RE: Product Enhancement by Voting?

Richard wrote:
>No, all is well.  As you noted above, that is what the language
>planners are supposed to bring.  It isn't perfect but it isn't
>bad.  Throw in some developers from the large software houses
>and a consultant or two (Al Barsa or his ilk) and you have a
>competent team.

That's the traditional model.  But last summer, we tried something
a little different - whenever we had a question of how a
particular enhancement should be designed, we came to this
mailing list for advice.  And you know what?  We got some very
useful feedback.  We got opinions from a wide variety of people,
and probably faster than if we just asked the usual group of
experts.  Also, one thing we discovered was that the opinions of
the experts didn't always match up with the opinions of average
RPG programmers.

As I mentioned before, we don't take a majority vote - one well
thought out argument can overrule dozens of other opinions.  We
examine the opinions and arguments on their technical merits, and
not on the quantity (or reputation) of the opinion-holders.


>But, there is a much more important problem.
>
>The AS/400 is under siege from the NT- and Unix-based software
>houses.  Based on their language choice, those people have a very
>small vocabulary - C.  Since the Unix/NT people represent new
>sales for the AS/400, IBM cannot afford to ignore them.  When they
>vote, somebody is going to listen.  New development money and
>inside access will be lavished on them - it would be foolish not
>to.
>
>There are thousands of AS/400 developers whose vocabulary doesn't
>include C and who, for several reasons, can't or won't get access
>to it.  For their future, they want you to add some of the new
>stuff they have heard about to the old language.  But not too much
>and not too fast - they don't want to be kicked out of their
>comfort zone -

(I like that term "comfort zone"!)

>and, of course, it has to be free.  Those people
>represent a group that upgrades their systems and adds machines to
>existing clusters - a steady source of revenue.  IBM can't afford
>to ignore them either.  We can see their power when we look at how
>long matching record lasted.
>
>[You compiler writer guys must have a lot of wisdom to live in a
>battlefield like that - and I'll bet you guys are QUICK too!]
>
>Based on this quick look and a profound ignorance of the facts, I
>would invest most of my development dollars on the new stuff.  I
>would build features that will help to sell the box into new
>markets.  I would spend small but not zero dollars on maintaining
>a solid connection between the old languages and the new operating
>system features because existing AS/400 customers are less likely
>to defect to NT and Unix if "their" features are regularly upgraded.
>In my blissful cocoon of ignorance, I would probably balance the
>funding 75-25 or 80-20 and that is real money.
>
>Circling back to the language enhancement issue for a minute, should
>the investment in RPG be increased to include a bunch of new language
>features?  Based on where the new money is coming from, investing in
>RPG would be silly!  Does anyone honestly think that there will be
>significant new application development using the RPG language?  Even
>an old crustacean like me can see where the new business is coming
>from.  IBM is not a social club and it isn't a charity - it has to
>fight hard to keep up and it has to get lucky to get ahead.  RPG is a
>poor bet for the future.
>
>Does this logic apply to anything else on the box?  Like the operating
>system for example?  In my opinion, maybe.  There are two things wrong
>with OS/400 and they are both pretty hard to fix - it takes too long
>to make changes and they can only come from one source.  I think that
>the OS discussion can wait for another note.
>
>Ladies and gentlemen, I can't claim to be smart (too much evidence to
>the contrary) but I can't deny that I am "experienced" and this is what
>my experience tells me.  It seems to me that a few of our plans and
>assumptions may need to be examined in the light of these ideas.

There's very little in your analysis that I would disagree with!

I've written here before about my recent experiences moving
outside of my own "comfort zone", learning about Java, Linux,
Perl, web programming, etc.  I've seen what we in the AS/400
world are up against.  As a result, I understand fully why IBM is
embracing stuff like Java and Linux and Apache.

I agree fully that "plans and assumptions may need to be examined
in the light of these ideas".  There are too many AS/400 people
who don't want to move out of their "comfort zone".  There are
still too many people who still use RPG III even though moving to
RPG IV is practically a "no-brainer".

I've been trying to encourage people to take an "open source"
approach to RPG software, but many still look towards IBM to
satisfy all of their software needs.  But that's unrealistic.  In
contrast, a Perl programmer (for example) has almost instant
access to hundreds of tools and software libraries available for
no cost.  As a result, Perl is one of the most productive
languages around.  Why can't that also be true for RPG
programming?

Regarding what to do with RPG, we're really between a rock and a
hard place.  Many people really don't care about new features in
the language (prefering to stay within their comfort zone), but
if we don't enhance the language, RPG programmers people will
think we're abandoning them.

And if we do too much, people will complain because they can't
keep up.  (That probably sums up pretty well the main complaints
against our V5R1 content!)  Unfortunately, if you want to stay
in your comfort zone, you'll be left behind since the rest of
the computing world won't wait for you.

Cheers!  Hans

Hans Boldt, ILE RPG Development, IBM Toronto Lab, boldt@ca.ibm.com


+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---