online > help, online coaching, etc. Is this the right place to send out a > message about such tools? Is OS9 an OS2 group? Am I in the wrong > place? Where should I go to send such a message? Any help would > be appreciated. > Most of us don't run OS2 (not even at gunpoint). ;) There are a few of the regulars here who have some knowledge of OS2, and maybe one of them will respond, but OS9 is not even distantly related to OS2. ******************************** "Don't take life too serious - it ain't nohow permanent" (Pogo) John R. Wainwright <> <> -*- 86125 8-MAR 23:32 General Information RE: Am I in the right place? (Re: Msg 86107) From: JHICKLE To: WEINSCHENK (NR) go gr cus 41 for os/2 forum; or go com pc for generic pc forum which has an windows/os2/and_somethin'else database. I use os/2 at work, but only because they pay me. -*- End of Thread. -*- 86113 8-MAR 19:59 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86111) From: RICKULAND To: CBJ You know, lots of folks claim their CoCo's 'do' 9600 bps. The honest ones append a note with an actual cps thruput around 500 up to 700cps..... it's important to note what we are test here is how low the other system can go before dropping connection. Until a CoCo can thruput 960cps it can't really do 9600bps, although it might be able to maintain the connect. I know, but I _am_ working on it! -ricku -*- 86114 8-MAR 20:03 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86113) From: MITHELEN To: RICKULAND (NR) Right, I think this has been Carl (and my) point along... Although, I saw a message the other day on the listserver that a guy could get 900 CPS sending from the CoCo to a PC with a 19200 baud link using the new rz/sz 3.24 for the CoCo (not sure what pathes or hardware mods ha may have also had) with a 9600 baud link, he was only able to get 500 CPS. -- Paul -*- 86120 8-MAR 22:22 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86092) From: NEHOWARD To: CBJ Carl, What problems are/were you having with 9600 baud and the BBS package? I run a BBS package that uses a 'un-modified' serial port(OS9 Based), using Sacia, that handles connection rates of up to 14,400 baud and does auto-ansi detect, Auto-RIP detect and auto-baud detect. The driver for this is written in Basic09 which is supposed to be VERY slow compared to machine language routines. I also use a standard, 512K Coco3 with the standard 6809 chip and a Disto Supercontroller II with the 2-in-1 board(HDII/RS232)... The system runs just fine for me... I am curious to know what problems you had/have with the 'high-speed' connections... Nelson -*- 86121 8-MAR 22:37 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86113) From: NEHOWARD To: RICKULAND (NR) True, the CPS is a tad slow on the coco. Even with mine 'nul-modemed' to another machine the CPS is slow at 9600 or even 19200.. The actual CPS is comparable to 4800bps. Connect rates are no problem though for a BBS at the higher baud rates... Mine handles it fine along with a series of auto-detect sequences for various things(goodies) that I have written into the front-end of the package. What is needed is a new serial driver that will allow a faster cps on file xfers. Even then, will the 1.7mhz processor keep up with it?? Nelson -*- 86123 8-MAR 23:13 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86121) From: MITHELEN To: NEHOWARD The single bigest problem with highspeed operation and an un-modified RS-232 pack is the 6551's "defective" handleing of CTS. This problem becomes VERY evident when using a streaming transferprotocal, such as Zmodem or StG NetXfr. And even more so if the receiveing system is at a LOWER baud rate then what the CoCo is set at. This single problem is what is really causing Carl, and other StG sysops problems with running as 9600 baud... Unforch, Carl tried to "repair" his RS-232 pack with a new version of the 6551 that handles CTS properly, but, for some reason it did not work with the new chip... The other big hangup with the 6551 ACIA is it's lack of an internal buffer... The CoCoIO from Connect, when it eventually makes it out to the market place with solve both these problems for the power BBS operator, and serious telecomunications nuts. -*- 86126 8-MAR 23:45 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86113) From: RANDYKWILSON To: RICKULAND (NR) Well, Rick, yes and no. The serial side of OS-9 can indeed hold a 9600 connect with full throughput. I have a version of SuperComm here that clears 940cps doing y-modem. This version has one thing ommited to get a true feeling for serial speed; it doesn't do anything with the data once validated. What really kills throughput on a coco is moving the data from the port to some place else. The screen is dreadfully slow (and I'm working on that part). Floppies are almost as painful. Hard drives start to bring in reasonable results (700+ with SC doing y-modem). Right now, a ram drive is by far the fastest way, and is recommended for doing short files. I clear 920cps on my 6309 system using Rammer. To get more speed and reliability out of high speed serial, several fronts need to be attacked. You're after one of the biggest ones; a buffered, smart, serial port. A 6809/6551 combination at 9600 is spending vast amounts of cpu time servicing the port for each and every byte. Your buffered board should drop this overhead by a factor of 8 (give or take). Another thing to be addressed that will affect preceived speed is screen writes. I'm studying and testing several ideas on this front. But what would really be nice, really get the 1655x throughput speeds up there, is a good mass storage scheme. How about taking some of your free time and looking into grafting a DMA and SCSI controller into a Disto 2-meg? :> Randy -*- 86131 9-MAR 10:43 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86120) From: CBJ To: NEHOWARD The problems I've had are as follows: At high speed bidirectional transfer of files via our networking program there are lost characters. This leads to retransmission of the affected packets of data and a lowering of CPS. I've actually seen speed drop as low as 240CPS at 14,400 when the transmission is done. Not very fast as my 2400 baud modem does better than that. I am using a CoCo3 with an unmodified COMM4 board and a disto HDA with a SC1. No floppy access is done while the tranfer of data is taking place. Carl -*- 86133 9-MAR 20:21 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86131) From: NEHOWARD To: CBJ (NR) Yeah, I understand the cps problem. I had originally thought you were exper- iencing problems with the actual 'connect' from the modem. I too, am looking forward to a RS232 port/driver that will handle all the data lines correctly as I run scads of netmail programs(3 packages total) with my BBS and could use the extra speed to cut the LD bills. I just opened the front end of my BBS software up to 14.4K as it was restricted to 9600-hardcoded before and would like to be able to utilize it!! A Comm4 board?!! I would 'kill' for one of those!!!! Know where I could get one for a 'reasonable' price??? That would give me 5 serial ports to play with here in-house!! I could throw a second node on the BBS and hook up 2 more workstations. What a gass that would be!!(I tend to 'stress' my poor ole machine)(have a workstation null-modemed with the bitbang drivers now at 19.2K with the BBS on-line also)Well, good luck to ya, wish I could help out more. I'm sure you have 'played' with buffer sizes in Sacia. I have mine set to 2K and things seem to work ok here. Could be something in the xtp setting if you do an xmode of the port. I dunno... Wish ya luck! Nelson -*- 86142 9-MAR 22:56 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86131) From: KSCALES To: CBJ (NR) > I've actually seen speed drop as > low as 240CPS at 14,400 when the transmission is done. Not very fast as > my 2400 baud modem does better than that. I am using a CoCo3 with an > unmodified COMM4 board and a disto HDA with a SC1. No floppy access is > done while the tranfer of data is taking place. Hi, Carl - If you have a Multipak installed in the system, my versions of the Disto hard drive drivers (scsi47.ar, cchdsasi.ar, sasi07.ar) may not support modem operation at steady streams above 4800 baud. (This is only WITH a Multipak. Otherwise, this should not be a factor, but you will naturally still be limited by overall system throughput.) ... / Ken -------------------------------------------------------------------------- Ken Scales Delphi:KSCALES Internet:kscales@delphi.com CIS:74646,2237 -*- 86147 10-MAR 07:04 Telecom (6809) RE: 9600 on a CoCo... (Re: Msg 86111) From: MARKGRIFFITH To: CBJ (NR) Carl, > I've never been able to get 9600 baud working reliably on a CoCo under > StG netxfr doing bidirectional transfer of data. Errors start piling up > after the first couple of packets. The problem most likely lies in the software, not the hardware. I know people that were running full streaming protocols with a CoCo3 Level II that were getting the maximum throughput. And that was with a standard RS-232 pack, but using a newer ACIAPAK driver. /************* /\/\ark ************/ (uploaded with InfoXpress Ver 1.01) -*- End of Thread. -*- 86115 8-MAR 20:20 General Information RE: DISTO Products running Low. (Re: Msg 86073) From: DISTO To: DBREEDING Well, the serial chip I use in the 4IN1 can do RTS/CTS. The newer 4IN1 boards does not use it. It is defaulted to on. This board uses a 10 pin connector. The older versions of the 4IN1 used a 16 pin RS-232 connector and uses the RTS/CTS lines. I have a couple of these left. Or, if you are a hacker, you could get a schematic set from me, and wire in the RTS/CTS yourself. It would require a chip or two. I don't recomend this unless you know a bit of electronics. -Tony. -*- 86138 9-MAR 22:07 General Information RE: DISTO Products running Low. (Re: Msg 86115) From: DBREEDING To: DISTO (NR) Say the older 4in1 DOES support RTS/CTS? Might just get one (dunno how long I will keep my coco; I'm thinking about upgrading to OSK (or perhaps a PC ), but I might give it a shot. Will one of these boards then support local hardware control? As an aside, what was the purpose of dropping the RTS support, economics or to make data terminal usage easier? Thnx.... -- David Breeding -- *** Sent via CoCo-InfoXpress V1.01 *** ^^^^ ^^^^^^^^^^ -*- 86139 9-MAR 22:08 General Information RE: DISTO Products running Low. (Re: Msg 86115) From: DBREEDING To: DISTO (NR) One more question... Do the cables come with the 4in1? Guess all I'd need would be the serial cable, if it requires a 16-pin connector, mine is 10 on the card side. -- David Breeding -- *** Sent via CoCo-InfoXpress V1.01 *** ^^^^ ^^^^^^^^^^ -*- End of Thread. -*- 86116 8-MAR 20:25 OSK Applications RE: system command (Re: Msg 86099) From: VAXELF To: PAGAN I agree that os9exec() is normally the way to go, except in this case, I tried everyway possible to get it to work using os9exec(). Finially gave up and stuck with system() in this case. I tried using if ( status = system(process) != 0 ) { error processing exit to parent } normal exit. but it didn't seem to work. Mabey I overlooked something. I'll give it another try. John D. -*- 86129 9-MAR 01:47 OSK Applications RE: system command (Re: Msg 86116) From: PAGAN To: VAXELF >I agree that os9exec() is normally the way to go, except in this case, I >tried everyway possible to get it to work using os9exec(). Finially gave up >and stuck with system() in this case. What are you trying to run? Stephen (PAGAN) -*- End of Thread. -*- 86117 8-MAR 20:28 Programmers Den RE: C's stack (Re: Msg 86096) From: WDTV5 To: BANANAMAN Yeah, seems like it might be a good idea, after all its supposed to exit if thjheres not enough room on the stack, not crash the whole machine in grand and glorius color! Which is what I'm getting if the stack request is several kilobytes. Cheers, Gene -*- 86118 8-MAR 21:17 Programmers Den RE: C's stack (Re: Msg 86108) From: BANANAMAN To: JEJONES Your point is well taken. Using malloc() to create local variables instead of declaring them in a recursive procedure would help the compiler out in those situations. But I've had these stack allocation problems in programs which don't even use recursion. (which is most of them. To me, recursion = debugging nightmare) I wonder what causes the compiler the confusion in those situations. --Andy -*- 86124 8-MAR 23:26 Programmers Den RE: C's stack (Re: Msg 86096) From: JHICKLE To: BANANAMAN Needn't bother disassembling stkcheck()... Source for that, cstart.r, and everything else (except whats in clib.l) comes with the C compiler. Look on the second disk in a directory called SOURCES. -*- 86140 9-MAR 22:25 Programmers Den RE: C's stack (Re: Msg 86124) From: BANANAMAN To: JHICKLE (NR) Thanks! --Andy -*- End of Thread. -*- 86119 8-MAR 21:35 General Information RE: MM/1A Mouse Mysteries Revealed?? (Re: Msg 86091) From: COLORSYSTEMS To: BOISY I have an Identity Systems trackball which also has a switch which has "PC" as one selection and "MS" as the other selection. I have to have it on the PC selection for it to work. Here is an ident of the driver I am using: Header for: msdrv Module size: $272 #626 Owner: 0.0 Module CRC: $9635B1 Good CRC Header parity: $1C44 Good parity Edition: $8 #8 Ty/La At/Rev $C01 $A000 Permission: $555 -----e-r-e-r-e-r System Mod, 68000 obj, Sharable, System State Process ------------------------------------ Zack C Sessions "Always in motion is the future." - Yoda -*- 86136 9-MAR 21:24 General Information RE: MM/1A Mouse Mysteries Revealed?? (Re: Msg 86119) From: JOHNREED To: COLORSYSTEMS Zack (& Boisy), I am using the same driver Zack just described (msg# 86119) with the cheapest mouse I could find. It says "Generic Mouse" on the bottom, and came in a white box that said "Mouse". Got it at a discount office supply store. I tried an Artec mouse first, then a Dexxa -- the cheap one worked best with orig MM/1 and still works with the MM/1A. As others have said, the thing seems to like cheap mice. ******************************** "Don't take life too serious - it ain't nohow permanent" (Pogo) John R. Wainwright <> <> -*- End of Thread. -*- 86122 8-MAR 22:39 General Information RE: MM/1A Mouse Mysteries Revealed?? From: BOISY To: MARKGRIFFITH I've disassembled the msdrv code. It appears to have two entry points: one which initializes the port and sets up the IRQ polling routine, the othe rwhich terminates the port and removes the entry from the IRQ polling table. It's absurd *not* to have a Logitech mouse driver, as this is the more popular variant. I'm willing to do the work to get Logitech mice to work on the MM/1 if I can get the commented source to do so. If I'm declined, the someone who is privy to the source should commit to the conversion. -*- 86127 8-MAR 23:57 General Information RE: MM/1A Mouse Mysteries Revealed?? (Re: Msg 86122) From: RANDYKWILSON To: BOISY Boisy, what are you refering to as "Logitech"? My understanding on things, at least back when the MM/1 was being developed, is that there are/were two mouse standards. One is MicroSoft. The other was Mouse Systems. At that time, Logictech used the Mouse systems standard. And so, as always happens in the world of buzz words, any mouse that used the Mouse Systems standard was refered to as "Logitech compatable". So, anyway, has Logitech since switched or developed a new standard? if so, got any idea what it looks like? I do have some mouse driver sources from an antique version of windio that could be used as a guide for studying a disasm of the current drivers. BTW, you pretty much got the picture of how the drivers work (mouse and keyboard). They are nothing more than replacable IRQ handlers for windio with three entry points (init, term, irqserv). Information is passed back andsd forth via a few fields in WData. Randy -*- 86146 10-MAR 07:04 General Information RE: MM/1A Mouse Mysteries Revealed?? (Re: Msg 86122) From: MARKGRIFFITH To: BOISY (NR) Boisy, > It's absurd *not* to have a Logitech mouse driver, as this is the more > popular variant. I'm willing to do the work to get Logitech mice to > work on the MM/1 if I can get the commented source to do so. If I'm > declined, the someone who is privy to the source should commit to the > conversion. Well, you might try leaving Kevin Darling a message, but don't expect an answer. You also might try a message to Kevin Pease. I don't have his Internet address with me, but his work number is 615-673-9100 extension 2792. I'd be happy to send anything to you if I had it. /************* /\/\ark ************/ (uploaded with InfoXpress Ver 1.01) -*- End of Thread. -*- 86128 9-MAR 00:35 Programmers Den CGFX From: BILLGRESENS To: ALL Can someone please tell me how to compile a program that calls the functions in the Tandy CGFX and/or the Mike Sweet CGFX? It always fails with a "CGFX.L is not a relocatable" error when I use the c.link and I get an unresolved external error with rlink. My meager brain seems to recall some problems w/the linker/assembler combo but I can't remember the solution. Thanx, Bill -*- 86141 9-MAR 22:28 Programmers Den RE: CGFX (Re: Msg 86128) From: BANANAMAN To: BILLGRESENS (NR) Are you using "-l cgfx.l" in your compile statement? If not, did you download it with xmodem? xmodem adds nulls which may freak the compiler out. --Andy -*- End of Thread. -*- 86130 9-MAR 06:32 General Information RE: MM/1A Screen Problems (Re: Msg 86090) From: SCWEGERT To: BOISY > The buffers were $14 and $50 respectively BEFORE I changed them. They > were increased in value, not decreased. So I've now read. Having already increased my /t3 and /t4 buffers to 2K to address similar problems, I made a bad assumption when you mentioned changing the buffers to 256. Thanks for the clarification. *- Steve -* -*- 86132 9-MAR 19:37 Games & Graphics RE: new user hard drive (Re: Msg 86105) From: TAULBORG To: RICKULAND (NR) I installed hyper io before putting os9 back on my harddrive.I had os9 on there before but something went on with it so I decided to put hyper io back on it instead of just os9!After putting hyper io on I decided to see if os9 could do anything with the hard drive and since hpio formats the hard drive with an os9 format os9 read the hard drive right away.Now my problem is getting a new os9 boot disk put together so that I can use the os9 stuff on my hard drive because the boot that I did have DIED for some reason.By the way I was never able to get any part of the boot to work off the hard drive so the whole boot process was on floppy and for some reason it don't work!!!! I am sure that once I can get a new boot disk put together with the drivers and stuff for the hard drive it will work fine!!I just can't do it without step by step instructions and ALL the books I have assume that I know a lot about os9 and I DON'T. -*- 86134 9-MAR 20:31 Programmers Den graphdev From: KKOBLIS To: ALL anyone know about dave morph? -*- 86135 9-MAR 20:47 Programmers Den RE: graphdev (Re: Msg 86134) From: THETAURUS To: KKOBLIS (NR) Sorry, never heard of him. Who is he? >Chris< -*- End of Thread. -*- 86137 9-MAR 22:07 General Information RE: Hello All... (Re: Msg 86084) From: DBREEDING To: REVWCP Thnx for the welcome. I could never afford to dload all the msgs (I have to pay Long/Distance in addition to Delphi, and at the rate this was going, it would have taken at least half a day. I use "IX", great stuff. As an aside, what is your telecom setup. Do you use data compression or anything? Type modem, Baud rate, etc. I'm trying to see if I can improve the speed of my CoCo. I have a Disto 4in1 and Hayes Optima 24. I'm wondering if the coco can use it. Of course the Disto board stock does not support RTS/CTS local control. TTYL.... -- David Breeding -- *** Sent via CoCo-InfoXpress V1.01 *** ^^^^ ^^^^^^^^^^ -*- 86143 9-MAR 23:10 General Information RE: Hello All... (Re: Msg 86137) From: REVWCP To: DBREEDING (NR) I am using a RS-232 pak, even though I have a 4-in-1. I am currently using a Patriot 2400 baud modem, nothing fancy. I hope to move up to a high speed modem at some point. With all best wishes, Brother Jeremy, CSJW OS9 User's Group Treasurer -*- End of Thread. -*- 86144 9-MAR 23:52 General Information os9 under From: TMF To: ALL Is the "international" os9 underground still in business? -*- 86145 10-MAR 05:03 General Information RE: os9 under (Re: Msg 86144) From: BROWN80 To: TMF (NR) I don't know how international it is but he is suppozed to still be in business There was a message here not too long ago that he had some problems with the earthquake and that he was going to have to resort to another double issue and then he should be back on track. I hope so. John Brown -*- End of Thread. -*- FORUM>Reply, Add, Read, "?" or Exit>