Slashdot Mirror
Where Are The Founders Of The Dial-Up Revolution?
RIMBoy writes "The Atlanta Journal-Constitution recently tracked down the founders behind the dial-up modem revolution. The founders of Hayes Micromodem set the standard with their AT Command set. While Dennis Hayes finds himself inducted into the Computer Industry Hall of Fame, at the same time he is broke (with a stop as a bar owner) and trying to find the next big thing. Dale Heatherington cashed out early and has dedicated himself to several projects, including ham radio."
It's been covered on slashdot many times so I'm sure people will remember, but there is a BBS Documentary in the works.
The history of such revolutions should be documented for future generations to learn from.
..Jeff Keegan
seven syllables explain TiVo: kee gan dot org slash ti vo
I remember reading that the 56K limit was legal, not technical (and that this legal limit is actually something like 53K:
"In the U.S., the FCC places a power ceiling on phone lines of -12dbm average per 3 second interval. X2 modems work within this by restricting throughput to 53kbps in the U.S. X2 modems can theoretically work at 56k, although they are constrained to operate 5% slower than this in the U.S. (Some users have reported occasional connections past 53kbps.)"
(from this page
Don't blame Durga. I voted for Centauri.
As I understand it 56K (or actually 53K) is the max transmission limit imposed by the FCC in the US. I seem to recall some fear about damaging the POTS infrastructure or something along those lines - not sure if I remember the rationale properly.
Relive the good ol' days at textfiles.com
Because that's all the bandwidth there is.
Most calls get digitized by the phone company, and the 53K modems take that into account to get almost all of the theoretical bandwidth. I know someone will correct me, but I think that most phone calls are digitized as 64Kb data streams. There may be some overhead in that, lowering the theoretical maximum throughput.
Of course, if all the phone companies upgraded their equipment to some different standard, they could probably support significantly higher data rates. But then again, isn't that called DSL?
Between the Central Offices, the connections are digital and multiplexed. The amount dedicated to each channel is 64k with 8k used for switching information. So while it's possible to run better than 56k over a phone line pair (DSL does it at least for limited distances), once you hit the CO, the 56k limit comes into play.
You asked for it ... ;)
In the US the phone lines are digitized with 8000 Hz and 7 bits, resulting in a bandwidth of 56 kbps. In Europe 8 bits are used, giving 64 kbps. I can't remember off-hand what Japan uses (they mix happily european and US standards )
So you can't go above 56k and hope to sell your modems in the US, thus losing at least half of your potential customers. It's just not theoretically possible.
This is somewhat off-topic, but I think the honor of that particular title should be "The World" http://www.TheWorld.com operated by Software Tool & Die. Since 1989, the first public dialup Internet Service Provider (ISP) on the planet. And we're still proud to be the best. These other guys may have set up the technology, but "the revolution" is another matter. Its like crediting K&R for starting Open Source. Not quite.
Back in 1995/1996 when 56K modems were becoming the rage I also folded shop and sold my mid-sized ISP that was serving 2 cities. Hayes modem cards in a 19 inch rack chassi were the standard then, 33.6 was the MAX you could get on a good day and ISP's like me that spent the long dollar for the real modems instead of a pile of crap sportsters like one company I remember you could get that speed. (I started as an ISP when 14.400 was the fastest you could get.)
56K killed it for most of us... T1's required for incoming lines as well as horribly priced interfaces for the 56K dial up side made it impossible for the medium/small guy to survive. the Small towns I was going into and started out with 3-4 modems now had a minimum of 24 incoming lines because of the T1 requirement. each dial in node now doubled all it's costs for operation and quadrupled it's costs for equipment.
Dial-up died when 56K came around.
Do not look at laser with remaining good eye.
I also remember using "split" modems which were asymmetric -- 1200 downstream and IIRC 150 upstream -- which prefigure today's ADSL.
Actually in europe they have E1 (~2 mbit as apposed to ~1.5mbit total), not a T1(aka DS1) with 30 channels and they can and do run something they call "E1 PRI" over those for 29 B channels and a D channel.
What you described is US PRI T1 which is 23 B channels with a D channel in the US at 64K each(this is what isdn service is based on, you can also run standard telco calls over them). US also has the standard T1 which is 24 channels as you described.
In Japan they call theirs a J1 (or PRI J1) and its based on the US standards, only in the yellow alarm generation/detection and the crc-6 calculation methods.
man motd
man login
touch ~/.hushlogin
More importantly, as I've mentioned Ward, with Randy Suess, also INVENTED THE BBS when this very same Dennis Hayes sent them one of his original 300 baud autodial/auto-answer modems.
Ward will tell you fun details like why CBBS looks for the modem's RING result and then sends the ATO to make the modem answer. CBBS never puts the modem into auto-answer mode.
Why? So that if the CBBS program wasn't running happily, the caller wouldn't waste money on an answered phone call to a BBS that wasn't working.
Ward takes more credit for CBBS than the MODEM* protocol because MODEM was written quickly to fix a problem (sending program files to Randy over the modem-modem link) but CBBS was planned. Ward says MODEM was a response "like a sneeze" He doesn't like taking credit for a sneeze.
* - The real name of the protocol is MODEM. Ward's original MODEM comm program had an option to auto-receive files,. XMODEM was MODEM with the option. When you're the first you don't put in version qualifiers.
Ever dream you could fly? Get up from the Flight Sim. I Fly
Those were the days when you would dial up some BB and hear EEEEE aaaaaa iiiii shhhhhh oooo bong bong bing (you get the point....)
Ah, but if you concentrate you'll remember that before 56k (or maybe 28.8) modems, it didn't do the boing boing noise at the end. It ended with the static sshhhhhhh, (and maybe had a short even "aeaea" tone or two over the static), and then cut out. The boing boing sound was a shocking late development in modem handshaking art.
There are no trails. There are no trees out here.
Jobs got Woz to write the Atari games.
And Jobs stiffed Woz on the money Atari paid them.
See:
This Google cache of www.woz.org/letters/general/09.html.
See also:
This second google cache of www.woz.org/letters/general/53.html
You'll need to scroll down a little, but the paragraphs to look at have highlighting in them.
I loved ZMODEM because it got files in bigger 1024kbit chunks?
:-)
Hold your horses, zmodem only went to 8192 byte blocks. 1Mb blocks would suck if you had to wait 20 minutes for each block to be retransmitted
Edwin
bash$
http://www.wa4dsy.net shows the data modem he constructed, and also has pages of info on his robots.