deja vu

[cellio]

In the early days, email was unreliable and slow. We had to route messages by hand, and each machine in the network might process its UUCP queue only once or twice a day. It could take days for a message to get to its recipient, and sometimes it didn't get there at all.

Then, the ARPANet became dominant and email was faster and more reliable. Most people using it were "practicioners" of various sorts; the average guy on the street didn't yet have email.

Then two things happened: spam, and widespread email access (making spam even more profitable). Now, almost 10 years after the green-card lawyers created the first piece of spam and AOL sent out its first mass (physical) mailing of software, we're all swimming in spam, and virii, and lures (for the unsuspecting) to "just click here", and the attendant side effects. So we have filters and auto-processing of various sorts, and this month most of us aren't inspecting bounced-message reports because of all the SoBig-generated false ones, so we don't even know if we're seeing all our mail or if the mail we send gets through. Sometimes we have to send messages several times, or actually tell someone "I sent you email", before the recipient actually sees it.

25 years after the rise of UUCP, email is unreliable and slow again. Oops.

Comments

Not having problems

[laid.livejournal.com]

The only filtering I run is checking the body of messages for "TVqQAA", which is the base64 encoding of the first few bytes of the .exe header. This stops just about all the trojan email. Ohh, I also have a filter in place for things that get sent to similar usernames on adelphia.net, and things that get sent to me@buf.adelphia.net.

Combined with the heavy handed method I use to deal with spammers and their providers(forwarding the spam that I recieved to everyone at the institution from the president to the stable boy), I have very little spam to deal with.

Re: Not having problems

[cellio]

I don't do any automated filtering, though my provider does, mostly for virii. (I didn't get 900 copies of the Microsoft "patch" that one of my friends did.) I delete 'em manually, based on subject/sender. (I'm not skilled with procmail.) But I've noticed that mail I send isn't always being received/seen, because of filtering or an over-active "D" key at the other end. That's frustrating.

Re: Not having problems

[filkerdave.livejournal.com]

I'm getting about 2-300 of the current MS patch mesages per night right now (as in, between my last cleanout of the night and my first in the morning).

Bleh.

Re: Not having problems

[goljerp]

Yeah. What if I want to send an E-mail to Joy saying "I love you" in the subject? What if I want to send an e-mail to my Goljerp mailing list, which is mainly BCC'd to stop people from accidentally replying to everyone? What if I wanted to send a non-viral .exe to someone for some reason?

Hmm... maybe if E-mail 2.0 required each message to be signed by a PGP key, and each person could get one free one a year. (They'd pay 2^n for each nth one after that; corporations could get 10 corporate keys free, and pay 2^n for each 10 keys afterwards... a company of 30 employees ought to be able to afford a one-time cost of $100...) Of course, there are other problems with this system, but at least you'd be able (ideally) to know that whoever sent the message is who they claim to be. Spammers could still exist, of course, but it would be more expensive, because any key used for spamming would be rejected...

Re: Not having problems

[geekosaur]

There are folks here (i.e. at CMU) working on such schemes. Unfortunately, the consensus is that no such scheme would ever achieve critical mass.

As for rejecting PGP keys of spammers: go look at the list of things SoBig.F did to infected machines again. Now combine that list with the presence of PGP keys and the ability to record the keystrokes used to unlock the private key... that battle has already been lost, spammers can already acquire and use people's public and secret keys if they so wish, and I fully expect them to do so if it should become necessary --- the reputations of their victims are meaningless to them; we know that already as well. (Don't assume they have any limits to what they will do based on civility, morality, ethics, etc. Even Canter and Siegel are still out there, hopping ISP accounts --- and hopping states as they get disbarred in each.)

Re: Not having problems

[goljerp]

Sigh. You're probably right...

Re: Not having problems

[miz-hatbox.livejournal.com]

I hear you. There are people I want to get in touch with that I haven't seen in a while, and my email address has changed. I don't mind carefully composing messages, but having to obsess over the perfect subject line just to get the message read? Sheesh. It's like having to draft little tiny cover letters for all my emails!

[cellio]

tiny little cover letters

Yes, it does feel like that sometimes. Gone are the days when I can put something like "hey, long time no see" in a subject line...

Email...

[patsmor.livejournal.com]

Interestingly enough, there was an article in CIO magazine last year about companies who had gone almost entirely to email going back to voicemail and (gasp!) phone calls because their email was so junky. Sigh.

Re: Email...

[cellio]

A similar trend: we've all pretty much abandoned Usenet (or so I gather) to go back to mailing lists.

Usenet

[patsmor.livejournal.com]

As far as I can tell, that's true....

Re: Usenet

[dglenn.livejournal.com]

Oddly enough, a few groups have started being useful again, after folks abandoned them long enough for the spammers to give up.

[dagonell.livejournal.com]

'Most people using it were "practicioners" of various sorts; the average guy on the street didn't yet have email.'

I apologize in advance, this is going to come out snarky to some folks no matter how I say it.

Most folks had a .edu account. And then when Compuserve et al started up, the academics grumbled that the bar had been lowered and now just anybody without any sort of technical background could communicate over long distances to everybody without racking up a huge phone bill to which all the ham radio operators I know were saying "That's what *we* said about Usenet!" :D

nit

[cellio]

Actually, back in the heyday of UUCP, nobody had .edu addresses. They hadn't been invented yet. :-)

...{seismo|rutgers}!cmu-cs-g!mjc

Re: nit

[gregbo.livejournal.com]

Even if domain names had (officially) been in existence back then, there still would have been a lot of people with .com, especially all the people who worked at Bell Labs and the other AT&T/Bellcore/etc sites.