return (a-b);
> }
>
> int main(int argc, char argv[])
> {
> int result = compare(INT_MAX,INT_MIN);
> assert(result > 0);
> return 0;
> }
>
> I propose move away from computing difference and consistently do
> something
> like
>
> if (a < b)
> return -1;
> if (a > b)
> return 1;
> return 0;
>
> for all datatypes instead.
>
>
>
>
>> -----Original Message-----
>> From: Lars-Erik.Bjork@stripped [mailto:Lars-Erik.Bjork@stripped]
>> Sent: Saturday, November 22, 2008 1:35 PM
>> To: falcon@stripped
>> Subject: Request for review bug#40614
>>
>> Hi all,
>>
>> The patch is described in the commit, available at
>> http://lists.mysql.com/commits/59617
>>
>> Have a nice weekend!
>>
>> --
>> Falcon Storage Engine Mailing List
>> For list archives: http://lists.mysql.com/falcon
>> To unsubscribe: http://lists.mysql.com/falcon?unsub=1
>
>
>
> --
> Falcon Storage Engine Mailing List
> For list archives: http://lists.mysql.com/falcon
> To unsubscribe: http://lists.mysql.com/falcon?unsub=1
>
