All technology carries the same lifespan: the period of time when it produces results, effectively better than alternatives. COBOL stands as a popular example of language technology that's outlived many of its gravediggers; we may never see a time, at least in our lifetimes, when COBOL code fades into the oblivion or irrelevance predicted for it.
Too often, technology gets written off in an ageist attack. Anything that is older, the tech mavens and marketing princes tell us, must be not as good. Wine is an exception. Perhaps one of the HP 3000's languages other than COBOL will be counted into that older-is-better club, and soon. The 3000 community, migrating or homesteading, can help.
Alan Yeo's ScreenJet Ltd. is polishing up a new technology to give fresh life to the many modules of Transact code in the world. Not so many as COBOL, to be sure, but then Transact was an HP product rather than an industry standard. HP purchased Transact from its creator David Dummer. The new future in TransAction, a product ScreenJet is polishing could save HP 3000 shops a lot of migration money.
ScreenJet's Yeo is looking for a few Transact customers to help fine-tune what's new for users who like the language launched in the 1970s. The technology proposes to give applications and modules which employ this fourth-generation language a new lease on life — and some of the target platforms to use Dummer's new product could be HP 3000s, too.
ScreenJet collaborated with Dummer in 2005 to create T2C, a Transact to COBOL conversion toolkit. But some customers don't want to embrace COBOL any closer. They just want their Transact modules, surround code or even applications to continue to do the work they perform. On another platform, perhaps in time, but most immediately, on the HP 3000.
You would need TransAction to cut across multiple environments, the kinds that the 3000 community is targeting. Windows. Unix. Even Linux. Dummer wrote Transact for the 16-bit HP 3000, the old "stack architecture" that because 32-bit PA-RISC. Now even HP is walking away from PA-RISC, toward the Intel chips and even AMD's architecture. Why should HP be the only company to push out its compilers for ascending environments?
If the prospect of testing out a new chapter in the Transact saga seems interesting — and there is no sales pitch involved, Yeo assures us — Yeo invites 3000 customers to contact him at [email protected]. A test run of a technology that could forestall a rewrite, or perhaps eliminate the need to find an expert on the nuances of Transact — that sounds like a way to keep a language productive, even at an advancing age.