Hi,

At April 6, 2004, 16:29, Martijn Tonies wrote:

>> > Forgot to metion Dialect 1 database.....
>>
>> This is why IBConsole, IBOConsole and Marathon didn't complain about
>> ambiguous fields. I think you've proven that ISQL does a check where
>> it shouldn't(on dialect 1 databases), and the bug is more likely in
>> ISQL. The ambiguous fields check isn't performed on a dialect 1
>> database, but on a dialect 3 database. The reason why it is this way
>> is so that old applications(using dialect 1) wouldn't be broken.

> I'm not sure, but I don't think the bug is in ISQL.

> If ISQL reports ambiguitythingy (or whatever you write it),
> it's most probably an SQL WARNING instead of an error.

I thought the warnings were only part of dialect 2, to help migrate
from dialect 1 to dialect 3.

> Or does ISQL have parsing etc build in?

I think ISQL does have parsing compiled in it, and probably does all
the checks for dialect 3, even if it's a dialect 1 database. Which
could lead someone to believe something is wrong with their queries,
if they are supposed to run without errors on dialect 1. But, dialect
3 would return errors.

--
Best regards,
Daniel Rail
Senior System Engineer
ACCRA Group Inc. (www.accra.ca)
ACCRA Med Software Inc. (www.filopto.com)