Explanation of my current policy regarding checking my e-mail

The situation until 2004.Oct

• Definitely spam for which my software knew where to complain, in which case it'd immediately send a spam complaint there less than a minute after the spam arrived.
• E-mail from somebody important, such as the sysadmin here, Mike Durkin, or one of my white-listed penpals or correspondents or collaborators regarding fighting spam, in which case my program would bleep my terminal to tell me that I had new e-mail and tell me what category it was in and whom it was from. I'd then quickly suspend whatever I was doing to see what the new e-mail was about.
• E-mail from any source whatsoever that had the magic ReSpam2002Jan08 keyword in the Subject field, as a way of new non-spammers to introduce themselves to me and communicate important information (such as if I was sending my spam complaints to the wrong address, each complaint told that keyword for the receiver of my complaint to get through my filter to let me know I was screwing up), and for new penpals to correspond with me long enough that I trust them enough to then put them on my explicit white-list so they wouldn't have to use the special keyword any longer.
• E-mail from anyone whatsoever which came to me forwarded from any of my Yahoo! Groups. So I even got alerted whenever anyone posted to a Yahoo! Group that I was monitoring. This included both regular groups that I managed or monitored, where I wanted to know immediately whenever anyone posted or joined or un-joined etc., and the one special group that I advertised publically RobertMaas@YahooGroups.Com where anyone could join and send me e-mail and thereby not need the special ReSpam2002Jan08 keyword even during very first contact with me. This was handy because most people making initial contact failed to RTFM and sent me e-mail without the word, and if it was from an IP address block that had been spamming me then of course my software would assume it's more of the same and send an auto-complaint. But via that special Yahoo! Group they were assured of *not* being treated as spam even even if they e-mailed me from the most prolific source of spam on the net.
• Likely spam from some source for which I didn't yet have CTW (Complain To Whom) data, not for that IP address block itself, nor for any upstream ISP, in which case the program would automatically connect to various online sources such as SpamCop and ARIN/RIPE/APNIC WHOIS to collect data and analyze it to see if it can figure out a spam-complaint source. But that case was very rare. Most of the time I had CTW upstream that could be used immediately, then later at my leisure I could run the program that gathers more direct CTW data if I wanted.

In addition to sending my spam complaints to a CTW address for an ISP that was responsible for sending me the spam, my program would automatically also send a copy to a Yahoo Group that I had created to keep copies of spam complaints. Here's an example of such a copy of spam complaint. Notice that it clearly tells how to get individual e-mail to me via that ReSpam2002Jan08 password in Subject field. In that way I would be protected from the deluge of auto-acks that flooded my mailbox for no purpose except to harass anyone who gets a lot of spam and complains about it all, but still the admin at the spamming site could get any important info to me if he/she bothered to actually read my spam complaint. To learn how to contact me.

Spam had made e-mail totally useless up to 2002 when I started writing my AntiSpam software. (Now you know why the special keyword has a date in 2002, because that's when I installed that special keyword and the software to deal with it. From that day until 2004.Oct, my software was working, and it got better as I developed it further.) There were hundreds of spam coming in for every legitimate message I ever got. I'd go weeks without any legitimate e-mail, meanwhile getting deluged with thousands of new spam. But my software made it all manageable, so I could receive e-mail from white-listed penpals, or from the sysadmin here, or from anyone new using the special ReSpam2002Jan08 password, and be alerted to any such new worthwhile e-mail about once ever few weeks when I got any, and in the mean time I could work on whatever I wanted to (writing more software, discussing issues on newsgroups, etc.) all those weeks between useful e-mail and not have to manually scan the utteer cesspool of inbox spam even once the whole time because my software auto-scanned it when I logged in and whenever new e-mail arrived. So I wouldn't have to spend any time whatsoever (except starting up my AntiSpam program every time I logged in) dealing with spam the rest of the time, and all the spam I got was automatically complained-about without hardly any further effort from me after I had spent more than a year building up all that AntiSpam software to maintain databases of IP address blocks and CTW addresses and upstream TraceRoute data etc. etc.

Note that BIFF here is broken. It says I have new e-mail, but doesn't give me the slightest idea where it came from. 99.9% of the time it's spam and it's a waste of my time to go check my inbox to see what it was. So with my AntiSpam /var/mail/rem-date-last-written loop running I could turn off BIFF and not be bothered at all when new spam came in, except for a little printout that my mail-check loop did when it had decided what kind of e-mail had arrived. So I'd just take a brief break from work whenever my screen had new-email stuff showing on it, then as soon as the handling of the new spam was finished I'd just refresh my screen and get right back to whatever I had been doing at the time.

What happened 2004.Oct that destroyed my way of quickly knowing if there was new non-spam e-mail for me:

• I was getting a very large amount of spam from Savvis. The admin at Savvis took it upon himself to complain to the admin here about my spam complaints. Apparently Savvis wanted to send spam and considered any complaints about them to be abuse of Savvis. The admin at Savvis bothered my own admin so much that one day I got a message from my admin ordering me to cease sending complaints to Savvis. So I was forced to change my software to check if the spam source was Savvis, and if so then not send a complaint there, but instead send it to a new Yahoo! Group where I collect such complaints.
• Then a few days later I was ordered by my admin not to send any more spam complaints to *anyone* whatsoever. So I was forced to change my software to send *all* spam complaints to various Yahoo! Groups where I was collecting archives of them per IP number block or ISP. I had to make several trips to public libraries to create the new groups because my only access from home is text only, whereas Yahoo! Groups requires seeing a visual image every time somebody creates a Group or joins one. Then after getting home I had to check each newly created group to make sure it was accepting e-mail from here, and then upgrade my AntiSpam software to send complaints to the appropriate Yahoo! Group. I was almost totally burned out by the effort it took me to modify my software to comply with my admin's no-complaints-any-more policy.
• Then very short time later, on 2004.Oct.22, I got a message from the admin here ordering me to cease *all* running of my AntiSpam software whatsoever, because it consumes too many CPU cycles. This left me with no allowed way whatsoever to (1) check incoming e-mail automatically to alert me if there was any non-spam and otherwise not bother me every time a new spam came in, (2) copy spam I'd received to a public archive where I could show people what I'm suffering, etc. etc.
• In all the time since that day, I haven't received any new e-mail here on my Unix shell account from that admin or from anyone else that I ever correspond with, only spam, spam, spam, spam, spam, and more spam. Every once in a while I check my system mailbox here, but I've never seen any legitimate e-mail in it when I look, so my time checking it was totally wasted, and I felt stupid for checking it. It's just not worth my time to check if maybe there's some legitimate e-mail for the first time since 2004.Oct.22, so maybe I won't ever again. In any case, since 2004.Oct.22 I've gone weeks without looking even once at my system inbox. In the future it might be *never*. So if you sent me any e-mail from 2004.Oct.22 to present, to my shell account mailbox, that's why I haven't replied. You need to remember when you sent me the e-mail, and what e-mail address you sent it from, and see instructions later below for how to alert me online.
• In 2005.Jan, somehow my e-mail address here on this Unix shell machine, which I had been keeping secret ever since we switched from the old tsoft machine to this new rawbw machine, must have gotten posted in public, because the rate of incomimg spam increased greatly.
• Although I hadn't been allowed to run my AntiSpam software since 2004.Oct.22, still whenever I logged in I ran a quick shell script that simply checked my system inbox to see whether there was anything from the admin here or anything that had the ReSpam2002Jan08 password in it. Nothing ever showed up in all that time until on May.01, when I saw that I had e-mail addressed *to* the admin here. It was spam! Apparently the admin here is now accepting spam for him but putting it in my inbox instead of his. Over time, several more of those spam to him showed up in my inbox, to where it was spewing out a whole crap of that every time I ran my shell script (which runs grep) at login time. Finally I edited my system inbox to trash his e-mail address in those several spam so it no longer shows up in my shell-grep script, so if he *ever* sends me legitimate e-mail I'll notice it at login without having to sift though all those false matches in spam. But if I'm online when he sends me e-mail, I won't notice it until many hours later when I log back in again and run that durkin/respam-grep script. Since October Mike hasn't allowed me to run any of my own AntiSpam software that alerts me when I get non-spam e-mail.
• 2005.July appx. 20 I finally figured out a new way to get an alert whenever somebody important (not spam) sends me e-mail. For several years I've maintained some CGI software that allows people elsewhere on the net to connect here and run various programs that I had set up. At first (2001.Jan) I had a demo of my SegMat algorithm for grading short-answer quizzes and for teaching spelling and foreign languages etc., applied in my demo to having users guess answers to riddles. Then (2001.Mar) I installed new software to guess what foreign language a body of text is written in. In late 2002 I wrote some more CGI software, including a login form whereby people could run FINGER to see if I was online (for example if they tried telephoning me and the line was busy), and where I could log in to send e-mail, and where I could demonstrate my flashcard algorithm and my students could run it themselves. I even had a feature where I could leave a short message for my users and they could see it when they log in. But the only way I could see whether anyone had logged in and might have seen the message I left for them, was for me to manually check the directory there, which I got too exhausted to do and didn't notice until May or June that Annie Boone had logged in in February and had at that time seen my message that hadn't been updated since last year sometime, oops. But I couldn't figure out any way that I could get bleeped at my terminal whenever somebody logged in. All the system programs to write to a terminal work only if they are run from a program that has a terminal of its own. They don't work from CGI scripts. But in late July I finally came up with an idea, based on a trick I had used on my Linux laptop to communicate between Gnu Emacs and BeanShell Java: I set up a FIFO on the disk and then one program can write to it while the other reads from it. So what I did here is set up a shell script that repeatedly tries to read from the FIFO and spews anything it receives out to its controlling terminal, which is mine here where I'm logged in if I invoked the script during the current login session. (I have my login script now automatically invoke it in background every time I log in now.) Then my CGI login script, upon accepting anyone's login, transmits some text to that FIFO to let me know, and immediately I'm bleeped with that info. Sometime shortly earlier, I set up a guest1 account, with password free, so that anyone anywhere even if they don't have an account can still log in to see demo of my flashcard system. Then I wrote a new feature in my CGI login system, whereby anyone logged in can send me a short online message. Whenevr they submit that message to be written to disk, the CGI program bleeps me to tell me that, so I know to immediately look at the disk file to see what they wrote. (And I can edit that file myself, so if the remote user goes back into that edit-message mode by refreshing that WebPage a few minutes after bleeping me, they can see anything that I edited into that file, so I can leave them such short replies, which they'll see if they do indeed get back into send-message with reloaded WebPage at that time.) So now for the first time since 2004.Oct.22 I have a way for people to bleep me with a message when I'm online.

As I've said many times above, the administration on this shell machine doesn't provide any system program that would tell me whenever I get new non-spam e-mail, and since 2004.Oct.22 I've been forbidden to run any of my own software to filter spam from non-spam and alert me to the latter when it comes in, so if you send me e-mail there's no way for me to know if you did, except if you bleep me via this new Web-based system I set up 2005.Jul.appx20. Also, disk space here on this shell machine is very limited. My allocation is 200 megabytes, and I'm already sitting at 199 megabytes, and that doesn't even include the spam I've received since October when I totally stopped reading my system inbox. I don't have enough disk allocation to hold the spam I've already received, much less any new flood of spam if I were to let this e-mail address be known in public. By comparison, Yahoo! Mail increased their allocation on *each* account to 1000 megabytes. My rem642b account already had over 237 megabytes of spam as of 2005.Jul.24 and is getting another hundred or so spam every day, but with 1000 megabytes allowed there's plenty of room to continue accumulating spam, and occasional false positives such as the very important e-mail I received from a fellow in London shortly after the bombings which I didn't see until several days late when I discovered them in my Bulk Mail folder. But at least on Yahoo! Mail I *can* continue to receive new e-mail without overflowing my allocation (for a couple more years anyway), so I recommend anyone wishing to communicate with me to send e-mail to my Yahoo! Mail address, and then connect to my Web site to alert me about the e-mail, so if it is in Bulk Mail folder I'll look specially for it.

Don't send me any e-mail on my shell account, because I haven't checked there on any regular basis since 2004.Oct. Send to rem642b@Yahoo.Com instead.
Update: I no longer check e-mail there. See here instead.