darrell@sdcsvax.UCSD.EDU (Darrell Long) (10/18/87)
School has started at most universities, so all researchers should be back and available. It is time to revive comp.os.research. When I was at DCS in Berlin, a lot of folks came up to me and said, "I read comp.os.research all the time, but what happened to it -- there's no traffic anymore, is it dead?" No, it's not dead, but we have suffered from a lack of submissions. Certainly as researchers we can't be content with the state of things: 4.4 will be the last BSD, Bell Labs isn't releasing any of its research UNIXes -- I'd like to hear from the (MACH | Sprite | DASH | Clouds) folks about this. SIGOPS is coming up, perhaps someone who is going could summarize? There was just a workshop on large grained parallelism, I know several folks who were there -- perhaps one of you could summarize? For my part, I'll try to work up a summary of DCS for those who could not attend. Other things: I'm told that DEC is really not cooperating with universities like it used to do. They really want to sell ULTRIX, so they've made it harder to get machines that will run BSD. They won't give out bus specs on the new VAXBI machines, so BSD won't run on them. This is all very interesting to us -- research is great, but it is not really great unless we produce something that ultimately makes it "out there" and gets used. Could it be that we are returning to the "good old days" of proprietary operating systems? ULTRIX as the new VMS, an interesting idea... The cooperation we get from manufacturers greatly influences how research projects progress -- reverse engineering bus specs is not fun. Could there come a time when most research is done on machines built by small companies willing to give out specs? What does this say about the commercial viability of the products of research? I've been raving too long, but I'd like to hear some speculation in tose directions. comp.os.research is NOT dead. But there are too many readers and not enough submitters. Take a few minutes and voice your opinion on the directions that OS research should take. Certainly everyone cannot be content with a future dominated by System V.