Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!iuvax!uxc.cso.uiuc.edu!uxc.cso.uiuc.edu!m.cs.uiuc.edu!hummel From: hummel@m.cs.uiuc.edu Newsgroups: comp.sys.amiga Subject: Re: Ami Fantasy Wishlist Message-ID: <7200126@m.cs.uiuc.edu> Date: 23 Aug 89 04:28:00 GMT References: <6203@ingr.com> Lines: 78 Nf-ID: #R:ingr.com:6203:m.cs.uiuc.edu:7200126:000:4932 Nf-From: m.cs.uiuc.edu!hummel Aug 22 23:28:00 1989 Written 7:02 pm Aug 21, 1989 by phil@ingr.com in comp.sys.amiga: > I have been reading the wishlist for ADOS 1.4, Amiga 3000- through 10,000 etc > 8-), and the UNIX, MINIX, and ADOS discussions and decided to add my three > cents worth. 8-) 8-) 8-) > > 1. Operating system - VM look-alike with close adherence to JCL standard. > 2. Editor - What else - TSO ( why give up now). > 3. Communication - RJE and Bisync ( full steam ahead). Oh yeah, a 300 > bps asynchronous line for connecting to OTHERs. > 4. Graphics - PLOT10, an established standard. > 5. User Interface - In this environment, We're tough, We don't need no > stinking user interface. > THIS IS NO FANTASY! where's FREDDY? He must be around in this dream somewhere. Don't forget: 6. Scripting Language- ReXX. Whoops; "It's in there!". Am I bordering on sacrilege here, or what? Alright, now here is what I REALLY would like to see in an A3000. This is going to be Commodore's keynote '030 machine, so it needs to set a standard that will keep up with further improvements in technology. - 68030 at 25Mhz, on an asynchronous "fast" bus, with a socketed crystal. Designed for operation with '030's up to 50MHz. 68882, clocked independently of the CPU. Some external caching hardware would be a nice addition, and might fall out of the work that will have to be done for compatibility with existing Amiga architectures. An interleaved memory scheme would be nice, though expansion granularity might be too high, even for 2-way interleaved. The control of advanced memory architectures may be best left up to future microprocessor designs, but I think the best computers will go far beyond the capabilities already possesed by the MPU. This is, in fact, one of the things that attracted me to the A1000 from the start. I doubt anyone ever has or ever will push a lone 68000 further than the crew from Los Gatos did. I have high hopes that the crew from West Chester will set the same high standard for the 68030 (you listening, hazy ?-). - Give those custom chips more bandwidth !-) What could you do, besides drive expensive, high resolution, very-large-palette video displays? Oh, well, perhaps target high-bandwidth access to chip RAM as part of the expansion architecture so that future video adapters could pull (and insert!) displays right out of (and into!) memory. That would sure blow away all of those folks trying to do real-time video on workstations! - SCSI on board. There are no excuses, and I don't expect to hear any. - RS-422 on board. This can always be cut down to RS-423 (~RS-232), so there's no reason not to provide top o' the (differentially-driven) line service! Who knows? Get someone to write the software and you could even co-exist with MacDrivel on Apple/LocalTalk. - Ethernet on board. This is 100% necessary for a machine in the A3000 class. At the outset, anyone who will want (and can afford) this machine will want Ethernet (less the few deadbeats :-) who want to use Novell's proprietary networking technology). Let THEM buy a board, but let this machine talk Ethernet out of the box. If you're worried about the effect on price, take it out in some "cost-reduced" '030 design, but it's a must in a machine that's going to break new ground. - On the system software end, start working something along the lines of Mach into the picture. The Amiga is too good for SysV and even SVr4 - it is a dinosaur. Do what you must for business/government, but you can bet that the world is headed towards UNIX, networks and distributed processing, and the ability for computers to work in any medium. Exec/AmigaDOS were ahead of their time AT ONE TIME, but Commodore had better be seriously plotting another major step forward. For the interim, they might be able to do some neat things with their Janus technology, though this will eventually prove a dead-end unless they effectively generalize it into a system for heterogenous processing, with hardware support a cut above the existing BridgeCards and radical developments in the design philosophy of Amiga application software. - If this is another two-layer motherboard, I will barf. :-) :-) :-) I'm looking forward to being impressed by Commodore next year. If nothing else, I want good reasons not to upgrade to a NeXT (or how 'bout one of those great new HP workstations?) when the time comes. The technology is still expensive, but prices fall as fast as newer machines are designed. < Lionel ---------- Lionel Hummel 409 Kenwood, Champaign, IL 61821 University of Illinois, Urbana-Champaign [H] (217)356-6379 [W] (217)333-7408 hummel@cs.uiuc.edu {pur-ee,uunet}!uiucdcs!hummel BIX: lhummel