Slashdot Mirror
On the Chaotic Evolution of Email?
TheCarlMau asks: "I'm doing research on the origins of email in the 70's and 80's. I'm particularly interested in how this technology was designed and implemented without any planned trajectory (ie: nobody sat down in 1970 and planned to create email as we know it today in 2006). As very little has been written on the history, I'm wondering if the Slashdot community could provide any insights, stories, or first-hand experiences? It seems to me, as a person who did not experience this 'revolution,' that the offspring of the ARPANET technology was hackish and sometimes chaotic. What do you think on this matter?"
Can be found by reading the Requests for Comment associated with that protocol. In other words there WAS planning involved- a good deal of planning- it's just that the end-users were completely different than the original audience- the original audience were arpanet researchers, whose system was so good it overtook the competeing FIDONet hackers- which resulted in spammers.
SJW: a person who perceives an injustice, and while correcting it, commits a greater injustice.
Then when I want to college (this wasn't much before every freshman was issued an email account and web space at orientation -- things snowballed really quickly) someone told me that there was a way to send messages by computers to other schools, for free. I went down to the bowels of the CS building and a moss-covered grad student gave me a Bitnet address that looked like the volume of the earth in cubic centimeters. In hindsight, the whole episode was like something out of Harry Potter.
What I'm listening to now on Pandora...
...when are features like PGP, anti-phishing, anti-spam, working rich text/html, etc. going to become part of a standard that everyone can use easily? Right now there are a bunch of "hacks" that patch up problems with email, and as such there is a huge difference in what some people get versus what others are stuck with. Is this something like IPv6 that will simply have to be mandated and rolled out with the cooperation of lots of large organizations?
Of course my article gets rejected
Ask Slashdot: On the Catholic Intelligent Design of Email?
Paying taxes to buy civilization is like paying a hooker to buy love.
Email was compelling from day one. The technology has changed, but only in details: bangpaths are gone and the abomination of HTML afflicts us. Popularity and exploits are results of the Metcalfe Effect.
But email is still very much email. `ytalk` has morphed into [G]AIM. WWW similarly unchanged although it has seen more technical changes, including a wholesale shift from gopher:
http://www.amazon.co.uk/exec/obidos/ASIN/074346837 6/qid=1137543821/sr=8-2/ref=sr_8_xs_ap_i2_xgl/202- 4708284-5091803
Provides a good background to how the internet came about, including a chapter on email.
Will get you going.
You are being MICROattacked, from various angles, in a SOFT manner.
There was email long before there were networks. http://www.multicians.org/thvv/mail-history.html describes early email on Multics, between multiple users time-sharing one mainframe.
It seems to me, as a person who did not experience this 'revolution,' that the offspring of the ARPANET technology was hackish and sometimes chaotic. What do you think on this matter?"
I think you should count yourself lucky you missed it. Just a few of the many joys:
Great. Well, now I know what I'm going to be having nightmares about tonight.
--MarkusQ
You might do well to check postings by Mark Crispin in the comp.sys.next.* USENET archives, sometime around 1988-1992. NeXT developed an early version of mail with fonts, colors, rulers (margins/tabs/spacing/justification.etc.), embedded pictures (TIFF, PostScript), attached files in arbitrary locations in the text, a picture of the sender, and embedded sound clips. Basically the program created an RTF file with attachments, tar'd it up, uuencoded it, and sent it as a plaintext message. Worked perfectly -- if you had a NeXT computer to read it on! To my knowledge, no other system had email even remotely as sophisticated as this.
Mark began working on a related project: MIME. This was done at U Washington, which developed MIME in conjunction with pine and pico. He spent a lot of time on the NeXT USENET lists posting vitriol about how much better MIME was going to be than NeXTmail. In retrospect the postings, and responses, give a lot of insight into how MIME was shaped, developed, and of course how it was influenced by NeXTmail.
First, let me say, Google is your friend. But since I'm really nice, I'll ask Ray Tomlinson:
http://openmap.bbn.com/~tomlinso/ray/home.html
"I sent the first network email in 1971 using a program I wrote called SNDMSG. I have written a brief account of the first email with the intent of forestalling some of the more common questions about that event. If you want to see what the computer used to send the first email looked like, you will find that here too."
The latest Slashdot meme.
The Catholic Church officially recognizes evolution as a valid theory, they do not believe in 7 day creationism or any of that other ID bullshit.
Here's quite an illuminating link on google groups.
judging from the subject lines of the spam i get i would say it's more chaotic evil
Snowden and Manning are heroes.
The issue with going right to the RFCs is that E-Mail was widely deployed before Internet access was. Corporations and government used inhouse systems such as IBM PROFS, Lotus ccMail, and even MS Mail. There were large non-RFC mail networks, including MCI, AT&T, and Worldcom's Lotus Notes network (that had something like a million users when the Internet was far smaller).
When Internet mail started to catch on in the early 90s, the Internet Mail capabilties were rather obviously kludged into these systems, usually with a funky addressing scheme such as "joeblow@example.com @ INTERNET", difficulty with file attachments, etc. Microsoft even introduced a X400 based product in 1994 where it was clear that SMTP was an after-thought. It was only around 2000 when SMTP was integrated into Exchange and Notes as a core protocol, rather than a gateway.
Many of the features that people from the Internet Mail tradition find distasteful, such as Top-Reply and Rich (html) Text come directly from the capabilities of corporate systems. Any sort of comprehensive history of email has to include these systems, rather than just the Unix boxes with their sendmail.
Finally, let me just complain that the RFCs for Internet Mail took a very simple spec and turned it into a complete fricken mess, with all sorts of ridiclous, overly-complex encoding crap for back-compatibility with 7-bit systems. It would be nice if someday someone flushed all this MIME crap and started over with a nice clean protocol like HTTP.
Business. Numbers. Money. People. Computer World.
I started a firm in 1985 that specialized in design and implementation of email systems. this usually required linking some existing mail services (PROFS, DISOSS, All-in-One, Wang Office, etc) together with online services (MCI Mail, Compuserve, etc) and LAN based mail (MHS compliant products, VINES Mail, CC:Mail, Quickmail, many others). In addition there were the first generations of Mail-enabled products (group calenders and other groupware) that had their own benefits and requirements.
ISO protocols and design considerations were the most important; UUCP was a joke and sendmail feeble, although we did eventually work with the first Motif/MIME compliant mail app. Interenet protocols did not matter until later on, and then largely because of the failure of the ISO community on one hand and the popularity of browser-enabled Internet on the other.
The largest problem that I saw (but few others at the time agreed with) was directories, directory-integration, and directory-enabled apps. Today this is ubiquitous (not just for mail but for all identity management) and therefore invisible. Thne it was a nightmare that required careful planning and usually lot's of custom coding.
Eventually we had some better tools: X400 backbones, X500 and meta-directory products, content managers, and some additional challenges (wireless particularly, PKI integration, etc) and finally serious Internet apps including the first great LDAP server (Critical Path) and the first scalable Internet mail server (Sun SIMS).
The events of 2001 lead me to shut down my firm early in 2002. I have not thought much about technology since but if you are interested feel free to contact me directly at terrymccarthynyc@yahoo.com for more history. I also may be able to put you in touch with some of the real pioneers.
Regards,
Terry McCarthy
Several solution will in combination stop most of this.
1. Accept email only if sender is in your address list just like IM is handling it.
For example: "Incoming email REQUEST from blah@sp.corp.com, which is not in your address list. Would you like to [Accept] or [Deny] this email?"
So the request (not the email itself) would have to be confirmed. Press [Accept] and the confirmation is sent back to the sender telling the server it is alright to send the email.
2. Email should be text only (no HTML/rtf or URL hyperlinking) so that anyone and anything (for example text readers for blind people) can read them. Sure, you can set up the client right now to do this, but I would like to see all clients denying any message which contain html since it is one of the biggest (if not the only) way that phishing use.
3. Secure the system so your PC does not become a spam-zombie-bot.
Anti-spam will be/is a makeover at best, the real root of the spam problem must be solved once and for all. 1 and 2 in combination will stop most of this if 3 is done properly.
And finally, as you said, PGP/GPG as standard practice.
I just wanted to say thanks to everybody that replied. I may be in touch with a few of you in the near future, if you don't mind.