lantz@Pescadero.UUCP (02/19/87)
I recently sent a message in response to Daniel Conde's request about a Unix-compatible version of V, but have reason to believe it didn't get posted outside of Stanford. Therefore, let me reiterate: As of summer 1985 I had an ``almost there'' version of V that eliminated most of the remaining incompatibilities from UNIX -- including providing fork, signals, and shared file I/O -- such that even more applications could be ported without change. This is in contrast to what we, Tektronix, and presumably all other V user communities out there are (still) doing -- porting applications to V by modifying the applications, however ``straightforward'' those modifications might be. There were two motivations for this project: 1) to make V more marketable (not for money so much as wide-spread use) and 2) to provide a base for getting REAL performance comparisons between a kernel-based UNIX and vanilla UNIX on the same hardware running the same application suite. Alas, this project was terminated due to the return of its implementor, Alan Hastings, to Bell and lack of interest on the part of prospective funding bodies. If someone out there is really interested in finishing this project, get in touch.