Highlighting IBM’s somewhat erratic position on its own Systems Application Architecture strategy, the veteran DOS/VSE operating system – which recently celebrated its 25th anniversary – has been given the precious label of strategic – an adjective seen as a virtual guarantee of survival and one usually reserved for the SAA offerings of which DOS/VSE is not a part – well it really depends on which IBMer you spoke to last. According to Computerwoche, the kiss of life comes just a few years after IBM looked on the brink of doing away with the technology, but now it looks likely that a new version of Virtual Storage Extended will be introduced this year under the name VSE/ESA – it was due this summer. VSE product manager Volker Lhr puts the U-turn down a change in thinking vis-a-vis the strategic value of VSE that was mainly a response to user demand – in other words users told IBM give us what we want – and that’s not a cut-down MVS, or we convert to DEC’s VMS. Around 1986, Lhr continues, we thought we could migrate the 22,000 VSE users up to MVS because both systems had similar strengths. But despite certain enticements and higher functionality, the majority of VSE customers were clearly not ready to move to the dearer and more expensive system – as Lhr puts it, we realised that not everyone could afford a Mercedes, and many were quite happy with their VW Golf. And now, despite the fact that many customers use their VSE system with 3090 mainframes while IBM has specified the lower System 370 environment as appropriate, and IBM has finally acknowledged the legitimacy of VSE with the designation strategic. Putting a slight dampener on the VSE users’ celebrations, however, IBM has stopped short of giving the system the full SAA-compliant stamp of approval: its attitude on VSE in the context of SAA is still summed up by a footnote that accompanied the announcement of VSE Version 4 in 1988 – namely, that VSE will only participate in SAA when necessary. And IBM is still set on encouraging its VSE customers to make what even Lhr describes as the quantum leap to MVS. The MVS carrot is now applications portability between the two systems, with the main goal being to allow CICS applications to migrate one for one from VSE to MVS.
