pre-www internet communication
13 views
Skip to first unread message
Kai Baker
Jun 1, 2017, 1:56:51 PM6/1/17
to collexion
I'm always fascinated with old computer technology and also fascinated by how things used to work. I was looking at some old bulletin board discussions (linked below) and I found some really exotic emails and what appears to be forwarding mechanisms. For example:
hoptoad!g...@lll-crg.ARPA
g...@hoptoad.uucp
Path: utzoo!mnetor!seismo!rutgers!caip!clyde!cbatt!cbosgd!ted
utzoo.UUCP
irdis%vpi....@CSNET-RELAY.ARPA
What's the deal with the ! and the % in the arpa emails? What's this uucp? Is that related to the usenet network? Are any of these systems still operable? These are details never covered in a history about the internet.
Paul Eberhart
Jun 1, 2017, 2:28:41 PM6/1/17
to coll...@googlegroups.com
Ooh, I learned this a while ago, even though it predates my exposure to the Internet by about a decade (and the Internet by about 5 years). That's not _quite_ Internet, not _quite_ usenet, and not really BBS, but in that weird era and niche in the late 70s/early 80s between completely disparate computers and networks and a interconnected standardized network.
What you are looking at is mostly UUCP https://en.wikipedia.org/wiki/UUCP from back when intermittent point-to-point connectivity was the norm. You can still talk UUCP, but unless you really work to set up your own little network, it's generally tunneled over TCP/IP and thus not super interesting.
Bang paths are specified routes. Rather than an _address_ you could give a _path_, like bigsite!localsite!host!user. That way messages would go to and cache at places with "fast" persistent connections and get passed on when the appropriate connectivity happened. This makes sense in the UUCP machines-periodically-calling-each-other context. Now we have persistent connections, and a small priesthood who understand the unrelenting horror that is modern BGP holding the universe together.
Percent addresses are [a shitty hack for] relays between not-generally-interconnected networks; that's a worst-case scenario type connection in your example: I think what is going on there is a CSNET (a different proto-internet, similar to ARPANET but especially for universities and researchers) address being sent to from an ARPANET address via a UUCP path. So the sender talks SMTP locally or over TCP to some box that has TCP/IP-UUCP bridge and a known chain of UUCP connections to another ARPANET site, which has a relay to a CSNET site the user is at.
The livinginternet folks have a pretty good description of connectivity in that era here: http://www.livinginternet.com/e/ew_addr.htm
What you are looking at is mostly UUCP https://en.wikipedia.org/wiki/UUCP from back when intermittent point-to-point connectivity was the norm. You can still talk UUCP, but unless you really work to set up your own little network, it's generally tunneled over TCP/IP and thus not super interesting.
Bang paths are specified routes. Rather than an _address_ you could give a _path_, like bigsite!localsite!host!user. That way messages would go to and cache at places with "fast" persistent connections and get passed on when the appropriate connectivity happened. This makes sense in the UUCP machines-periodically-calling-each-other context. Now we have persistent connections, and a small priesthood who understand the unrelenting horror that is modern BGP holding the universe together.
Percent addresses are [a shitty hack for] relays between not-generally-interconnected networks; that's a worst-case scenario type connection in your example: I think what is going on there is a CSNET (a different proto-internet, similar to ARPANET but especially for universities and researchers) address being sent to from an ARPANET address via a UUCP path. So the sender talks SMTP locally or over TCP to some box that has TCP/IP-UUCP bridge and a known chain of UUCP connections to another ARPANET site, which has a relay to a CSNET site the user is at.
The livinginternet folks have a pretty good description of connectivity in that era here: http://www.livinginternet.com/e/ew_addr.htm
On Thu, Jun 1, 2017 at 1:56 PM, Kai Baker <eigen...@gmail.com> wrote:
I'm always fascinated with old computer technology and also fascinated by how things used to work. I was looking at some old bulletin board discussions (linked below) and I found some really exotic emails and what appears to be forwarding mechanisms. For example:
hoptoad!g...@lll-crg.ARPA
g...@hoptoad.uucp
Path: utzoo!mnetor!seismo!rutgers!caip!clyde!cbatt!cbosgd!ted
utzoo.UUCP
irdis%vpi.csnet@CSNET-RELAY.ARPA
What's the deal with the ! and the % in the arpa emails? What's this uucp? Is that related to the usenet network? Are any of these systems still operable? These are details never covered in a history about the internet.
--
You received this message because you are subscribed to the Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to collexion+unsubscribe@googlegroups.com.
To post to this group, send email to coll...@googlegroups.com.
Visit this group at https://groups.google.com/group/collexion.
For more options, visit https://groups.google.com/d/optout.
Paul Selegue Eberhart (PAPPP) - http://www.pappp.net/
Messages from this account may be in regard to PAPP...@gmail.com, pse...@uky.edu or pse...@engr.uky.edu.
Messages from this account may be in regard to PAPP...@gmail.com, pse...@uky.edu or pse...@engr.uky.edu.
Dave
Jun 3, 2017, 4:16:03 PM6/3/17
to coll...@googlegroups.com
Wikipedia is your friend:
https://en.wikipedia.org/wiki/UUCP
I remember the days of the pre-www internet, mostly from some personal experience (My brother operated a BBS for a few years.). Most of the BBS systems were stand-alone, and only supported local users, although there was some networking. The problem was that the only communications medium available was dial-up telephone lines, starting at a ridiculously slow speed. My first modem was a 300 bit per second acoustical modem, although I pretty quickly upgraded to a 1200 bps direct connect modem. From there, the speeds leap-frogged up to 2400, 4800, 9600, 14,400, 28.8K, 33.6K, and, finally, 56K bps, at least on phone lines which would support it (Ours only handled about 33.6Kbps reliably).
https://en.wikipedia.org/wiki/Modem#List_of_dialup_speeds
The problem, of course, is that just about any non-local dial-up connection incurred long-distance rates/fees, and these could stack up quite rapidly (For those old enough to remember, inter-state calling was often cheaper than intra-state calling.). Given that a 1200 bps connection only transfers data at a theoretical maximum of 120 characters per second, and, practically, by the time you account for hand-shaking, error detection/repeats/retrys, and headers, the actual effective rate was usually well under 100 characters per second. Thus, given that a lively discussion could often run for many 10s or 100s of K in text, the cost of forwarding those messages could run into some real money. And, if any user was silly enough to try and embed any kind of graphics, they'd find themselves quickly banned.
As a result, a lot of BBSes were strictly stand-alone and local. Of the few which did do message exchange, most restricted themselves to a rather narrow group of messages.
If you're interested in some of the other esoteric communications techniques, remember that ham radio also developed a messaging system, using AX.25 packet mode.
https://en.wikipedia.org/wiki/AX.25
There were a couple of problems with this. One was that, due to the strictly non-commercial nature of amateur radio, absolutely no commercial traffic was permitted. Another was that there were language restrictions (e.g., no cussing, etc.). The communications speeds were abysmal (300 bits per second for HF, and usually 1200 bits per second for VHF). Still, some messages did get distributed via this mode.
Dave
--------------------------------------------
On Thu, 6/1/17, Kai Baker <eigen...@gmail.com> wrote:
Subject: pre-www internet communication
To: "collexion" <coll...@googlegroups.com>
Date: Thursday, June 1, 2017, 12:56 PM
I'm always
fascinated with old computer technology and also fascinated
by how things used to work. I was looking at some old
bulletin board discussions (linked below) and I found some
really exotic emails and what appears to be forwarding
mechanisms. For example:
hoptoad!g...@lll-crg.ARPA
g...@hoptoad.uucp
Path:
utzoo!mnetor!seismo!rutgers!caip!clyde!cbatt!cbosgd!ted
utzoo.UUCP
irdis%vpi....@CSNET-RELAY.ARPA
What's the deal with the ! and the % in the
arpa emails? What's this uucp? Is that related to the
usenet network? Are any of these systems still operable?
These are details never covered in a history about the
internet.
https://www.cs.rutgers.edu/~cwm/NetStuff/Human-Nets/Volume9.html
http://www.columbia.edu/kermit/ftp/e/mail.87a
http://tech-insider.org/data-security/research/1986/1003.html
--
You received this message because you are subscribed to the
Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails
from it, send an email to collexion+...@googlegroups.com.
https://en.wikipedia.org/wiki/UUCP
I remember the days of the pre-www internet, mostly from some personal experience (My brother operated a BBS for a few years.). Most of the BBS systems were stand-alone, and only supported local users, although there was some networking. The problem was that the only communications medium available was dial-up telephone lines, starting at a ridiculously slow speed. My first modem was a 300 bit per second acoustical modem, although I pretty quickly upgraded to a 1200 bps direct connect modem. From there, the speeds leap-frogged up to 2400, 4800, 9600, 14,400, 28.8K, 33.6K, and, finally, 56K bps, at least on phone lines which would support it (Ours only handled about 33.6Kbps reliably).
https://en.wikipedia.org/wiki/Modem#List_of_dialup_speeds
The problem, of course, is that just about any non-local dial-up connection incurred long-distance rates/fees, and these could stack up quite rapidly (For those old enough to remember, inter-state calling was often cheaper than intra-state calling.). Given that a 1200 bps connection only transfers data at a theoretical maximum of 120 characters per second, and, practically, by the time you account for hand-shaking, error detection/repeats/retrys, and headers, the actual effective rate was usually well under 100 characters per second. Thus, given that a lively discussion could often run for many 10s or 100s of K in text, the cost of forwarding those messages could run into some real money. And, if any user was silly enough to try and embed any kind of graphics, they'd find themselves quickly banned.
As a result, a lot of BBSes were strictly stand-alone and local. Of the few which did do message exchange, most restricted themselves to a rather narrow group of messages.
If you're interested in some of the other esoteric communications techniques, remember that ham radio also developed a messaging system, using AX.25 packet mode.
https://en.wikipedia.org/wiki/AX.25
There were a couple of problems with this. One was that, due to the strictly non-commercial nature of amateur radio, absolutely no commercial traffic was permitted. Another was that there were language restrictions (e.g., no cussing, etc.). The communications speeds were abysmal (300 bits per second for HF, and usually 1200 bits per second for VHF). Still, some messages did get distributed via this mode.
Dave
--------------------------------------------
On Thu, 6/1/17, Kai Baker <eigen...@gmail.com> wrote:
Subject: pre-www internet communication
To: "collexion" <coll...@googlegroups.com>
Date: Thursday, June 1, 2017, 12:56 PM
I'm always
fascinated with old computer technology and also fascinated
by how things used to work. I was looking at some old
bulletin board discussions (linked below) and I found some
really exotic emails and what appears to be forwarding
mechanisms. For example:
hoptoad!g...@lll-crg.ARPA
g...@hoptoad.uucp
Path:
utzoo!mnetor!seismo!rutgers!caip!clyde!cbatt!cbosgd!ted
utzoo.UUCP
What's the deal with the ! and the % in the
arpa emails? What's this uucp? Is that related to the
usenet network? Are any of these systems still operable?
These are details never covered in a history about the
internet.
https://www.cs.rutgers.edu/~cwm/NetStuff/Human-Nets/Volume9.html
http://www.columbia.edu/kermit/ftp/e/mail.87a
http://tech-insider.org/data-security/research/1986/1003.html
--
You received this message because you are subscribed to the
Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails
Kai Baker
Jun 3, 2017, 8:50:19 PM6/3/17
to collexion
Thanks so much Paul! I can count on you for knowing the most obscure ins and outs of computer history. That is great! I love seeing how things used to work. It was much more organic, in the sense that it was new and still developing, which made it interesting. I really miss that and I want to bring it back.
On Thu, Jun 1, 2017 at 2:28 PM, Paul Eberhart <papp...@gmail.com> wrote:
Ooh, I learned this a while ago, even though it predates my exposure to the Internet by about a decade (and the Internet by about 5 years). That's not _quite_ Internet, not _quite_ usenet, and not really BBS, but in that weird era and niche in the late 70s/early 80s between completely disparate computers and networks and a interconnected standardized network.
What you are looking at is mostly UUCP https://en.wikipedia.org/wiki/UUCP from back when intermittent point-to-point connectivity was the norm. You can still talk UUCP, but unless you really work to set up your own little network, it's generally tunneled over TCP/IP and thus not super interesting.
Bang paths are specified routes. Rather than an _address_ you could give a _path_, like bigsite!localsite!host!user. That way messages would go to and cache at places with "fast" persistent connections and get passed on when the appropriate connectivity happened. This makes sense in the UUCP machines-periodically-calling-each-other context. Now we have persistent connections, and a small priesthood who understand the unrelenting horror that is modern BGP holding the universe together.
Percent addresses are [a shitty hack for] relays between not-generally-interconnected networks; that's a worst-case scenario type connection in your example: I think what is going on there is a CSNET (a different proto-internet, similar to ARPANET but especially for universities and researchers) address being sent to from an ARPANET address via a UUCP path. So the sender talks SMTP locally or over TCP to some box that has TCP/IP-UUCP bridge and a known chain of UUCP connections to another ARPANET site, which has a relay to a CSNET site the user is at.
The livinginternet folks have a pretty good description of connectivity in that era here: http://www.livinginternet.com/e/ew_addr.htm
On Thu, Jun 1, 2017 at 1:56 PM, Kai Baker <eigen...@gmail.com> wrote:
I'm always fascinated with old computer technology and also fascinated by how things used to work. I was looking at some old bulletin board discussions (linked below) and I found some really exotic emails and what appears to be forwarding mechanisms. For example:
hoptoad!g...@lll-crg.ARPA
g...@hoptoad.uucp
Path: utzoo!mnetor!seismo!rutgers!caip!clyde!cbatt!cbosgd!ted
utzoo.UUCP
What's the deal with the ! and the % in the arpa emails? What's this uucp? Is that related to the usenet network? Are any of these systems still operable? These are details never covered in a history about the internet.
--
You received this message because you are subscribed to the Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to collexion+unsubscribe@googlegroups.com.
To post to this group, send email to coll...@googlegroups.com.
Visit this group at https://groups.google.com/group/collexion.
For more options, visit https://groups.google.com/d/optout.
--Paul Selegue Eberhart (PAPPP) - http://www.pappp.net/
Messages from this account may be in regard to PAPP...@gmail.com, pse...@uky.edu or pse...@engr.uky.edu.
--
You received this message because you are subscribed to the Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to collexion+unsubscribe@googlegroups.com.
To post to this group, send email to coll...@googlegroups.com.
Visit this group at https://groups.google.com/group/collexion.
For more options, visit https://groups.google.com/d/optout.
Paul Eberhart
Jun 4, 2017, 11:44:18 AM6/4/17
to coll...@googlegroups.com
Part of why I included that quip about BGP is to remind that the functioning of the Internet is (by design) still a continuing miracle of consensus and cooperation among the people who make it work. It's more formal than it once was, but the informal consensus traditions are baked into the processes for things like IETF (Internet Engineering Task Force, the people who bless RFCs as standards).
What I do miss is the easily-comprehensible simplicity of early computers, we've got what Danny Hillis (founder of Thinking Machines, among other things) has been calling "The Entanglement" going on, where in order to handle all the edge-cases and make things "easy," technologies have become incomprehensibly complicated even for experts. One of my favorite computing talks is a 2012 SciAm interview with Bill Joy and Danny Hillis, https://www.scientificamerican.com/podcast/episode/the-coming-entanglement-bill-joy-an-12-02-15/ about that.
What I do miss is the easily-comprehensible simplicity of early computers, we've got what Danny Hillis (founder of Thinking Machines, among other things) has been calling "The Entanglement" going on, where in order to handle all the edge-cases and make things "easy," technologies have become incomprehensibly complicated even for experts. One of my favorite computing talks is a 2012 SciAm interview with Bill Joy and Danny Hillis, https://www.scientificamerican.com/podcast/episode/the-coming-entanglement-bill-joy-an-12-02-15/ about that.
Kai Baker
Jun 4, 2017, 1:32:32 PM6/4/17
to collexion
The supposedly coming entanglement sounds real gross and ugly. Its aesthetic is absolutely horrid.
Kai Baker
Jun 5, 2017, 12:58:40 PM6/5/17
to collexion
The first time I used the internet was in 1994. By 1995 I became hooked when I realized I could research previously difficult to find information. Our first internet provider was eWorld. We connected through a 2.4k baud modem. I am quite familiar with long distance calling back then since we lived in Taiwan and we had to call across seas to the US. My parents didn't expect it would be used much. So when they got an $800 phone bill one month they decided to switch to a local internet provider. :p
I'm real fascinated by old networks. Apparently Linux has a DECnet port and 3 people already downloaded it this week. Is DECnet still used? If so, how do I connect to it, assuming I install it on my linux kernel?https://sourceforge.net/projects/linux-decnet/
irdis%vpi.csnet@CSNET-RELAY.ARPA
What's the deal with the ! and the % in the
arpa emails? What's this uucp? Is that related to the
usenet network? Are any of these systems still operable?
These are details never covered in a history about the
internet.
https://www.cs.rutgers.edu/~cwm/NetStuff/Human-Nets/Volume9.html
http://www.columbia.edu/kermit/ftp/e/mail.87a
http://tech-insider.org/data-security/research/1986/1003.html
--
You received this message because you are subscribed to the
Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails
from it, send an email to collexion+unsubscribe@googlegroups.com.
To post to this group, send email to coll...@googlegroups.com.
Visit this group at https://groups.google.com/group/collexion.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups "Collexion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to collexion+unsubscribe@googlegroups.com.
To post to this group, send email to coll...@googlegroups.com.
Visit this group at https://groups.google.com/group/collexion.
For more options, visit https://groups.google.com/d/optout.
Mike Andrews
Jun 10, 2017, 7:41:51 PM6/10/17
to coll...@googlegroups.com
DECnet was more a LAN thing than a WAN thing -- think of it more
like Novell's IPX or Microsoft's NetBEUI inside of a corporate
network. It's kinda specific to the OpenVMS operating system,
which was designed concurrently with the DEC VAX-11 computers
(successor to the PDP-11), then ported to DEC Alpha, then Intel
Itanium, and now an Intel Xeon port is in progress. DEC
themselves were eaten by Compaq in the late 90's, and Compaq was
then borged by HP. OpenVMS mostly uses DECnet for file sharing
and clustering -- and nowadays they can run that over TCP/IP
instead.
The only time I saw DECnet used on a WAN basis was between some
Ohio universities, to carry BITNET traffic from one school's VAX
to another school's VAX. BITNET was something that supported
email, instant messaging, very very very crude file transfer, and
basically nothing else, none of it exactly in real-time. It
pretty much ran only on IBM mainframes, or VAXes running software
that emulated it, and mostly connected colleges rather than
companies. It was shut down in the mid 90's when it became
obvious the Internet could do everything it could but better (I
was the one that got to unplug Wittenberg University from BITNET).
To unsubscribe from this group and stop receiving emails from it, send an email to collexion+...@googlegroups.com.
Mike Andrews
Jun 10, 2017, 7:41:56 PM6/10/17
to coll...@googlegroups.com
Intra-state was taxed different than inter-state -- just like sales tax
is now. Has to do with the Commerce Clause in the US Constitution.
is now. Has to do with the Commerce Clause in the US Constitution.
nx
Jun 10, 2017, 8:23:12 PM6/10/17
to coll...@googlegroups.com
Speaking of pre-www, I'm highly interested in building a gopher service for 2017.
Kai Baker
Jun 10, 2017, 9:43:55 PM6/10/17
to coll...@googlegroups.com
That sounds interesting. Perhaps we can convince other gopher enthusiasts to organize and create an internet parallel to http service? ^_^
--
Reply all
Reply to author
Forward
0 new messages