This letter was posted by Bruce Campbell on Feb 5th, 1998.

While the conversion from Watstar to Waterloo Polaris proceeds, many are asking "Why are things changing ? It worked before".

Watstar has always been an evolving network, right from the day it was first installed in 1984 running DOS 3.0 until today, running both Windows 3.11 and Windows 95. As operating systems, applications programs and needs advance over time, Watstar advances to run the new software and meet the new needs. This is not always a painless process. People often don't like change, so it is a tough decision for when to move to new software, and discard the old, but it has to be done... eventually.

Some may ask "Why discard the old at all ? Why not move to the new software, and keep the old ?". Contrary to what you may think, as your computer magically seems to keep on working all by itself, large computer systems don't run by themselves. There is a team of system managers, technicians, consultants, etc. all working to keep UWs' computers working. As new software and operating systems appear, computer departments are expected to become expert in the new systems, and make them available. To keep supporting the old systems, while supporting new systems requires an ever increasing number of staff, or - and this is the tough part - a decision must be made to discontinue supporting old systems. (The alternative to never support new systems is possible too, but UW may lose its' reputation of excellence at some point).

Certainly, many can install their own standalone computer, do their own backups, buy their own new software, keep their own printer full of paper, and make their own decisions as to when to change things. But, as soon as they want to share files with others, run the same software as others, print on network printers, and have problems solved by others, being on an island starts to look less appealling. Having full control also means having full responsibility.

The conversion from Watstar to Waterloo Polaris started September 1997, and will be completed September 1998. This move to Windows 95 hasn't exactly been an aggressive one for Engineering Computing. Windows 95 was released early 1996, and it is now two years later. Engineering Computing didn't start the conversion process right away, but it couldn't wait forever. Users want and need the new features of Windows 95, and Windows 95 applications.

But why not allow both Windows 95 and Windows 3.11 ?

It's the support cost. Support isn't free - it requires staffing. The two systems are so dramatically different at the network level, that supporting both is equivalent to halving our support ability. The deadline of September 1998 is important, since some network advances in the works cannot be released until all systems are converted. The network advances will allow for a much faster login time, convenient (and standard) remote access to the users N: disk, and improved printing speeds, among other things. There is a cost with the move to Windows 95, because of its' increased hardware requirements. There is a cost in supporting both Windows 3.11 and Windows 95, in terms of staffing requirements (or reduced support ability). There is a cost to ignoring Windows 95, in terms of the impact on the credibility of UW as a leader, and the loss of efficiency as faculty/staff/students use outdated software, and encounter difficulty sharing files with users of modern systems.

There are three choices, none of which is right, and none of which is wrong, but moving ahead at a reasonable rate seems the best choice, so we're going to Windows 95 over a one year period, two years after the product first hit the market.