RE: CFINT: I understand it now...

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> -----Original Message-----
> From: Leif Svalgaard
>
> In what way?  In fact, a well written server program is far MORE efficient
> than a traditional monolithic green screen application.
>
> ===> I think you are overstating the point here. They are about the
> same when talking processing efficiency. The server approach is
> far MORE efficient when it comes to maintenance, but that is not
> the issue here.

You're right, Leif.  A server program is more efficient than the
corresponding monolithic program, but the sum of the client program and the
server program is roughly equivalent.  However, a server can provide
performance boosts such as leaving data paths open and caching small tables
that simply aren't as efficiently implemented in a monolithic program that
is invoked in different interactive sessions.  Correct?


> Let us assume for a moment that everybody went where 'IBM
> wants them to go' and converted everything to run client/server.
> That would remove the CFINT revenue and if as some (e.g.
> Jon Pais) have claimed that revenue is essential to the viability
> of the platform, then the platform will die when everybody is
> doing client/server. The only saving grace is that doing client/
> server may require a lot more processing power forcing people
> to buy bigger boxes thus enhancing IBM's revenue to offset the
> CFINT tax,

Ah, but I don't agree with Jon's assertion.  I think that if the iSeries
were used as a true server box (and not as an ODBC server!) it would run
rings around the competition and would in fact sell MORE, rather than less.
In any case, it's irrelevant to my point that, for the end user,
client/server processing is better than monolithic programming.


> If as you say, server programs are far MORE efficient, people
> can get by with smaller boxes further eroding IBM's revenue
> and thus the viability of the platform. I think I may be missing
> something here, but I can't see what.

You're assuming that the interactive cost is required to keep the platform
viable.  If that's indeed the case, then the box will die no matter what we
do, because interactive programming is, whether we like it or not, a dying
animal and is being replaced by client/server techniques.  The only question
is whether we as a community are willing to stand up and develop
architectures that take advantage of the box's strengths, thereby making it
the server of choice, or whether we want to take the easy route of ODBC
development, one which makes the iSeries just one more entry in the
bloatware sweepstakes.

Joe Pluta
www.plutabrothers.com