RE: DB and App Agnosticism on the iSeries (thinking about: Questionabout UDB on iSeries)

> From: Pete Helgren
> 
> My concern is, if we concentrate on writing ONLY applications that run
on
> the iSeries, then we have two challenges when it comes to selling:  1.
> Convince them to buy an iSeries and 2. Convince them that our
application
> is superior to any other (and, that there is greater value in that
> combination than elsewhere)

Well, if you can't do the second, then you really have no business doing
the first.  But let's continue this discussion.

> We sell to school districts and the O/S politics are very high there.
> They
> have folks who know Microsoft server products and, increasingly,
Linux.
> When we tell them that they will HAVE to buy a server that no one in
the
> district has experience with, we have an uphill battle.

Yes you do.  You have the battle of selling somebody the best solution
as opposed to simply making money off of whatever they choose to buy.
But that's an ethical question, and let's continue past that particular
sticky wicket.


> If we can tell them
> "I don't care what you are running hardware/OS-wise, as long as you
want
> to
> live with the limitations and hassles of WinTEL or LinTel, you can run
our
> software" we don't have near the battle.

Of course, that in turn tells them there's really nothing that special
about the iSeries, since your software will run on Windows or Linux.
Especially since I'm relatively certain your sales reps aren't going to
be spending a lot of time on any perceived "limitations", Windows or
not.  But that's another ethical issue we'll avoid for this discussion.


> OF COURSE we recommend an iSeries,
> but they may balk.  Saying we can also run on other platforms may
close a
> sale we might otherwise lose.

Now, are you SAYING you run on Windows with the idea of actually trying
to get them to an iSeries, or are you actually trying to sell Windows
software?  The former is yet ANOTHER ethical quagmire, and so let's once
again assume the better path and assume you're trying to actually write
competitive Windows software.


> That is the battle that rages here.  Write only for the iSeries, or
write
> in ways that can run on pother platforms.

So, let's assume that you're past all of the ethical issues I've raised,
and you really intend to write software that will run on Windows and
Unix boxes as well as the iSeries.  The only possible architecture
you're talking about is a browser interface, using JSP and servlets.

That means you're going to be putting up your browser-based solution
against native iSeries applications on the one end, and Windows/Unix
solutions on the other.  And while there may indeed be a few niches
where a browser-based solution might win, it seems to me you're going to
have some serious problems.  Let's see why:

1. From the iSeries side, you will be stepping down in feature/function,
reliability and productivity.  In effect, you're offering an inferior
product to your iSeries customers.

2. From the Windows/Unix side, your programming costs will actually
escalate.  SQL is *NOT* the same on every box, and of you'll have to
support various combinations of OS and database.  Testing costs will
rise, as will support costs.  At the same time, you'll have to reduce
your price point to compete with other solutions.  Finally, you'll have
many of the same battles as with the iSeries native solutions, because
there will be native Windows/Unix solutions that run faster and look
better than yours.


So, your simple "run on more platforms, make more money" concept isn't
quite so simple.  After you take into account multiple development and
test environments, multiple support centers, reduced feature/function,
increased low-end competition and the inevitable lower price point, it
seems to me that you had better be able to sell a LOT more units of your
product in order to justify moving to a lower common denominator.


But hey, that's just me.  I only saw one 500 million dollar company go
bankrupt doing what you're suggesting.  What do I know?

Joe