DEVEL.THD --- Copyright 1987 by Phil Wheeler An original compilation of Compuserve Model 100 Forum messages for use by Forum members only. When doing assembly of Model 100/102/200/Kyo machine language programs on my Model 100 with TDD2, I've often considered setting up a development environment on a bigger, faster desktop. These messages discuss approaches to doing just that. Message range: 161793 to 161912 Dates: 12/1/87 to 12/3/87 Fm: Mo Budlong 76167,3310 To: all Re 100/102 Telcom Differences I've done a lot of development on the 100, Option Roms, Assembly type stuff and have hit a bug/snag/anomaly etc in the 102. I develop large .CO programs using CPM's MAC assembler and use an RS-232 loader to load the Intel Hex file and translate it and poke it into memory. The procedure is download a short BASIC loader using RUN "COM:88n1e". The BASIC loader is too slow for large files, so the first file I down load is an Intel Hex file for a machine code loader. This loader is poked into memory by the BASIC program and then CALL'ed. The machine loader is very fast and I can transfer 20k .CO files in no time. Problem: This technique works 100% of the time for the Model 100. For the 102 which I recently started working on the result is erratic. It seems to depend on the telcom program running on the Host computer. The Wang PC PCTTY program always worked for the 100 and the 102. SmartCom running on an AT clone always works on the 100 but frequently fails (75% of the time) on the 102. The error always appears in the first 100 or so bytes of the final .HEX file that is downloaded. Why is there a difference between the 100/102? The Disassemblies I have done do not appear to show any significant difference, but I haven't been through all the COM code in detail. I'm scraping up another 102 to test to see if the one I have is unique. Can anyone help?? Has anyone heard of a difference in the 100/102 in the RS232 handler? Fm: Phil Wheeler 71266,125 To: Mo Budlong 76167,3310 There are some difference files, I think in DL8. Diff is mainly in the keyboard matrix -- and I think there are others. Not that I know of in Telcom, tho. I always use FLTIBM.COM (see FLTIBM.DES, DL3) for such chores. Works on PC-compatibles. Versions for COM1: & COM2:. Give it a try and let me know if it doesn't do what you want, reliuably. Now -- what computer are you running MAC on? Do you have a V-20 or one thos NEC chips that will run CP/M, or what?? Fm: Mo Budlong 76167,3310 To: Phil Wheeler 71266,125 Thanks for the tip Phil. I'm out of town for a few days and will check it out when I get back. I'm running UniDos on an Equity III+. It comes with a Z80 chip on a plug in board. Unidos is a TSR. Once installed I presume it filters the DOS command line if DOS fails to find a program you requested. Anything named with a .CPM extension is picked up and run under a CPM emulation using the Z80 board. It is very fast. I also have Wang PC. The Wang runs a non-IBM DOS. The PC comes with what they call a 928 Com board that allows the PC to be used as a standard Wang terminal on the Wang VS Mini Computer. The 928 board is a Z80 that takes control of the terminal when you are logged on to the big VS. Wang provides software that also allows you to use the board for a CPM emulator. Since most compiled languages (aside from Borland's stuff) is MS-DOS rather than IBM PC based, Iworked on that Wang PC happily for years. I have a collection of development tools for the Model 100 all developed under CPM (emulations) including a C compiler. I have since ported the C compiler over to MS-DOS so that it runs in native DOS mode rather than under a CPM Emulation. The output is 8080 Assembler. I haven't finished writing the cross assembler yet, so I still use MAC. Anyway I'll check the notes on telcom differences and thanks for your help. Fm: Phil Wheeler 71266,125 To: Denny 76701,40 Hmmmm..just discovered that I have a cross-assembler in my Z-150 and did not knwo it. I'm running with a V-20 which emulates an 8080. And (which I had forgotten) in my CPM directory I have MAC.COM, the standard CP/M macro assembler. It will assemble M100 source files (tho they must be in the ROM2/Cleaseau style, it seems) and make an Intel Hex file, to be laoded with HEX2CO.BA (or such) in DL8. The last part may be the rub: Slow, I suspect. But this is a possible alternative to the use of CS asm with the TDD -- and would let me deal with big sources in the PC.