Ebb and Flow

Voice Card  -  Volume 17  -  John Card Number 15  -  Sat, Dec 1, 1990 6:13 AM

This is a response to Vol 17 Paul 7 ("Yes, we want it")...

Paul and I have been engaging in an interesting debate in our last few tapes and perhaps it is time to bring some parts of this discussion into the wider circle of Archipelago. This is a long card, but bear with me.

Almost from the beginning of Archipelago, Paul has been advocating a feature called "Distant Links" and I have been resisting this idea. On the surface, the idea is very simple. Currently, we can only establish reply links between the current volume and the previous volume. Paul asks, "Why not let us reply to ancient voice volumes, thus establishing links between the current volume and volumes in the distant past?"

As it turns out, this simple idea is enormously difficult and dangerous to implement. There are two basic reasons why this is so.

First, since it's not feasible to re-release a potentially large subset of all voice volumes with each new issue, Distant Links would have to be established by modifying each member's set of stacks after he receives his packet, that is, by remote control. Consider that a single link could theoretically trigger a cascade of updates across a wide range of volumes (since the hand buttons on all "sister" nodes would also have to be modified). Also, more and more members are being forced by shrinking hard disk space to archive ancient voice volumes onto diskettes. Thus members would be forced into some kind of installation procedure at the beginning of each session that might involve finding and inserting diskettes. There are all sorts of ways that something could go wrong during this procedure, and it seems likely that not all members would always install all such links, so that we could no longer assume that all collections were identical.

That's the first major category of problems. The second has to do with the fact that distant links involve jumping into the future as well as the past. In order for this feature to work, I have to build stacks today that are capable to linking to stacks that won't be created for many years to come. I have to add icons in current stacks for members who haven't even joined yet. And I have to contend with the distinct possibility that the design of voice cards may continue to evolve in the future.

These "interface-with-the-future" problems looked impossible to me at first glance. But upon deeper reflection I believe that it is possible to overcome these problems, provided that I make some significant changes to the underlying system architecture and stick to a few basic principles in years to come "no matter what." The problem is that these changes would be very difficult to implement and would come with a high price tag: larger, slower, more complex stacks, and, I fear, some restrictions and added complexity for the users.

This is a very difficult and technical issue which neither Paul nor I have fully explored. The two major problems give rise to dozens of more detailed complications. So far I have been able to find a way around each new complication, but always at an additional cost. The bottom line is that Distant Links are almost certainly possible, but are difficult and dangerous as well.

Why should the rest of you care? Well, it turns out that there are deeper issues at stake here. In some of his recent cards, Paul has raised the possibility of a more advanced technology that could forever free Archipelago from the clutches of the antiquated Postal Service. That is, we could someday soon transmit voice cards over the phone line. No more stamps for me to lick!

I have no problem with this. But there's more. Paul suggests that we could greatly speed up the pace of our conversations (when we wanted to). Instead of waiting for the next volume, each member could dial into command central and obtain all the new cards that had come in since the last time he called and could transmit new cards as soon as they were created. O brave new world!

My reaction to this visionary idea has been cautious. Even assuming we can pull this off (it's definitely possible, but it's not quite practical just yet), I wonder "Do we want it to happen?"

Paul is suggesting a fundamental change in the very shape and rhythm of Archipelago. There would be no more volumes and no more deadlines implicit in the structure. Instead, there would be a continuous stream of voice cards with members dipping into that stream as often or as seldom as they wished. In such a system there would HAVE to be distant links. We would no longer be synchronized and so some members could race ahead of others, making many exchanges before the slower members could jump into the fray.

There are already such systems in place, and Paul has access to one of the most magnificent. Such on-line conversations have a distinct flavor and many advantages (and some disadvantages as well). I am not necessarily opposed to any of these ideas, but I think it's worth considering what we would have to give up in order to make this technological leap.

I rather like the ebb and flow of the current volume-by-volume system. I like the fact that we are synchronized, and that we have periods of rest followed by clear deadlines. The lack of distant links keeps us focused on the present. And, like the issues of a fine magazine, each voice volume is a distinct creation with a personality all its own. Every month or so we all come together and create a tangible thing. There is a wonderful sense of closure here, a sense of completion, that would be lost in a continuous on-line structure.

The deadlines implicit in the volume-by-volume structure of Archipelago are perhaps it's most crucial feature. In his last card Paul states that "the current Archipelago use of U.S. mail forces an artificial pace on all conversations." But is this really true? In the month spent by each member creating a new packet, only three days, or about 10%, is consumed by Postal delays. What happens during the other 27 days?

Just because we could reply instantly to each new voice card doesn't mean we would. In fact, without the monthly deadlines and the familiar rhythm of the current system, I fear that many of you would drift away altogether. Structure, even when it imposes certain limitations, is not a bad thing, and in fact some structure is vital. In Paul's brave new world, the longer you delayed in picking up the phone, the farther behind you would get. How long would it be before some of you stopped picking up the phone altogether?

Archipelago is to letter-writing what On-line systems are to phone-calling. The phone is more efficient than the pen, or at least faster, but that does not mean the phone is BETTER than the pen. There's a lot to be said for letter writing. My point here is that the most important issues before us are not technological but psychological. It's not what we can do that matters, but what we want to do.

This debate, which started with an obscure technical question, comes down at last to this much more fundamental question: "What do we WANT out of Archipelago?"

On his last tape, Paul broached this question by asking me what I want out of Archipelago. Here's what I came up with:

  • I want mail. To some extent, Archipelago is a vast scheme to increase the amount of mail I receive in my lonely little mail box. On this level the current system is far better that the proposed phone-in system. Phone calls are interruptions but receiving mail is like finding packages under a Christmas Tree.
  • I want an audience. I love to create wacky new stacks and write strange stories. But it's just no fun unless SOMEONE SOMEWHERE gets to see these creations. On this level I have a slight preference for the current system. It's somewhat more satisfying to present my creations as part of a distinct volume than it is to merely dump them into the stream. There is a sense with each new volume of "the curtain going up."
  • I want to be remembered. This is perhaps the most selfish desire of all. It comes down to a futile attempt to cheat death. When I'm on my deathbed I want to be able to hand a little pile of disks or a cube or something to whoever is at my side and say "Here. Here is my life. Examine it. Learn from it. Enjoy it. Pass it on."
  • Finally, I want to preserve the essence of my friends. I want to have something I can hold on to that will remind me what they were like and what they did and how they changed.
I want to make Archipelago so big and so varied and so strange that it will be remembered and so outlast me. And I want copies scattered around so as to increase its chances of surviving. None of these desires are particularly logical or noble, but they are, nonetheless, MY desires and on the whole I couldn't be happier with the way Archipelago is working out.

Now then, if any of you are still with me, I really want to know:

What do YOU want out of Archipelago? Am I right in resisting the fundamental changes that Paul envisions, or am I just being stodgy and old-fashioned? In any case, Archipelago can certainly be improved: have you any suggestions? In particular, do any of you having a craving for some kind of Distant Links capability? Speak up.

I think what Paul and I are both reaching for is a clearer vision of what we want Archipelago to become. The technical problems can all be resolved. But the central problem is deciding what we want to do. A good way to start is for each of you to say just a few words about what you want.