Re: Which of the SYSIBM tables/views show the row count for

On Fri, Mar 14, 2008 at 12:53 AM, Tom Liotta <qsrvbas@xxxxxxxxxxxx> wrote:

I'm pretty sure that he means something about how _programmers_ are
allowed to code programs. But it's not clear at all because
_programmers_ can write code on Windows as well as on whatever
systems run Oracle or any other "real RDBMS".

I'm looking to understand why DB2 for i5/OS is inferior because it's
in control for "native I/O (whatever that means)" when other RDBMSs
are not.

And I noticed that you didn't provide an answer for that either...?

Well mostly because i'm indifferent about it. "native I/O" on i5/OS
doesn't hurt anyone. I mainly see it as a way to continue developing
legacy applications without significant overhaul, and as such it
represents strong backward capability by IBM. That's a good thing. The
bad thing is that people still use it for new applications.

Fine. Dave knows such stuff in detail, enough to consider DB2 under
i5/OS not to be "real". He gave reasons -- one of which had to do
with "native I/O". It seemed to be an important reason, but it
wasn't defined in any useful sense.

I'm hoping to be educated by him about it so I can understand his point.

Maybe you can explain...

Well, i of course do not know Dave's Motivation and points. So i can't
really help you out here. I stated my personal opinion about native
I/O above, and in general what Dave is complaining about are the
developers that are stuck in the 80ies, just partly the database (and
there mostly about the terminology that IBM invented and the fact that
there are other methods besides SQL available).

Yes, IBM decided to use it's own terminology for DB/2 on i5/OS. I can
live with that. To me, a file is a table. I can understand when
someone says "file" in the i5/OS context, but to me it's always a
table.