rs79.vrx.palo-alto.ca.us

Cyberculture Clash


Cyberculture Clash


In 1995 the ad hoc UUCP network was larger than the Interent. By the next year it wasn't and vanished shortly after. The nascent www and ubiquity of dial up access consumed the UUCP network. But what of UUCPs governance? Does that mesh with IP network governance model of today? What might we have missed, and at what cost?
The UUCP and NSF-IP networks grew up side by side from different environments. Both are well documented and the governance of each reflects the culture they emerged from.

The UUCP network began with the birth of Unix and was simply a way to copy files from one machine to another, presuming there was a direct point to point connection to the other machine by way of a serial port or dialup modem usually. RS232 serial ports for low speed data communication had been an integral part of nearly every computer made since the mid 1970s - it was how the computer talked to the printer. Serial ports and printers we found with nearly every computer in that era, they were ubiquitous.

So it seemed like the most logical thing to do in the Unux environment where "cp" meant "copy file" to make "uucp" mean "copy this file to the other machine I'm connected to"; the priter was unplugged as it was more useful for a computer to talk to another computer than to make a printout. Paper, your day of judgement was decided on that point in time.

This was on the easy coast, in the Bell labs that were the sole repository for all things Unix, spread between Cherry Hill, NJ (R & D), the birthplace of Unic and C and Napierville (Operations) where the machine "ihnp4" lived.

On the west coast at USC/ISI a parallel effort was being made with the IP and NCP protocol suits. At the time when the Unix program uucp was being written, the IP network community was working on FTP which did the same thing as uucp - copied a file to another computer, but unlike uucp it could be any computer on the IP network, not just one that you can a direct connection to. This was to be the reason for the eventual widespread adoption of the IP network, but we're decades away from that happening at this particular point in time.

So computers began putting files on each others disk drives, computers that were attached by a myriad of cables to each other. The physical link had been made but of course there was no software to use. There was no network software to speak of and certainly no network applications. Clouds where something that brought rain.

And this was how email worked. Email itself had been around as an application for a while, because as soon as timesharing was invented and you had the ability to let more than one person at a time use the computer (Previous to this, you ran your deck of punched cards, waited for it to think, then got your printout. Don't laugh, this got man to the moon and back. What have you done lately on Twitter?) people were not the only person using the computer and it seemed natural to want to leave a note for other people on the computer.

This was a powerful concept, so powerful it's the 19th article in the UN declaration of human rights - the right to communicate, a principle so lasting lasting it even explains the success of things such as the Twitter of today.

Email systems wern't immediatly hooked up to these new network connections. There was a time when poeple were merrily emailing each other: "Uh, the printer is on fire." or "slept through class. up all night hacking. anything interesting in class?" but that was just to other users on the same computer. But, even despite the lack of a network connection, group collaboration had alredy begun made possible by the availabilty of a multi user computer: the "gripe" file at the University of Waterloo in 1975 or the "pnews" program at Lord Elgin High School in Burlington Ontario in 1974 were exmaples of would become our online communites and blogs of today - with no network yet. These two were just waiting to happen.

And they did happen, it did not take very long for networking to begin to change the nature of application software.

At first, the IP network people would FTP a file to another machine, and poeple at the other end would check a certain place in their computers filesystem to see if anything had shown up there. I'm not kidding, that's really how it worked and to a great extents works today. Your email program is still exactly just this process, automated with a graphical interface.

In the Unix community poeple used the uucp program to copy files to one another as a sort of virtual way of passing notes around, but it did not take very long for the unix email program to be able to send a message to a machine one "hop" away, that is the next machine over, the one it had a cabled connection to.

Machines suddenly had to have names, because you needed to identify what machine you were on the network so other machines would now where to send a response. And if you were on a machine called bob, instead of using email to contact a user called "dmr" you owuld instead say "alice!dmr" which meant "on the machine ``alice'' the user named ``dmr''".

This expanded very quickly as machines learned to act as intermediary "hops" and not endpoints so a path of machines was exlicitly listed: bob!carrol!ted!alice!dmr and when your environment are all unix

.