Subj : FreeBSD To : Roy J. Tellason From : Neal Robles Date : Tue Nov 20 2001 04:51 pm Hello Roy, -=> Quoting Roy J. Tellason to Neal Robles <=- NR> Although I quickly switched to Linux as soon as I got a NR> distribution (FT-Linux from What PC? magazine), I still have good NR> memories of Minix. It was my first exposure to Unix, running on a NR> dual-360K floppy XT without a hard disk when I couldn't afford a NR> better computer. RJT> You can actually run it on an XT? How well did it run? It was pretty zippy. Multitasking too. I never got to install it on a hard disk, because I don't have a reliable MFM HD anymore (anyone here discarding a 20 or 30-Meg XT HD? ;->). The version I used (1.2) might not have been practical for getting real work done. One flaw was the static memory allocation stored in each program's header, which you had to manually change if the program crashed or behaved oddly. Since the 8086/8088 didn't have memory protection hardware or a memory-management unit (MMU)--even the 286 didn't have paging--you can't really run a production multitasking or multiuser OS on it. NR> Minix reminds me of the circa-late 70's personal computers. The NR> computers were simple enough and well documented enough to let you NR> get in and learn everything and possibly modify the design or come NR> up with your own. RJT> That's one of the things I really liked about CP/M, that you could RJT> get a fairly complete grasp of *all* of it. Completely and totally, RJT> and you could also do a heck of a lot with very little, small RJT> executables, and so forth. But I started wanting to do too much in RJT> terms of actually *using* computers and too many of those things RJT> weren't possible on that older and more limited hardware and OS. The RJT> same thing started happening with dos, and I sure didn't see putting a RJT> clown suit on it as being much of an improvement, instead demanding RJT> and using up more resources for that fancy interface... RJT> And that's a lot of the reason why I've moved on in this direction. Although I've read a lot of good things about CP/M, and even read an article about it in an old issue of Popular Electronics (with discussions of the CCP and TPA?) I never actually got to use it. My hands-on microcomputer experience started in the pre-CPM era on a Radio Shack TRS80 Model 1 with no floppy drive and 8K of free RAM. On a bad day, you might get only 4K. I programmed in BASIC and assembly language. I didn't actually use an assembler, but rather a monitor. A monitor is like having a permanent copy of DOS "debug" in ROM. This is one feature that I really miss in all the newer computers. Later on, the quality of life improved immensely when the Trash 80 got a 32K expansion unit and a single external floppy drive. However, there was always a screwdriver on the table beside the floppy drive because every few minutes, the drive belt would slip off the drive wheel and we'd have to open up the box to put it back. Those were the good old days.... CP/M didn't exist yet, we used TRS DOS at the start, then later, NewDOS by Apparat. RJT> I can remember a program that I ran across that called itself "UZI". RJT> This stood for "Unix Z80 Implementation". I thought it rather odd that RJT> this "OS" claimed half of the required 64K of system ram for itself, In an 1975 issue of Byte (the 2nd issue?) there was hot discussion about whether an OS was even practical at all, given the limited RAM most systems then had. One reader's letter complained about Pascal requiring 130K just for the compiler. Anyway, by the time Z80 systems and CP/M were king (the pre-IBM era), my interests had moved to other directions. I think I would have enjoyed what you experienced. What system were you running CP/M on? Regards, Neal ___ Blue Wave/386 v2.30 [NR] --- Platinum Xpress/Win/WINServer v3.0pr5 * Origin: The File Bank BBS! Makati. Philippines +63-2-896-3116 (6:751/321) .