Any amateur radio operators in the Bay Area doing any digital modes on VHF/UHF?
There’s a pretty neat Winlink project written in Go [1](https://github.com/la5nta/pat ) that I’ve wanted to use. If anyone is out there interested in collaborating, please let me know.
I have one of these sitting on the bench and have dreamt for a while of setting up some kind of BBS on 70cm. Life has perpetually gotten in the way though. Also… I’m a long long way from the Bay Area :). 73 VE5AJA
Reminds me a bit of a HP mainframe I was an operator on in the mid 80s. We had essentially a paper/typewriter style terminal so that it would record all input and output direct to continuous paper that would be saved for auditing if needed.
Type a command, read your book until you heard it print a response, go read the response and do the next step. Repeat.
I had similar on a VAX 11/750. The LA120 DECwriter was connected to the system via a serial port mux adapter.
One night the mux adapter glitched and reset itself back to defaults, which turned on local echo, so it reflected 'login:' back to the VAX, which responded with the password prompt.
This prompt was reflected too, giving an incorrect password response...which was reflected back to a new login prompt...etc..rinse, repeat all night.
Everything was printed on the LA120 until the fanfold paper ran out.
I came in the next day and could not see the LA120 due to the amount of paper on top of it.
My Model 100 had a built in 300 baud modem. It was a modular plug vs an acoustic coupler, but still 300 baud.
The nice thing about the terminal program on the 100 was that you could upload text files. So, on a BBS you’d enter New Message, then upload the text file. And it loaded just like typing it in.
No Xmodem or anything, just “type it in”. Not a real problem at 300 baud.
But the nice feature of the 100 was that even though it was a 40 column display, you could set the right margin to something like 78 or 80, and it would word wrap your lines. So it would format fine for the BBS. Hardly rocket science but just super useful.
Also, 300 baud worked really well with a 40 x 8 display. Filled it up “fast enough” vs a 80 x 24 display.
More than usable for BBSing in bed (difficult to do with a VT100).
As I recall from talking to older people, the model 100 with an acoustic coupler was an extremely popular solution for journalists to quickly file stories from remote locations/while traveling. All you needed was the model 100, a payphone, and the coupler.
I worked at Radio Shack in the early 1990s. I was shocked to see that Radio Shack was still selling the Model 100s (acutually, model 102 by that time) as it was at least a 10 year old design by then. The manager told me that journalists still liked to buy them, because it was rugged, super portable, ran on batteries, had a great keyboard, and they could upload their text over a phone line. All I can say is their articles must have been very short as the memory limitations of the device were severe.
I might not remember this right, but I don't think that the memory limitation was a factor, as they were often used with rudimentary dialup terminal software as a dumb serial terminal dialing in to a newspaper run system. The story would sometimes be typed live from the reporter's notes directly into the shell session of the remote (minicomputer, etc) system on the far end. This mean that even if there was 5-8KB of text it wouldn't be a problem.
Then any formatting/line wrapping/typo and other mistakes from the raw submitted story would be fixed by copy editors from their local workstations.
Something neat I discovered about my first 300 baud terminal is that I could record the connection on a tape recorder and play it back and the whole process would work for a replay.
There was just wide enough tolerance and such a small part of the audio spectrum that it worked. Stopped working with 1200 baud modems. And 2400bps modems initially had a fight over compatibility, 9600bps went completely out the window, hayes vs everyone else.
My first modem was a 300 baud Hayes modem that was given to me. I used it on my Atari 1040 STE. It was ... slow. The next bump was to 1200 baud, and that definitely made things a bit more bearable. A new family computer came with an internal 2400 baud, and then eventually I got my own USRobotics 14.4k. It was akin to getting on the highway after being stuck on rural roads with a low speed limit. I still miss those connect tones (14.4k being my all-time favourite.) I still have a USRobotics 56k vEverything modem.
Interesting that the terminal appears to have at least an 80 character buffer, as it is able to print after an LF and while performing a CR.
My first modem was 300 baud and while it was painfully slow, text interfaces of the era tended to be very efficient. File transfers of high resolution images is where it tended to get very painful, especially since watching the block count increase (if even that) wasn't anywhere near as engaging as actually navigating through menus and reading information.
I was fortunate to be a sysop on one of Sydney's popular BBS. At first we offered some MUDs and chat. Then multilayer doom and ultimately IRC gateway before it became redundant. But through that I learned Linux, then novel which started my career. Forever
grateful for that modem.
'87 BBS kid here, started out at 300baud on my C64 and worked my way up as things progressed.
No leap in speed was more visually impressive and delightful than going from 300 to 1200 at 40 columns. Going from 1200 to 2400 at 80 columns felt weirdly slower.
It wasn't the state of the art in the early to mid 1980s, but that's when they were still popular. The famous Supra Modem 2400 arrived on the shelves in the late 1980s. I had a 14k4 modem in 1994, and a 28k8 one in 1995.
Add.: the 14k4 modem cost me an equivalent of about $350 in 1994.
Not only was it expensive, but it was a crazy time to buy a modem. There was also 16.8k and 19.2k, which were proprietary if I recall correctly. (I had a 19.2k modem, and was overjoyed when I found a BBS that I could connect to at that speed.) Around that time you also had 28.8k and 33.6k, which I think were standardized by were incredibly short lived.
I remember sticking to what was effectively 14.4k until I jumped to cable internet. It did not make much sense to upgrade since ISP rates were lower for slower modems and serial modems were also more expensive. When I did jump to cable internet, I took some flak because of my obsolete computers. I was far happier with an always-on connection on an obsolete *ix box than dial-up on a faster PC (since I didn't have the money to have both at the time).
Their popularity was because of aggressive marketing and sponsoring from USR, rather than being technically better than the competition. In the 1990s loads of Swedish BBSes had small USR advertisement blurbs on their login screens as thanks for the discounted hardware. The push even made it all the way into the elite boards (invite-only warez BBSes). Source: I ran two BBSes in the 90s and spent a few years co-sysoping a large elite board along with doing a lot of "modem trading" (slinging warez between boards).
Could be. I ran into a number of BBSes that dropped subtle hints that they hosted stuff that was not accessible to regular users. I always had the impression that it was adult content, but it could very well have been pirated software. (These were publicly listed BBSes.)
Most pirate BBSes publicly advertised themselves, as a form of bluster.. But having the phone number and getting an account were two very different things. :-)
You had a Courier or nothing. They were amazing if your ISP (or warez bbs) had them at the other end. Proprietary compression / connection algorithms, I guess?
Pretty much the only modem I got ACTUAL 56K speeds with was a Courier V.34 that I flashed to become a V.90
The standardized compression is called v.42Bis, but it was in fact a feature present only on few of USR's models while most of the other brands provided it with consistency. It worked well on plain-text content - I remember hitting 90 KB/sec download speeds when testing large text files on a 28k8 connection - and it made a noticable improvement when bouncing between menus and reading messages and such, but very limited use for actual file transfers since practically everything was already lha/lzh/lzx/dms-compressed.
My first connection was 1200 bps, and now I have 2 Gb/s. That's over 1 million times faster.
Downloading a program for my Atari ST at maybe 500 kB in size took an hour. Now it would be difficult to even measure the time needed. It's difficult to explain just how slow downloading things used to be.
An XT class luggable of mine had a Memory/RTC/GamePort/Parallel/Serial/InternalModem card installed when I acquired it. Some research on the card showed it was released in 1986. If they were being put on multifunction cards then the market had to already have been full of dedicated cards.
I was around and using modems at the time but really don't remember the brief era of 1200 baud modems. The family TRS-80 Model 1 had a Novation CAT 300 attached which my father's work let him take home when they disposed of it around 1980. In the mid-80's my C64 used a 300 baud Commodore 1660 modem. Next thing I remember, I was shoving an internal 2400 baud modem into a new Packard Bell PB500 XT class machine. That modem got well used.
1200 bps modems were available for the IBM PC much earlier than that. I think by 1982 your could get a 1200 or even 2400 bps modem for the PC. The AT models in 1984 standardized the faster speeds.
300 baud modems were still popular in the 1980s, because most people still used home microcomputers, and that's what was available for those low end machines. Atari even sold an acoustic coupler modem for their 8-bit computers that was very popular as it was quite inexpensive.
> There were others available slightly earlier, but they were commercial-grade and prohibitively expensive
They were not “commercial grade” exclusively. I had a Hayes Smartmodem 1200 in 1983 or 1984. It was absolutely consumer. In fact the case was identical to the Smartmodem 300 but with an extra LED or two. I hooked it up to an Apple 2e. It was indeed expensive, though.
Consider yourself lucky. I had a 2400 bps modem in 1995, my first year on the internet. The ISP recommended 9600 bps but everything seemed ok for 14 year old me. IRC was fine and most websites didn't have too many pictures.
Ditto here, I got a 2400 ISA internal modem for Christmas 1993.
I catapulted headlong into CompuServe, found it largely boring, found a copy of the 313 BBS list in some file area, and started calling local boards.
I was "that bastard tying up the line for too long with his slow-ass modem" once in a while, but mostly I was tossing QWK/BlueWave mail packets which only took a few minutes anyway, and I tried to do my long file transfers in the middle of the night.
As soon as I got on the internet and discovered alt.binaries, I scrambled to upgrade. 28.8's were out but still quite expensive, and they had pushed 14.4's into affordable territory so that's where I landed. I think that must've been around 1995.
I had a 14.4K modem, but the dial-in pool for my ISP only supported 2400 bps speeds. It took me most of the day to download the first C&C demo file which was 10 MB IIRC.
before 1987, really. When 1200 bps modems were released and the standard for that finalized, the 300 bps were a lot less costly and occupied the low end of the market.
At the point that 1200 and 2400 bps modems became available, most everyone who could afford it in the BBS communities upgraded. By 1992 a 2400 bps modem was a very low cost ISA card for a normal x86 PC and commonly included with most desktop PCs sold.
My first paid programming gig was $20 for an Atari BASIC USR function (a way to call machine code from BASIC) to compute the XModem checksum for a block. Doing it in BASIC was fast enough for a 300 baud modem, but introduced a noticeable delay for a 1200 baud modem.
That $20 came to right something around $1/byte. Imagine if we got $1/byte for a node project or Win32 app nowadays... :)
Wikipedia claims 1200bps was available in the 70s. https://en.m.wikipedia.org/wiki/Modem I did start with a 1200bps modem back in the 80s, even then it was considered quite slow but was enough for email reading and basic bbs tasks. It sucked for any kind of download though.
1200 bps didn't really show up as something people could buy until the early 1980s, however. In the mid 1970s if you had a 1200 bps modem it would be equivalent today to the cost of having a cisco or juniper router with several ports of 100GbE interface in your house. They were in use for university mainframe/minicomputer data links and such.
Don't know about state of the art, but in the early 80s, 300 bps was the economical choice, relatively speaking, for home users. 1200 bps were around but were a good deal less affordable and in quite a bit of the country, less reliable due to the general state of signal quality on the phone network at the time.
There were some short lived quirky technologies at the end of the dial up era, such as using multiple 56k modems bonded into a single connection, the increase throughput. Since this required two telephone lines on both ends, in addition to four modems, only the most data hungry went to that degree of effort. Most simply got broadband when it was offered.
As someone that worked at a dial up ISP and had an ISDN connection at home, modem bonding was really fascinating. At the ISP side, these were all T1 lines, so every line was technically 64k capable. The analog 56k modems rarely made a true 56k connection. However, there were some customers that lived one telephone pole away from the CO, and we would see them connected at 57600. One close customer would show at some oddball speed like 58k or 59k (though we don't know what they actually got at their end).
But for ISDN, each channel was the full 64k and that would bond two channels to 128k. I had a cisco modem that would let me use either channel as an analog phone. So it would connect up both channels, and when a call came in, drop down to one 64k channel. It was incredibly reliable. Even when I got DSL, I kept the ISDN for quite a while because it had more consistent upload speed.
I was on ISDN for about a year, using a 3Com modem. Compared to dial-up, it was amazing as the connect time was only 3-4 seconds. No more listening to the buzzy connect sequence trying to guess what rate it would negotiate.
But something happened to the line and BellSouth could not get it working again. After a couple of rounds of "It passes our tests" and "It still doesn't work" I gave up and went to Time-Warner's RoadRunner cable internet.
My ISDN ISP was Uzi Nissan, of nissan-dot-com fame.
I worked on the drivers and firmware for a few ISDN cards back in 1997-1999. One of my performance optimizations for the BRI cards cut ping times from ~30ms to ~16ms. The Motorola MC68302's puny 68000 CPU took a long time to memcpy() packets from the buffer the HDLC controller would DMA them into back to the shared memory with the host. Eliminating that copy was more of an improvement than I expected.
The PRI (T1) cards using Intel i960 CPUs, which were quite nice embedded CPUs. The HDLC controller was capable of using individual 64k channels or bonding them together for fractional of full T1.
It was a bit late in the lifetime of ISDN, but I learned a heck of a lot about embedded systems from that job.
ISDN was the standard for "next generation" phone lines. It was never really meant for data. The design was to handle converted analog signals their new digital network. The demand for true broadband upended that plan, tough, and not it's just voice over much faster IP technologies.
This was just something I remembered observing (ads in computer magazines for 300 baud modems for $299, 1200 baud modems for $1199). But I just ran across a reference in wikipedia:
>The modem market in the 1970s was very simple and stagnant. Modems tended to sell at US$1 per baud. Hayes saw no need to be different—the original Hayes 300 baud modem sold for US$299 retail. At that price point, Hayes could build a "Cadillac of modems", using high-quality components, an extruded aluminum case, and an acrylic front panel with a number of LED indicators. As the modem market expanded, competitors quickly copied the Hayes command set and often the Hayes industrial design as well. To compete with Hayes on price, early competitors manufactured modems using low-cost components that were often unreliable. Hayes quickly gained a reputation for high quality, and for a time held a 50% market share.
There’s a pretty neat Winlink project written in Go [1](https://github.com/la5nta/pat ) that I’ve wanted to use. If anyone is out there interested in collaborating, please let me know.