My opinion would be to support the official standard ( despite the fact I
also think it's braindead to have the user 'guess' at the
maximum length they might need.)  If using a sentinel value (-1) doesn't
conflict with any existing practice, I'd say 'extend' the
'standard' in that direction.  The alternative would be an install/build
time choice of behaviour, which might be useful for ingres-only sites who
don't need the 'compatibility' with other drivers.  I've always found the
DBI to be too low-level to be able to write a 
RDBMS-agnostic script/app anyway.  Don't take that as a criticism of the
dbi.  I find the dbi to be the most compelling reason outside of the perl
core itself for using perl.  

Mark.

Ingres/Oracle DBA/Developer.

-----Original Message-----
From: Mike Battersby [mailto:[EMAIL PROTECTED]
Sent: Thursday, June 12, 2003 8:17 AM
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: Ignore LongReadLen? (Also DBD::Ingres)



Dear DBI developers,

I've been doing some more work on the long data type support in 
DBD::Ingres (which unfortunately still hasn't managed to make it into the 
official distribution), and now have support for the Ingres byte and
long byte binary data types as well.

I have a question about long data types though, which hasn't been 
resolved by checking other DBD drivers, as they all seem to do things 
differently.

Is it preferable to support LongReadLen and LongTruncOk, or just to return
the entire field?  To be honest, in my own work I never have a need for
limiting the length, and it seems ludicrous to make the user guess what
the largest size data is if you can just return the whole lot anyway by
allocating memory dynamically.

Maybe there needs to be a sentinel value (-1?) for LongReadLen to indicate
fetching everything?

At the moment I have two versions of DBD::Ingres in development, one which
supports the Long* settings exactly as per the DBI pod, and one which
ignores
them entirely and always returns the complete data.

Any hints about the correct way to deal with long values would be 
appreciated.  I consider it essential to allow the programmer to fetch all 
the data without having to guess at the size.

Henrik: if you read this and you ever get time to look at my long
data type stuff, drop me a line first so I can zap you the most recent 
version.

Cheers,

 - Mike Battersby

-- 
Mike Battersby <[EMAIL PROTECTED]>        The University of Melbourne
Fetch my pgp key from: http://ariel.ucs.unimelb.edu.au/~mib/pgpkey.txt



**************************************************************************************
This e-mail may contain information that is privileged, confidential
or otherwise protected from disclosure. It must not be used by, or
its contents copied or disclosed to, persons other than the intended
recipient. However, the contents of this e-mail may be intercepted,
monitored or recorded by Insurance Technology Solutions Limited
for the purposes of ensuring compliance with its policies and 
procedures. Any liability (in negligence or otherwise) arising from any
third party acting, or refraining from acting, on any information
contained in this e-mail is excluded.

If you have received this e-mail by mistake please notify our System
Administrators at [EMAIL PROTECTED] and delete this e-mail.

It is the responsibility of the recipient to ensure that the onward
transmission, opening or use of this message and any attachments will
not adversely affect its systems or data. Please carry out such virus and
other checks as you consider appropriate. No responsibility is accepted
by Insurance Technology Solutions Limited in this regard.
**************************************************************************************

Reply via email to