Dr Wheeler gave Club 100 permission to post all of his Model 100 work. EPROM1.THD --- Copyright 1988 by Phil Wheeler An original compilation of Compuserve Model 100 Forum messages for use by Forum members only. The Model 100/102/200 are designed to accept an auxilary ROM. Several commercial ROM's are available (e.g., Super ROM and Ultimate ROM II) and have been proven popular -- and there are custom ROM burning services advertized. It is possible to "roll your own", and users here have done so. These messages discuss the desirablility of programming a "math ROM" for the Model 100. A companion file (EPROM2.THD) discusses some of the "techno-mechanical" issues involved. Message range: 166820 to 166974 Dates: 4/1/88 to 4/4/88 Sb: #a "math" rom? Fm: SCOTT T. SCHAD 73720,1166 To: All I'm considering tackling a "math" rom for the model 100, similar to the math and curve-fitting roms for my HP-71B. It seems to me that lots of model 100 applications are engineering/controller-oriented, and the 100 sorely needs BASIC commands for trig functions in degrees, matrix math, base 10 logarithms, etc. But my question would be: how can I make the routines available to BASIC? I'm a m/l novice, and while I can deal with separate .CO files I have trouble visualizing a way to hook into BASIC and enhance it. Any comments, discussion, pointers to previous work would be appreciated. Followup: a Forth rom seems like another real possibility, given the high caliber of the Forth implementation to be found in the DL's. The HP-71B is, to my knowledge, to only portable computer that can take Forth in rom, but it is nibble-based and hard to use as a learning tool. A model 100 Forth chip coupled with Brodie's Learning Forth would be THE ideal teaching tool, especially since a page-by-page comparison of his book with M100 4th appears in the DL's. Forth would at least let one "get at" the machine in a better way than BASIC and a less demanding way than Assembler. Many of the higer-level math routines I am considering implementing are commonly found as Forth algorithms. Of course you'd have to work out compatibility with the new Tandy disc drive, but at least with a rom-based Forth you would have room for more than three screens. Fm: JAY HOLOVACS 74756,413 To: SCOTT T. SCHAD 73720,1166 SCOTT (these are off-the-cuff comments not really thought out, but might be of interest to you.) A forth kernal in ROM could be very interesting there are a couple of forth adaptations in the DL's but the effort of porting fig forth to rom may be worth it. If you get the permission of fig, there 8080 model should be almost a perfect fit, with changest only for m100 file structure. the TDD does not favor track and sector reads/writes so the easiest bet would be to use .do files as text streams rather than have a true screen oriented forth. Compiled programs could be saved as .co memory image files. some suppliers (such as Traveling Software) offer a rom burning service, possibly they could involve some of the m100 expertise required for such a project (bank switching is not for the faint hearted) if you can find enough takers to make the costs worth while. Good Luck Jay Holovacs Fm: SCOTT T. SCHAD 73720,1166 To: JAY HOLOVACS 74756,413 Using .DO files seems like an ideal solution, given that the editor is already in rom. I have nothing against screen-oriented Forths, (I know I should use the term "block-oriented" now), but having used a couple of Forths on the Macintosh and one on an Amiga that digested standard text files I much prefer it. I frankly am not knowledgeable enough to tackle the disc I/O problem...but I have downloaded all the Forth files in the DL and find they furnish 90% of the work which would be required for an implementation. There may not be a large market. But I can think of no better play toy than a model 100 with Forth in rom. It might be possible to port lots of good software to the model 100, if the Forth model was rigiorous enough. The HP-71B forth chip is "nibble-based", using 4-bit "words" and 20-bit numbers, etc. It is just non-standard enough that learning the language on a 71B is tough. But the speed increase is remarkable. Comparable 71B BASIC program execute 10 times slower than Forth routines. And you can easily call routines in Forth from BASIC and vice-versa. Wouldn't it be nice to have some Forth 64-bit integer words floating around in high memory that you could call from BASIC for super-precision financial calculations, etc.? The whole idea excites me. And if my talents aren't sufficient to bring it to fruition, perhaps the TS folks can help. I'd be willing to fund a limited run of the chips, just to sell them at cost to interested parties. Sb: #166820-#a "math" rom? Fm: Stan Wong 70346,1267 To: SCOTT T. SCHAD 73720,1166 A math or a Forth rom sounds interesting. I've spent quite a bit of time fooling with building my own rom so I can add a few points (I got it to start working but dropped the project before I finished. I'll get around to finishing and publishing my findings one of these days). First, if the code is operating out of the option rom, the Basic rom is not available so any CALLs for Basic service must be provided for in the option rom. Second, if you intend to call these functions from Basic there must be a fixed place that must be set up in memory to intercept the call and switch roms back and forth. It's all doable but the math rom part might get a bit trickey programming-wise. I'm not familiar with Forth to venture any opinions. If you decide to go ahead with the project, I'll try to finish up my rom project and publish the results. #: 166916 S8/Tech/Programming 03-Apr-88 17:00:59 Sb: #166912-a "math" rom? Fm: Paul Globman 72227,1661 To: Stan Wong 70346,1267 You can have the option ROM call STD ROM routines and return to the OPTION ROM. Tricky code, but once understood, is easily implemented. I am uncertain as to how soon that material will be published, but if knowing that it can be done will inspire you to forge ahead, then "go for it"! Fm: SCOTT T. SCHAD 73720,1166 To: Stan Wong 70346,1267 Your comments have clarified the picture more than you know. And while I would look forward to seeing your techique for making eproms I wonder if you might well have a little 'cottage industry' of custom chip-making should things take off. Lots of us have wanted this or that program on a rom for years, but the $500 initial setup that TS or PCSG charges is unreasonable. How much hardware would someone need to burn his own chips? Can the model 100 be the controller? Fm: Stan Wong 70346,1267 To: SCOTT T. SCHAD 73720,1166 Cottage industry? Yes I guess I could but soldering those EPROMS to the circuit carrier is messy for klutz like me. Actually, I stopped at the point where I could load a .CO file into the EPROM and read it back into RAM, sort of like a romdisk. My next step was to allow multiple .COs to be loaded and read back. The step after that was to allow Basic programs to be stored (a few minor technical details to be worked out first). What would be hard is to get a Basic program to run from the rom. From Paul Globman's message (#166916) it sounds doable but I haven't figured out all the details yet. If you are interested in something like this, let me know and we can pursue it further. Fm: JAY HOLOVACS 74756,413 To: SCOTT T. SCHAD 73720,1166 Well...good luck. I'm sure a number of people on the sig would be interested in a more or less reasonable unit if it ever can be put together Jay -eof-