Good question. I know that 19,200 makes a substantial qualitative difference compared to 9600 when using TELCOM to connect to a UNIX shell. For loading BASIC programs, I don't know that it would make a difference as I usually just let it run and forget about it while it tokenizes.I only mentioned loading from BASIC because that was the lowest maximum download speed in your table (600 baud).
Timing test for a 16K BASIC (in ASCII) sent to a Tandy 200: Command Bits per second Time Effective bps LOAD "COM:98N1ENN" 19200 58s 2793 LOAD "COM:88N1ENN" 9600 59s 2745 LOAD "COM:68N1ENN" 2400 78s 2076 LOAD "COM:48N1ENN" 600 270s 600 So, as Mike predicted, there is no difference between 9600 and 19,200. What was interesting to me was that, although those rates are able to go faster than 2400 bps on average, when the connection speed is 2400, the effective rate is lower. That shows that the tokenization speed on my Tandy 200 varies. It is rarely faster than 9600 but it is sometimes slower than 2400, which is probably why Mike suggested 600 as a safe maximum speed. And, sure enough, when I try 600 bps, the BASIC tokenizer is able to keep up perfectly without ever having to send XOFF to pause the transmission. This also shows that any PC serial port that is able to send a BASIC file at 9600 has to have XOFF/XON working well. While it is not conclusive, I would guess that the same connection would work fine at 19,200. —b9 On Tue, Nov 1, 2022 at 8:57 AM MikeS <[email protected]> wrote: > Maybe I should have said that XON/XOFF is hit-and-miss depending on the > hardware instead of 'unreliable' ;-) . > > It's not always easy to know whether a particular setup will work or not; > I've seen USB cables recommended that I couldn't get to work and, > conversely, had no problems with others that supposedly did not. > > FWIW, I normally also run at 19200bd but thought it would be safer to > recommend 9600 bd in this discussion since it usually works no matter what > and makes very little difference in actual throughput. > > Have you ever compared actual BASIC download speed at 9600 vs. 19200bd? > > m > > > ----- Original Message ----- > *From:* B 9 <[email protected]> > *To:* [email protected] > *Sent:* Tuesday, November 01, 2022 3:33 AM > *Subject:* Re: [M100] Notoriously S.L.O.W BASIC posted - help speeding it > up appreciated > > > On Mon, Oct 31, 2022 at 12:11 AM MikeS <[email protected]> wrote: > >> Reliable maximum download speeds on the M100 without handshaking are >> around: >> BASIC: 600 bd to allow time to tokenize and store. >> TERM: 2400 because of the slow LCD scrolling. >> TEXT: 9600 since it does not display while loading. >> > > That sounds right when the other end does not have hardware-level > XON/XOFF, but it should be much faster with a better UART on the PC, like > an OX16C950 > <https://pdf1.alldatasheet.com/datasheet-pdf/download/161698/OXFORD/OX16C950.html> > or a chip from FTDI or MOXA. As soon as the M100 sends XOFF, the UART chip > in the serial port automatically stops the flow of data. No lag, no lost > characters. I connect using 19,200 bps in BASIC (*LOAD "COM:98N1ENN"* ) > and it works perfectly when I use certain USB serial adapters (and not > others). > > Has anyone with a Model 100 tried using a serial card or USB adapter that > supports "automated in-band flow-control"? Do you see the same high-speed > connection as I do on my Tandy 200? > > —b9 > >
