taylor@hplabsc.UUCP (02/26/87)
[this is something from a discussion on the well about the future of mail systems...various people were making comments about system A being better/worse than system B and so on, and I just had to jump in] To me the most important feature of an electronic mail system, as with any system on a computer, is that it allows me to do what *I* want and manages to figure out and act based on that. Let's restate this a different way: Any good system matches the expectations of the user, and not the other way around. So this means that if people want to have an electronic mail system that models the familiar U.S. Postal Service, then it should do the whole job, with return-receipt-requested acknowledging that the remote host got the message, with "registered personal mail" that sends a message (that can be verified authentic) when the user actually accepts (reads) the mail, and so on. With a system like this, we can have addresses like "username/local-address/city/state/country/unique- identifier-for-faster-sorting". If we want to exceed this system, we need to realize that it will not necessarily be a 'best fit' or even particularly compatible...but I think it is of vital importance for us to get beyond the parochial postal deliver systems we have now (most of the postal service is based on ideas from hundreds of years ago, stuff that was 'innovative' when they were using Ponies to get mail across the US). But let's look at the U.S. Postal Service as a model for a bit before we dismiss it out of hand... In the past ten years, postal service has, for the most part, improved quite a bit, but at quite a cost. Postage has at least doubled in this time and the complexity of sending mail has increased (e.g. nine-digit zipcodes). And yet, even with this increased performance, the actual services offered by the postal system are relatively few. Ten years ago there wasn't "overnight express mail", but is that really such a big step? For that matter, we've had "registered mail" for a LONG time, but it is still only used in special cases. One interesting thing that the Postal System offers that there is no current electronic analogy to is 'insured mail'. This implies that the postal service acknowledges that `not all mail is created equal'. Yet the mail systems available via computer assume that the message is of equal importance and value (actually, there are some exceptions, but we'll get to them later)... The postal service has improved noticably due to the introduction of competition to the previously monopolized mail delivery area, but even the competition seems to acknowledge the lack of excitement and profit in the day-to-day delivery of mail - they are almost exclusively in the 'overnight mail' section of mail. UPS, on the other hand, has become the major shipper of packages by taking another 'uninteresting' area and wresting control away by offering a service custom designed for the shipping of packages. The U.S. Postal Service still ships packages, and oftentimes for less, but it is a more common perception nowadays that UPS is for packages and the U.S. Mail is for letters and (!) bills. There are a couple of other interesting phenomena that occur with U.S. Mail that we should mention before moving into the "Electronic Age". First off, junk mail is more than just an annoyance - it is a 'cultural artifact' of our times. It is a sign that our mail addresses are no longer private property, but are yet another commodity to swap about (the Whole Earth even sells their mailing list for quite a profit. Indeed, there are magazines that make money solely from selling their mailing list and from advertising revenue - they don't even charge subscribers postage cost). This is really quite a significant change, I think. In previous years where you lived was very personal information, held quite closely, and considered almost like the 'true name' of a witch or warlock. This is quite a sociological change, one that would certainly have disturbed George Orwell. (and this change was the motivation behind the introduction of 'post office boxes' for people who had home mail addresses (used to be for people who DIDN'T have that...)) The second interesting phenomenon is the reputation that the U.S. Postal service has garnered throughout the years. There are very few people who won't tell you a horror story or two about 'the mail system' (not to mention the 'urban myth' of "it's in the mail...") and yet the system is incredibly reliable, and rather amazingly fast overall. So the question arises as to WHY people perceive this apalling entity when we have such a fine service... (note that I don't mean to imply that it is perfect! Far from it! But it is certainly an impressive system when you consider the sheer volume of mail that flows through the system each day...) I really don't know what has caused this reputation, but I'd suspect that part of it is peoples' nature to be suspicious of large, monolothic organizations, and the postal system certainly is that. It is also one of the few government operations that straddles the line between private enterprise and public ownership, so that causes confusion and resentment too. With these thoughts in mind, let's finally turn to the area of Electronic Mail, what we're trying to accomplish and what we expect from a system. The first thing is for us to be able to build up over time personal address books that we can zip through and get a full address by just remembering the first letter of the last name of the person we want to send to, if not even a more sophisticated system, where I could say something like "send the following message to Joe in Berkeley" and have it say "woah! There are 6 people named Joe that live in Berkeley and two in Oakland. Which do you mean?". We could then either 'play 20 questions' or simply browse the list of people it has picked out. Now we hit our first interesting privacy question...I have a phone book at home that contains the names, address and phone numbers of my friends, business associates, and other people and places that I want to be able to access easily. At home it is implicitly private - you would have to break in to my house and rummage about to find it. On a computer the concept of ownership is much more fuzzy, however, and I don't particularly feel that I could have an `on-line' copy of my addresses there. Again, not something to solve, per se, just something to consider. Once the system has agreed with me that I am going to send mail to someone that exists, it should then check that the address I have is valid and current, and if it has changed, ideally, to tell me what the new one is *and update itself*. This is quite a step from the Post Office sending back "Adressee unknown" mail. At this point it would be nice to have it able to tell me "Dave, are you aware that Joe is on vacation in Hawaii for two more weeks?" before I compose a message about a party the next night. Privacy issue #2: while it is nice to be able to let people know that you aren't able to answer your email at the point they send it, what is really needed is a system that knows who all these people sending mail are, and can send different messages (or none at all) to them based on the relationship with you. (e.g. if one of my best friends drops me an email message it should say "I'm on vacation", but if I get 'electronic junk mail' it shouldn't tell them ANYthing about me. This is 100% analogous to the warnings in telephone answering machine booklets about not saying "I'm not home" or "I'm on vacation"). To compose the message, I fully expect at *least* the same ability to send stuff that I have with a regular mesage. I have received (paper) mail with cartoons, articles clipped out of magazines and newspapers, photographs, and even (ahem) items of clothing. While we can argue that transmitting physical things via an electronic medium is ridiculous, I think the concept of being able to trivially add doodles, attach pictures and other text and so on are *vital* to the success of the medium. This also means that composing the message should be similarly trivial - it certainly doesn't take much learning to write a letter, once you learn how to use a pencil and how to form letters. (In fact, you can send meaningful mail even before you learn how to 'write', per se. Anyone who has received illustrations from a child knows that). The key I'm trying to point out here is that we need to have a system that allows us to *COMMUNICATE* (in glowing neon). We 'send email' on computers currently, bashing about the ASCII character-set to force us to have the (vital) ability to send non-word information. This has been in printed language for thousands of years - just consider the function of punctuation or the ubiquitious parenthetical comments (like this). To have computer systems become an effective medium for communications (again, that word should be in flashing neon) we need the ability to transmit far far more information than even the most sophisticated multimedia systems allow. Having NAPLPS graphics, for example, is nice, but it does NOT allow the spontenaity of a doodle. A side thought: graphology is something that most of us know enough about to look at someones writing and say "they were tense" or "what a feminine script". That is because even at the level of the formation of individual letters we are communicating - in this case emotions and personality trends. Yet more information that cannot be transmitted via computer. How many people have a shoebox in their closet with old letters? Or a stack of love-letters wrapped in a bag to take out and read on a rainy afternoon? I do. But I also have quite a large collection of old electronic mail, and I only use that for reference purposes (as in "umm...you said something about that a while back...let me go look"). I think the biggest thing we need to watch out for in this medium (electronics) and with electronic mail specifically, is the loss of the personal traits and the sheer ability to *communicate*. I'll close this incredibly long posting with a suggestion that everyone go browse the book "Zen and the Art of the Macintosh" and think about why it is such an innovative book... -- Dave Taylor --
taylor@hplabsc.UUCP (03/03/87)
(Note from the submitter: sorry - but this must be kept anonymous if posted since I am involved with one of the unannounced products. I don't think this reveals anything, but I know that they are quite paranoid) In your posting you say: >To compose the message, I fully expect at *least* the same ability to >send stuff that I have with a regular mesage. I have received (paper) >mail with cartoons, articles clipped out of magazines and newspapers, >photographs, and even (ahem) items of clothing. While we can argue that >transmitting physical things via an electronic medium is ridiculous, I >think the concept of being able to trivially add doodles, attach pictures >and other text and so on are *vital* to the success of the medium. There are a number of R&D efforts along these lines. I recall one at BBN which has been written up occasionally. The R&D efforts involved too much expensive equipment to be practical for anyone outside a few very rich government agencies. However, there are already postings in the mac.binaries where the full repertoire of Mac graphics tools can be sent electonically. There are also products emerging that are specifically aimed at various commercial markets to support the electronic shipping and annotation of documents. In fact, if you expand your definition of electronic mail to include the new generation of facsimile equipment this capability is already widespread. All of these are low enough cost to be commercially affordable. Personal affordability should arrive within a decade. The only real gap is in the user interface arena, where nothing is yet available that can compete with the ease of use and expressive power of pencil, pen, or crayon. Here I don't even know of any close contenders. [Note from the moderator: The system being discussed here is in fact the Diamond-II multimedia mail system from Bolt Beranek and Neuman (BBN), a part of the EXPRES project (a DoD funded project for the submission and tracking of documenents). I think it would be very interesting to have someone from BBN reply to this comment about the expense of the Diamond system (a Sun 3/75 to start) and it would also be interested to have some people from the EXPRES project discuss it here... --- Dave]