Midrange News for the IBM i Community

Posted by: Bob Cozzi
Rogue Programmer
Cozzi Productions, Inc.
IBM Too Big to Succeed?
has no ratings.
Published: 10 Jun 2014
Revised: 14 Jul 2014 - 1074 days ago
Last viewed on: 21 Jun 2017 (2281 views) 

Using IBM i? Need to create Excel, CSV, HTML, JSON, PDF, SPOOL reports? Learn more about the fastest and least expensive tool for the job: SQL iQuery.

IBM Too Big to Succeed? Published by: Bob Cozzi on 10 Jun 2014 view comments(8)

Is IBM Too Big to Succeed?

I've been listening to talks about IBM's Java-based GUI Developer Tool, known as "Rationsl Developer for [insert product ID here]. For us IBM i folks, this is RDi or "Rational Developer for i".

Overall RDi is an okay product. It does a lot, about 80 percent of that capability is uninteresting to IBM i developers, and of that 80 percent, at least 70 percent will never be used by IBM i developers.

RDi isn't by any stretch of the world "modern". It does a few very cool things, like the code outliner, compare tool, subroutine nesting/jumping, and the navigator. The screen design tool for DDS Display Files is okay, but a bit awkward to use, but not bad. For RPG editing its okay, and often works as expected. The only serious bug in the tool has been immediately after compiling where the double-click to select and cursor-to-select features break and you have to know the secret to get them working again. (The secret is to double click on a line that contains no errors, but contains some code and force it to select something. Then it usually starts working again.)

I'm ont trying to bash RDi it is what we have and now that PC's have gotten fast enough to "fix it with hardware" the product is finally usable for most people.

But RDi today, save the few cool items I've point out, is doing very little more than what I wrote back in 1992 in my "Visual RPG" product for windows. This was a programmer's editor for RPG and related software, but it also had support for most other languages. It did the syntax highlighting, had a source file member directory tree on the left side (double-click to download) and one-click compiles. It too used the Events files (as does RDi) to retrieve the compiler error messages. A double click on the message windows and it too jumped to the line with the error. Visual RPG actually did it better because it didn't leave the errors embedded in the source until the next compile.

My point here is that trying to get people to use RDi by referring to it as "modern" vs SEU is a really stupid idea. Why? The AS/400 was announced back in 1988 and arguably was the last time SEU (Edit Source) was substancially updated. IBM did CODE/400 which was a good first-effort at a Windows-based (it was original OS/2-based) RPG editor. Visual RPG was introduced in 1992, some 4 years after 5250-based SEU. RDi has virtually nothing new compared to Visual RPG, so how is technology that is 4 years younger than SEU considered "modern"? It's not.

The fact, most (all?) major platforms have gone through several iterations of Development Environments--today they are all based on the supported User Interface of their host operating system. IBM made the wrong decision back in the mid-1990s when it decided it could not do a native GUI for this operating system. So we have 5250 or "green screen" as its called. But to try to convince people that SEU which is 26 years old, (I know it came out in release 2 of CPF on S/38, but I'm making a point here) is some how worse than RDi which uses technology that is 22 years old is like saying "You need to move to this new car based on the 1992 Buick Skylark because the one based on the 1988 Corvette is old technology".

All this continues the point, that IBM does not now, nor has it ever, known how to market anything.



Return to midrangenews.com home page.
Sort Ascend | Descend