As a reminder, if you are licensed to CICS Transaction Server Version 3 or
higher, or IMS Transaction Manager Version 9 or higher, your first Rational
Developer for System z license is available at no charge. You can open PMRs
with the CICS or IMS teams for specific RDz use cases related to CICS or
IMS. (Example: building CICS Web services.) General RDz support is not
included, however.
Yes, per-developer ("Authorized User") and per-concurrent developer
("Floating User") licenses are available. It is possible to buy some of
both as well. Just choose whatever mix makes the most sense.
OK, now on to price. "Oh my God, that's X dollars!"
...Well, yes, RDz costs money. (Yes, as mentioned, IBM does not always
charge list price.) So does any other professional tool, including drills,
hammers, teeth cleaners, diamond cutters, delivery trucks, and CAT
scanners. But the back-of-the-envelope math to build an RDz business case
isn't too complicated. Branham Group says there's about a 20% productivity
gain (ceteris paribus) when using RDz:
http://www.branhamgroup.com/CustomContentRetrieve.aspx?ID=2002290
Let's assume a fully burdened FTE cost per year of $72,000 ($6,000 per
month) per developer. That's probably too low for the U.S. and Western
Europe -- and probably absurdly low for contract programming services --
but let's use it anyway. And let's cut Branham's estimate in half and
assume a 10% productivity gain. But let's also assume there's a learning
curve, so for the first month it's -30%, the second month is -20%, the
third month is -10%, the fourth month is 0%, and then in the fifth month
the 10% productivity benefit kicks in. Let's assume the license is $5,000
(little or no discount) and the subscription/support renewal is $1,000 per
year (ditto). And let's assume there's no change to peak monthly MSU
consumption on the mainframe. (Some would argue there often is a little CPU
benefit with RDz, but I'll be conservative and assume not.) Here's how that
all works translated to cash flow:
Month 1: -$6,800 ($5,000 for the license and $1,800 for the productivity
hit)
Month 2: -$1,200
Month 3: -$600
Month 4: $0
Month 5: $600
Month 6: $600
...
Month 13: -$400
Month 14: $600
Month 15: $600
...
Month 25: -$400
Month 26: $600
...
Under these rather conservative assumptions you reach the break even point
in the middle of year 2, and after that you're getting a substantial
positive return on investment. YMMV -- insert your own numbers. (Including
a net present value discount rate adjustment, which I'm not including here
for simplicity -- and because interest rates are extremely low in the U.S.
and Western Europe.)
Please also note that I'm assuming like-for-like capabilities and just
considering a straight-up productivity gain for pounding out/maintaining
COBOL and/or PL/I code. That's a simplification that's too conservative.
RDz makes *possible* certain types of development that just flat out aren't
possible (realistically) with ISPF/TSO, so it helps reduce development and
deployment costs elsewhere. People often forget that IT isn't always about
just keeping the lights on, so to speak. It should also be about
contributing to business innovation and efficiency. RDz's support for Web
services, the CICS Service Flow Feature, UML modeling, and several other
functions are very much in that category. I'm also setting aside the
multi-language support (JCL, REXX, Assembler, COBOL, PL/I, C/C++, Java,
EGL, etc.) and graphical user interface support for Fault Analyzer, File
Manager, and Application Performance Analyzer (which boosts productivity in
those areas, too). It isn't realistic to ignore those benefits, yet in the
above example there's still a strong positive ROI.
I suppose some people might say this is all well and good, but (insert name
of "evil" person in my organization) won't support a business case even
with a positive return on investment in ~18 months. OK, perhaps, but is
that IBM's problem or yours (and your shareholders')? :-) Heck, even if
you're capital-constrained, IBM offers financing for software purchases --
they've thought of that, too. Most rational (no pun intended)
businesspeople would be quite happy to invest in projects with ~18 month
ROIs.
I am not necessarily suggesting that *every* developer should have RDz.
However, there are strong business cases for a large percentage of
developers to get RDz.
- - - - -
Timothy Sipples
IBM Consulting Enterprise Software Architect
Based in Tokyo, Serving IBM Japan / Asia-Pacific
E-Mail: [email protected]
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
