PROTOCOLS...

Hi, All! Today I made the hardware part of LINK. Works in the terminal - I serve it randomly bits - prints ;) I don’t know what to do next. Can someone tell me the protocol? transferring data so that they turn into bytes? ;) Can do it like in ABC Morse - long stick, short stick ;) But this, IMHO, is perverted... Bye All...

Hello Konstantin! 27 Jul 02 13:46, Konstantin Butorin -> All: KB> Today I made the hardware part of LINK. Works in the terminal - I submit KB> random bits - prints ;) I don’t know what to do next. Maybe someone KB> will tell you the data transfer protocol so that they turn into bytes? ;) KB> Can do it like in Morse code - long stick, short stick KB> ;) But this, IMHO, is perverted... Transfer bytes as you wish! The task of the driver for the link (from the MMD/EMC side) is simple and simply transmit the bytes issued to it over the line and receive the bytes from the line and issue upstairs. How it will work is of no concern to anyone :) It all depends on what kind of link you have (how many wires) If there are two signal wires in both directions, then one is the data, the second is the data strobe. And receiving strobe confirmation. If one signal wire goes in both directions, then “0” is transmitted one type of impulse, “1” - another. Thankfully speaking your version with Morse code ;) (Why didn't you like it? ;) Alexand

Press RESET immediately, Konstantin! 27 Jul 02 13:46, Konstantin Butorin wrote to All: KB> Today I made the hardware part of LINK. Works in the terminal - I submit KB> random bits - prints ;) I don’t know what to do next. Maybe someone KB> will tell you the data transfer protocol so that they turn into bytes? ;) KB> Can do it like in Morse code - long stick, short stick KB> ;) But this, IMHO, is perverted... First, explain what a transmission channel is. Bit to one one way and a bit the other way? Score it, or do 2 beats at least and slow down your athlone to the level of a ztnik or connect via serial port.

Press RESET immediately, Aleksandr! 29 Jul 02 19:02, Aleksandr Majorov wrote to Konstantin Butorin: AM> It all depends on what kind of link you have (how many wires) AM> If there are two signal wires in both directions, AM> then one is the data, the second is the data strobe. This is an inefficient way to encode data. With this coding, in one change Only 1 bit of signal can be transmitted. But 2 wires can have 4 different condition. If the transmission of each new piece of information is accompanied by a change state interface, you can encode 3 different states at a time. Can be transferred information 1.5 times faster. For example, for the time being 8 signal changes to your way is transmitted only 8 bits could be transferred to 12 bits. You can of course say fuck it such nonsense if it is enough to increase the speed of signal change, but then the frequency of the signals is the wires will also increase, and even now she has probably gone beyond all limits for a typical Chinese LPT cable.

Hi, Alexander! KB>> stick, short stick ;) But this, IMHO, is perverted... AM> Transfer bytes as you want! I want the PC to not have to write anything. There are terminals - let them work hard... And on spec write a driver or program for serial port. AM> It all depends on what kind of link you have (how many wires) Whatever is more convenient for writing a program, that’s what I’ll do. 2 wired for now AM> If there are two signal wires in both directions, AM> then one is data, the second is data strobe. Is it the same on a PC with a null modem? Correct me if I'm wrong, but I see it all like this: 1. Check to see if the strobe signal is one? 2. If it is one, then we accept the bit. Otherwise - see point one 3. We accept 8 bits, combine them into a byte, check if necessary and go further. A couple of questions then. If you make the strobe signal continuous (which will increase speed), then how do computers synchronize between bits? If the issue is resolved synchronization, then the second question arises - why then do we need a strobe wire? Well, if the strobe signal is interrupted, then the third question: on PC in Is this done on the serial port? Or are the data blocks synchronized there?AM> If one signal wire goes in both directions, then “0” is transmitted AM> with one type of pulse, "1" with another. Do you mean "impulse type"? How to ask it and how to receive it at the other end? AM> Good to say your version with Morse code ;) AM> (Why didn't you like it? ;) So there's an extra brake... Bye Alexander...

Hello Konstantin! 03 Aug 02 15:23, Konstantin Butorin -> Aleksandr Majorov: AM>> Transfer bytes as you want! KB> I want nothing to be written on the PC side. There are after all KB> terminals - let them work hard... And on the spec write a driver or KB> program for the serial port. You will have to install something like 580BB51 on Speck. Or emulate this matter. AM>> It all depends on what kind of link you have (how many wires) KB> Whatever is more convenient for writing a program, that’s what I’ll do. Bye 2x KB> wired In the sense of data there and data back or 2 wires there and back? AM>> If there are two signal wires in both directions, AM>> then one is the data, the second is the data strobe. KB> Is it the same on a PC with a null modem? And skip knows him, he has never worked on a PC null modem :) Now I looked at the diagram - there are nine wires used there. Although I heard somewhere that less is possible - about 4 or 5... KB> Correct me if I’m wrong, but I see it all like this: KB> 1. Check if the strobe signal is one? KB> 2. If it is one, then we accept the bit. Otherwise - see point one Having accepted the bit, we wait until the strobe is reset!!! KB> 3. We accept 8 bits, combine them into a byte, check if necessary and KB> moving on. I did it a little differently in my link. Yes, these are additional brakes, but I wrote “to make it faster and more demanding, but the speed will do” :) There is a Data and Strob output on each side. There are also the same entrances. The transmitter sets the data bit to Data and sets Strob to “0”. And waits until the receiver sets the strobe to “0”, after which turns the strobe to "1" and so on. Receiver: waits until the sender sets the strobe to “0”, reads the data bit, sets its strobe to “0” and waits until the transmitter sets his strobe to “1”, after which he sets the strobe to "1". Ugly, redundant... But it works! :-) KB> Then a couple of questions. If you make the strobe signal continuous (which KB> will increase the speed), then how do computers synchronize between bits? If KB> solve the issue of synchronization, then the second question arises - what for KB> then do we need a strobe wire? Hy and if there is a strobe signal KB> interrupt, then the third question: on PC in the serial port like this KB> done? Or are data blocks synchronized there? Mlyn, well, you asked. :) I don't know now how it all works. And it was never necessary... AM>> If there is one signal wire in both directions - then “0” AM>> is transmitted by one type of impulse, "1" - by another. KB> Do you mean "impulse type"? How to set it and how to get to the other end of it KB> accept? How to do it when saving to tape ;-) Essentially you output “1” to the port, pause, output “0”. By measuring the “pause” you determine whether it was zero or one. AM>> Good to say your version with Morse code ;) AM>> (Why didn’t you like it? ;) KB> So the brake is additional... Are you in a hurry to get somewhere? :) My version of the link produces 2000...2500 bytes per second. This takes into account the fact that MMD itself polls the link, draws on the screen, writes to disk... (2500 - this is in turbo mode with recording to RAM disk) Isn't it enough for you? :) Alexand

Press RESET immediately, Konstantin! 03 Aug 02 15:23, Konstantin Butorin wrote to Aleksandr Majorov: KB> A couple of questions then. If you make the strobe signal continuous (which KB> will increase the speed), then how do computers synchronize between bits? By time. KB> If you solve the synchronization issue, then the second question arises - a KB> why then do we need a strobe wire? I already wrote - it is not needed. Without it, it's 1.5 times faster. KB> Well, if the strobe signal is interrupted, then the third question: on PC in KB> is this done on the serial port? Or the blocks are synchronized there KB> data? Asynchronous interface. Each bit has its own time interval in a parcel consisting of /about/ one byte (see my other letter). Reception starts when the start bit is detected and is received sequentially all bits start from zero. The start bit is coded log.0 (+12v), stop bit log.1 (-12v) is encoded. Communication typically uses 8 bits of data, 0 parity bits and 1 stop bit. There can be more than one stop bit. Actually a stop bit is simply a pause in transmission; it occurs whenever the transmitter nothing to transmit and at least one bit interval between the lasttransmitted bit before and start bit of the next sending. That is, for transmission of a byte, you need to send at least 10 bits (start, data, stop). KB> Do you mean “impulse type”? How to set it and how to get to the other end of it KB> accept? Information can be encoded by voltage, current, pulse width...

Press RESET immediately, Aleksandr! 22 Aug 02 21:03, Aleksandr Majorov wrote to Konstantin Butorin: AM> And skip knows him, he has never dealt with the PC null modem :) AM> Now I looked at the diagram - there are nine wires used there. AM> Although I heard somewhere that less is possible - about 4 or 5... Minimum 3: general, transmission, reception.

Hello Aleksandr! August 22, 2002 (it was 21:03 then) Aleksandr Majorov in his letter to Konstantin Butorin wrote: AM> My version of the link produces 2000...2500 bytes per second. AM> This is taking into account the fact that MMD itself polls the link, draws on the screen, AM> writes to disk... (2500 - this is in turbo mode with recording to RAM disk) Hmm... And for me it’s 3000-3500 on average. Why? Although I don't mind :))) AM> Isn't it enough for you? :) No :) Vega/ex-Style Group. FIDO: 2:5030/885.34 ZXNET: 500:812/5.13 E-mail: vega56@mail.ru

Hello Vlad! 27 Aug 02 00:31, Vlad Sotnikov -> Aleksandr Majorov: [skipped] AM>> My version of the link produces 2000...2500 bytes per second. AM>> This takes into account the fact that MMD itself polls the link and draws on AM>> screen, writes to disk... (2500 - this is in turbo mode with recording to AM>> RAM disk) VS> Hmm... And I have 3000-3500 on average. Why? Although I don't mind :))) Mystic... :) Although... Are you receiving/transmitting to RAM or to a physical disk? And what is your PC doing at this time? My link usually works in the background - maybe and it slows down. AM>> Isn't it enough for you? :) VS> Not :) Enough for me too :) Alexand

Hello Aleksandr! August 27, 2002 (it was 1:41 pm then) Aleksandr Majorov in his letter to Vlad Sotnikov wrote: AM>>> My version of the link produces 2000...2500 bytes per second. AM>>> This takes into account the fact that MMD itself polls the link and draws on AM>>> screen, writes to disk... (2500 - this is in turbo mode with recording to AM>>> RAM disk) VS>> Hmm... And I have 3000-3500 on average. Why? Although I don't mind VS>> :))) AM> Mysticism... :) AM> Although... Are you receiving/transmitting to RAM or to a physical disk? On the hard drive. Quite fast. AM> And what is your PC doing at this time? As a rule - nothing. AM> My link usually works in the background - maybe AM> and this slows down. I have it in the background with the extras enabled. The link to the task is hanging. Therefore Sometimes I even launch from pure DOS. AM>>> Isn’t it enough for you? :) VS>> Not :) AM> I have enough too :) You bet :) Vega/ex-Style Group. FIDO: 2:5030/885.34 ZXNET: 500:812/5.13 E-mail: vega56@mail.ru

Hello Vlad! Tell us about what you do. This is me about protocols and the connection between the PC and Speccom. What successes have you already achieved? It’s better to write to me by e-mail: axor@mail.ru And one more question. Do you have the latest versions of MMD? And then we collected XTR here, and We still use the old one. If there is, throw it on the soap as well.