Where they've gone: TV George, on from HP
3000 pro uses open source version control

Can a new IDE push a migration forward?

We're looking for evidence of this concept in the 3000 community:

"If you get more modern development tools, then a migration to a new platform is worth the effort. Well, more so."

We've received reports from the user community about what they count upon when they need to code for their 3000s. Some answers have run to Whisper Programmer Studio (a fine piece of software that lost its vendor, WhisperTech, years ago) or the QUAD Editor. Even EDITOR/3000 got some votes in our simple poll.

A robust editor -- Robelle's Qedit, ready for both 3000-hosted use and Windows, cited most often -- is a great tool for keeping HP 3000 software maintained. When a company grows fast, however, the change sometimes seems to spark a major uptick in the demands on development. The next thing you know you're being challenged to support mobile versions of applications, integrating with existing apps that were developed elsewhere on other platforms, or tying into expansive Integrated Development Environments (IDEs) in the company that acquired you.

We're thinking Eclipse there, but the product names don't matter as much as the concept.

Those larger companies and wider scopes of app delivery can trigger a need for a bigger IDE. We're pretty sure of that, but just missing some stories that it's been true in the 3000 world. We're on the lookout for reports that the bigger IDE can help push a 3000 migration down the slide. At customers like SBCTC (that college consortium in Washington State) and at an Australian insurance company, 3000 sites became Unix and Linux operations. Did they base their decision, in part, on getting a chance to use big IDEs for growing application development needs?

Most of the migrations we have tracked were sparked by two things: a loss of faith in the 3000 because HP's support was ending; or the need to have modern hardware that was still being upgraded by the manufacturer. The other most-cited reason was a departure of 3000 expertise from the IT staff. Also near the top of the list: application software vendors, selling the off-the-shelf solutions, then leaving the MPE marketplace.

But it stands to reason that modernizing what the vendors call a "legacy server" environment includes a shot at more elaborate development tools. The 3000's architecture was simple and elegant in its initial generation. Not much has changed to expand tools into an Eclipse-like scope. 4GL suites came closer than anything, when you recall the modules from Speedware and the Powerhouse tools. It seems an extensive environment has never been needed at most 3000 sites, based on what we've heard from the homesteading customers.

The sites who have made their transition, however, are the ones who could report about the value of using these Eclipse IDEs. Visual Studio becomes the tool that's mostly to be employed by migrated customers. For a former 3000 developer, these IDEs would have to be very friendly to COBOL -- because that's the language in almost universal use among 3000 home-grown applications.

Did you acquire better tools for development -- or bigger ones, anyway -- when you made your migration? The most interesting question of all might be: Did this extra development capability have an impact on your decision to migrate?

We've said in the past that one community analyst considered this "buying a new car because of its tires." That seems unlikely. But maybe it's more like "buying a new car because it's got an integrated GPS navigation system and a hybrid engine." You can go further on that kind of vehicle with the same energy cost, and have a much better idea of what's all around your environment.

Tell me (email to [email protected]) if the concept has made any sense to you -- if you're migrated, or are nearly ready to cut over. I would love to talk to you (512-331-0075) as well.