In this paper I discuss VM's strengths (and weaknesses), and also argue that VM is a unique resource for IBM - and important to IBM's future - and describe steps IBM must take to preserve it.
There have been a lot of discouraged people in the VM community in the last few years, and comments that VM may be doomed to wither away. These perceptions result from events that emerged in the late 1980s: IBM's complacency and inactivity in the face of changing technology, arrogance that led to ill-considered policies, and the appearance of competitors in traditional markets.
Most important, personal computers and workstations came to dominate traditional VM/CMS application domains of end-user computing, "office", and program development. IBM didn't understand that their existing systems had to evolve to cooperate and compete in this environment, and let their market and technological advantages evaporate.
Despite this erosion, VM remains an extremely valuable resource, with unique capabilities available on no other computing platform. In particular, VM has evolved from a best-of-breed interactive system to a best-of-breed platform for building and running server applications. VM has remarkable gifts to offer in many domains, is underexploited in most institutions in which it is used, and could be extremely valuable for many sites not running it. Many institutions that tried to "get off VM" found it very difficult to replace VM's functionality and low cost.
As a result, VM continues to be very successul in a wide range of applications, and has had a number of surprising new opportunities due to the advent of the World Wide Web, and the failure of many "legacy client/server" efforts. This has caused some institutions to reevaluate plans to "sunset" their plans to get off mainframes.
IBM has always had an ambivalent attitude towards VM. One of VM's problems for IBM has been that it's just too efficient; IBM would much rather that customers use a "strategic" platform that just happens to require much more expensive iron to run. In later sections of this paper I'll discuss IBM's attempts to move customers to inappropriate, and often inferior technology.
As a result, IBM has always had an element that wished VM would disappear. and predicted VM's imminent demise many times in VM's 25-year history. These predictions turned out to be foolish, and VM has outlived many of its detractors within IBM. This has also had the effect of alienating many IBM customers, and reducing IBM's credibility.
Incidents of IBM staff telling customers that "VM is dead" have been reported since VM's early days. I don't think IBM realizes that insulting parts of its product line damages its entire brand reputation. This type of unprofessional behavior does not help convince me to buy anything from IBM. Fortunately, this type of self-inflicted wound seems to happen less frequently now, though a lot of damage was done. If you see an IBM person badmouthing VM, report it to the VM management team, and they'll see that it is dealt with appropriately.
IBM's traditional lack of marketing support for VM has cost it dearly in the past and is hurting it very badly now. In the mid-1980's, IBM had the chance to make VM into the force Unix and Windows subsequently became. VM's technology and customer base, and the market domination IBM enjoyed then, could have resulted in VM being as influential and widely used as Unix is today. IBM wasted this opportunity, losing billions in revenue that have gone elsewhere. If you think this is an exaggeration, consider the growth experienced by the then-tiny Unix market, or consider how "e-mail" used to be synonymous with "PROFS". IBM owned this lucrative market, and gave it away to other vendors.
Fortunately, IBM's VM development organization has carried on enhancing and promoting VM. Contributions like the brilliant CMS Pipelines by John Hartmann (of IBM Denmark) have helped keep VM a modern, high productivity environment. Developments outside IBM have also caused new use of VM, most notably VM web serving, pioneeed by Rick Troth and now a competitive marketplace with multiple web products for VM.
Today, VM is a niche environment, though still a profitable, sizable and productive one. Many customers that could be retained on VM or attracted to IBM equipment running VM have moved to non-IBM equipment. More will do so, unless IBM reverses this neglect.
There is good news: the downsizing trend has bottomed out. Many sites have suffered considerably trying to get off VM, and have abandoned their "get off VM" or "get off mainframe" attempts. VM is a profitable business for IBM, and a secret weapon for this institutions smart enough to use it (You can't derive strategic advantage from computing if you use the same technology as everyone else.)
Low cost and flexibility make VM highly attractive for delivering computing value. VM's superior application environment, with unique, powerful, and well-integrated services like REXX, CMS Pipelines make it possible to quickly create sophisticated applications. REXX sockets and CMS Pipelines in particular led to the VM web servers now creating a renaissance of VM activity. VM is by far the most suitable mainframe operating environment for web activity. New VM functions, such as OpenEdition and ADSM, and the forthcoming Java and NetRexx support, make it interoperate better with Unix and PC systems, and provide important value-added services.
VM is IBM's only plausible offering for many applications, not AIX, OS/2, OS/400 or OS/390 (MVS) or Lotus Notes/Domino. IBM must market and enhance VM's considerable strengths. To do otherwise will increase the erosion of IBM's customer base to other vendors. My experience is that applications leaving VM generally leave the IBM product line altogether.