Path: utzoo!attcan!uunet!husc6!purdue!decwrl!nsc!stevew From: stevew@nsc.nsc.com (Steve Wilson) Newsgroups: comp.arch Subject: Re: stack machines (Burroughs) Keywords: RISC, real-time Message-ID: <5158@nsc.nsc.com> Date: 10 Jun 88 16:57:57 GMT References: <1521@pt.cs.cmu.edu> <1532@pt.cs.cmu.edu> <476@pcrat.UUCP> <2868@louie.udel.EDU> <370@dlscg1.UUCP> <3147@polyslo.UUCP> Reply-To: stevew@nsc.UUCP (Steve Wilson) Organization: National Semiconductor, Sunnyvale Lines: 68 In article <3147@polyslo.UUCP> dorourke@polyslo.UUCP (David O'Rourke) writes: >different versions. The MCP is such an old design that a recent design >document estimated that it would take 180-200 man years to change the way >sectors were handled. Right now MCP assumes 180-bytes per sector, well so >people wanted that to become a variable sector size, and that is what is >going to take them so long to implement. Not exactly modualar software. There are hysterical reasons for having 180 byte sectors. As I recall even the ancient and honourable B1000 series used the same sector size implying that it is a company media standard. It maybe time for a change in such things, but when you're fighting 25 years of history there is alot of momentum to do things a certain way. > > But for the first time in years Unisys is getting the "new breed" of >programmers in, and most all of the people that I was hired on with are out >to change the way Unisys does software. Most of the new people I worked >with were as upset as I was about the lack of software tools, and they are >working rather quickly to fix it, just before going back to school I was >playing with an editor that I'd actually consider using. It was written >by one of the new programmers in his spare time and frustration with the >current editor. > Now CANDE, etc ain't that bad! Just compare it to what was available before such things. I KNOW you haven't had to use PUNCH CARDS! This again isn't saying there isn't alot of room for improvement. Heck, I was yellling and screaming at the large systems human interface clear back in 1979! Just to get sentimental on ya, when I was in school we had this old, ancient 360 model 50. It could run a couple of jobs at one MAX. The Burroughs MCP's where multi-programming at the same time that OS/360 was released back in the early 60's. There was a point in time where the MCP was a beautiful thing to behold. The ALGOL compiler could compile a couple of thousand lines a minute versus the stuff that was available else where which were running in the hundreds of lines ball park. Now, I've heard some friends of mine, who I have it on good authority where highly responsible for much of the early work on the MCP lament about the monolithic structure of the MCP. Hind sight is always 20/20 you know! > Also people are taking a good hard look at MCP and wondering what to do >about it. So you might see some significant changes in the next few years. >But I really don't know, it was just mostly shop talk at the water cooler, >but the ideas, and frustration with the current system, are there so there >might be a change. > > Anyways there's a recent Unisys employee's observations > > If anyone thinks I speak for Unisys they need some mental help!! > >-- >David M. O'Rourke > >Disclaimer: I don't represent the school. All opinions are mine! Well, David, maybe you can help bring an old architecture back to life. Besides, Mission Viejo is a MUCH nicer place to live than say... Pasadena! Steve Wilson National Semiconductor [EL grad '79] [ Universal disclaimer goes here!]