Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!willett!ForthNet From: ForthNet@willett.UUCP (ForthNet articles from GEnie) Newsgroups: comp.lang.forth Subject: FPC forth????????? Message-ID: <542.UUL1.3#5129@willett.UUCP> Date: 24 Feb 90 21:54:31 GMT Organization: Latest link in the ForthNet chain. (Pgh, PA) Lines: 84 Date: 02-24-90 (01:45) Number: 1754 (Echo) To: MARK SMILEY Refer#: 1725 From: MIKE SPERL Read: NO Subj: BRECALL Status: PUBLIC MESSAGE 1Mark, > An excellent first step. I am eager to see the results. Sorry to bbe away so long. Have my 386 running now, and have DOS 4.01 installed and have been reading all the directions, so had to put aside the Project for a while, but I'm finally back in businness. >I noticeed in your note to Pete that you are getting a 386 machine. >Sounds nice. If you don't mind my curiosity, what >kind/confiiguration did you decide on? I got a 20MHz AMI motherboard with 4MB dram (takes up to 16) and a 64K sram cache as well as a 387. The AMI bios is something else: built-in diaggnostics with windows, menus; analyzes the hard disk for bad sectors (and reported the same addrresses as the factory in the case of three hard disks I tried with it), does the lo level format and automatically tries various iinterleaves to find the best (2:1 for my two disks), tests the floppy drives, vga boardd, communications board, etc., etc. ... The board accepts any 8 or 16 bit peripheral so I can use all my old XT-clone boards with it, even the ramdisk, which I tried out to see if it would really work, aand it ran flawlessly (but it's now in my wife's xt-clone; who needs it with 4MB ram?). The i/o from the peripherals is cached so the cpu sees only 0 wait state 32 bit sram when the cache is enabled. I have to slow it down to 8MHz for VKAL! They make the same board to run at 25 and 33 MHz, but this is the best buy at 1300. If you decide to upgrade (you should!)), I'll tell you where I got that pprice. The good news: I backed up everything seeveral times before trashing my old computer. The badd news: I have so many backups I'm confused about what goes with what, and it'll take a while to sort everything out so i'm back where I was two weeks ago. I finnallly got around to upgrading to f-pc ver. 3.5, and got most of my tools installled in it. New-expect even workss with Tom's editor now without forcing you to do a warm boot to get bacck to forth. And I love his new browse/view system, but miss the editor's help screens that accompanied previous veersions (I don'tt use it enough to be fully conversant with aall the keystrokes). >> there's a place forr your header information without disturbing >> the gif fformat. > Great! Is there a limit on the sppaace, or is it variable? Set it up any way you like. The size is arbitrary, because a special mark '!' and count identifies it and a gif viewer that doesn't understand it will just sskip over itt. If you want to work on it, I'm uploading the front-end code in DGIF1.ZIP for you to examine, and you'll see where your header goes. The setup you sent me will be just fiine, or,, if you wish, you can modify it. Hope I didn't cause you any trouble suggesting a change in DEFER. I've looked into it and the change needs to be made, but it will take some doing. DEFER is defined no less than thrree times during metacompilation, and words are deferred before DEFER is defined! So we'll have to see which defer is doing what and change them all, including one in the metacompiller itself. Maybe someone more familiar with the metacompiler could be interested to help? Meanwhile, I thought consideration might be given to a more effficient jump table mechanism to direct a program to the correct code, eliminating the use of defer if possible. What do you think? Seems to me DEFER is a bit overused in f-pc. Between the timee one pressess enter and the time 'ok' appears, there are at least a dozen deferred words! I single-steppedd the code to see. And there are 45 deferred words (found by fgrepp) just in the kernel! And about 300 in the whole systtem! For 2000, that's every sixth wworrd! - Mike - ----- This message came from GEnie via willett through a semi-automated process. Report problems to: 'uunet!willett!dwp' or 'willett!dwp@gateway.sei.cmu.edu'