At some point in February 1992, shortly before I was promoted to DP, Magic instigated a complete rewrite of the entire board. The plan was to introduce new features, tidy up old inconsistencies and bugs, and generally produce a brand new, stable, and altogether better version of the board. This project never really got off the ground, for a variety of reasons, not least of which was the announcement that the central VAX/VMS service was to be replaced with a collection of (Silicon Graphics) Unix systems.
Mike Brudenell and Laurence Tyler, of the Computing Service, kindly arranged for the provision of an account on tower, the central Unix compute server, and work on Jorvick V gradually started. Unfortunately, due to various other pressures (my finals, ERiK's DPhil, Magic's job) this didn't progress much further than its VMS-based predecessor.
The demise of the VAX fast approaching, and realising that having even a hastily-cobbled-together, hacked, unstable version of Jorvick available on the Unix systems was preferable to none at all, I ported the existing Jorvick II code, ripping out any subsystem which relied too heavily on low-level VMS functionality. Talker went out the window straight away - all the interprocess communication was done by means of VMS mailboxes, rendering it highly non-portable.
Unfortunately, insurmountable file-locking issues rendered this version unstable in a highly unpredictable manner, so the plug was pulled after a few days. No alternative version of the software having been produced, Jorvick went down with the sinking VAX.
Date: 2003/09/04 17:03:17