home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.vms
- Path: sparky!uunet!paladin.american.edu!news.univie.ac.at!hp4at!mcsun!sunic!seunet!enea!sommar
- From: sommar@enea.se (Erland Sommarskog)
- Subject: Re: Plans for mixed Alpha/VAX cluster with common system disk?
- Message-ID: <1992Nov22.233249.8288@enea.se>
- Organization: Enea Data AB
- References: <27A00A11_00168E00.00963B3804E014C0$437_1@UK.AC.KCL.PH.IPG>
- Date: Sun, 22 Nov 1992 23:32:49 GMT
- Lines: 25
-
- Nigel Arnot (SYSMGR@IPG.PH.KCL.AC.UK) writes:
- >If DEC don't do something like this, I'll be very disappointed. Also it'll
- >be a big mistake: the confusion caused may well result in Alpha+VMS getting a
- >bad name and even loss of customers to the competition.
-
- Competition with whom? To Sun, who have had Sun-3 (68xxx?) and Sun-4
- (RISC) since God knows when.
-
- I must have been working too much in the Unix world lately (shudder!),
- but I fail to see that this is any problem.
-
- >If they can't do this, there's a trivial confusion-avoider. Simply create
- >file types .AEXE, .AOBJ, and .AOLB for alpha code, plus qualifiers /VAX and
- >/ALPHA on utilities like LINK and LIBRARY, which default to that native
- >for the processor on which they are running.
-
- And break a lot of programs, command files, MMS files and so. NO THANK YOU!
-
- >However, it would be much, much nicer to have the ability to create EXEs
- >that contain both sets of code and map and execute that appropriate for the
- >processor on which they are run.
-
- And then when someone ports VMS to a third architeture?
- --
- Erland Sommarskog - ENEA Data, Stockholm - sommar@enea.se
-