Bob Kerns: Thinking

The Computer History Pantheon

[This is a post I made to Livescribe's forum on 2008-06-01. Livescribe is no more; the closest equivalent are some tablet apps.]

I was listening to the start of Jim Marggraff's podcast this evening, and he made mention of Vannevar Bush's Memex idea.

It occurred to me that probably most people don't really know the origins of the ideas behind all this fancy hyperlinked multimedia technology we're coming to take for granted these days. And this site needs a link...

Vannevar Bush was an MIT professor, Vice President and Dean of Engineering from 1932-1938. His idea originated during this period, but was put on hold during the war period, while he (among other things) ran the Manhatten Project and advised Presidents Roosevelt and Truman.

But after the war, he dusted off the idea, and published it as an article in the Atlantic Monthly. This single article, in July of 1945, over 9 years before I was even born -- this ancient article, is the true root of all our modern idea of Internet, and of the World Wide Web.

A number of years ago, Atlantic Monthly put this article up. If I recall correctly, they continued to make it free even while they experimented with paid web access.

The article is entitled As We May Think. I highly recommend anyone who has not read it before, and is interested in the direction and possibilities take a little time out to read this article. It is written from the perspective of microfilm technology, which younger readers won't even have heard of, and yet it anticipates technologies that weren't developed for another 50 years, and which still have not been applied to the extent outlined in this article.

I know of no visionary article, on any topic, at any time, which has had the impact, the long-lived influence, and the amazing degree of fidelity into the future as this article.

It has inspired all of the major players in the evolution of what we do with computers: Douglas Englebart (inventor of the mouse and much more), Ted Nelson, who, in his 1974 book, Computer Lib/Dream Machines, coined the term "hypertext". This book tried to be a prototype of what it advocated -- extremely non-linear in form, leading you on a journey from point to point throughout the book. I sure wish I still had my copy....

And let's not forget J. C. R. Licklider, whose 1963 memo, Memorandum For Members and Affiliates of the Intergalactic Computer Network led directly (and quickly) to the creation of the Internet. Nine years later, when I came to MIT and Project Mac, the lab he founded and later directed at MIT, the internet (then called ARPAnet) was already up, running, and developing its own culture.

There are two projects which, combined, start to come close to what Vannevar Bush envisioned 70 years ago.

The first is Wikipedia itself. (There is a reason I chose Wikipedia articles for my earlier links!).

The second is Gorden Bell, best known for being VP in charge of Engineering at Digital Equipment Corporation (DEC). He is now capturing his entire life digitally, in the MyLifeBits Project.

Against this historical backdrop and vision of capturing, integrating and communicating our daily human experience, the pen stands out as a holdout, the last important mode of human communication to integrate.

I myself, have shunned the pen in preference to the computer for these 36 years, precisely because of this lack of integration.

This is why it is so interesting, and exciting, for me personally, to see Livescribe focus on a platform for re-integrating the pen. It's not about capturing handwriting. It's not about capturing audio.

It's all about the integration. Pen with audio with location with events with phone with network with ideas captured with ideas borrowed with ideas shared.

Vannevar Bush's vision was constrained by the technologies he had at hand -- largely mechanical and photograhic. Yet look how far he saw.

When you look at what the Pulse Smartpen can do today, don't look at it as being about capturing writing done in the old way. Try to look ahead at what writing could become for you, once integrated with the rest of your Life Bits. How would you use writing differently?

You, too, can be a pioneer.

In this historical canvas, Lifescribe is just a small yet-unpainted corner. But I'd argue it is important (if it succeeds) precisely because they occupy that last unpainted section, that last unmet need.

That last blank space, where will be writ, by Livescribe and others -- by all of us, the final signature, of the end of the Age of Paper, and the true beginning of the Digital Millennium.

Addendum:

It just occurred to me, I should add one more project to the list that are completing Vannevar Bush's Memex vision:

Brewster Kahle's Internet Archives captures the history of the internet. If you're looking for something that used to be there, then you need only go to the archives, and check the Wayback Machine. If it was there long enough, the Internet Archives will have one or more snapshots of it. It is also other film & video, and audio, text, and other digital artifacts.

UTF-8, Replacement, and Data Corruption

I happened upon a nice-looking Lisp implementation, ABCL (Armed Bear Common Lisp), which runs in and compiles to the Java JVM, integrating with the JSR-223 Java Scripting API.

Unfortunately, I found on their blog, this article.

[Note: Corrected URL]

Basically, they're trying to handle "bad UTF-8 input" by substituting a character ('?' in this case) for what they don't understand.

I tried to post a response, but my first attempt wouldn't OpenID authenticate against my typepad account (this blog), and resulted in my finely-crafted prose, better than this in many ways, being discarded.

So I rewrote it with what you see below, and tried again with my Google account. First it complained that my HTML was > 4096 characters -- and submitting a smaller partial input also made the same complaint, despite being well below that limit. Presumably some rewriting to HTML was pushing it over the limit.

Finally, I go it to accept it -- only to get a Google error message that the URI was too long to process. Integration failure!

So I post this on my own blog, and direct it to any and all who would consider trying to be nice by "handling" the "bad input".

---

Please, please do not do this!

Your intentions are good, and this seems like an obvious good thing -- but it's not. It's a trap.

This is such a serious mistake as to completely disqualify your system from usage for many purposes (unless one bypasses your text input stream handling entirely!). What Java does here is correct and vital. Do not bypass it.

I quite understand the instinct to just "make things work" in the face of invalid input. For example, Interlisp's DWIM -- Do What I Mean. More aptly titled, "Do what Warren Teitelman taught Interlisp to guess that I might mean."

This had the unfortunate effect of allowing bugs to slip into systems undetected.

You're thinking along the lines of Jon Postel's "be conservative in what you do, be liberal in what you accept from others", but that's a different situation, seeking robustness in establishing communication, not integrity of the data -- and even there, it had a lot of unfortunate consequences and has largely been abandoned in favor of more strict compliance.

Or consider the history of HTML, and the myriad messes which have resulted from browsers being liberal in what they accept coupled with page authors being sloppy.

I consider what you are doing here to be EXTREMELY seriously bad. You are, of your own initiative, silently corrupting data. There are few worse things a computer can do!

I cannot begin to tell you how often I have had to track down bad data caused by EXACTLY this behavior. Such data is usually detected FAR from the bad input. Often, extensive and expensive manual fixup has been required. Often, repair has simply been impossible.

You state "not everybody produces strict ascii or strict utf8 sources." Well, kind of true. People often produce strict ISO-8859-1, for example. That's the situation here, apparently. Rather than fix your real bug (reading an ISO-8859-1 file as UTF-8, thus potentially corrupting the data, if you're unlucky, or getting an error if you are lucky, as in this case) -- you are simply eliminating the error -- guaranteeing you'll be unlucky and the mistake will go undetected, until someone has to investigate why there's this funny '?' in some output somewhere, when some quite different appeared in the source code!

Input files can be in strict ISO-8859-2, -3, etc. or strict shift-JIS, or various strict combinations of {EUC, ISO-2022} and national coding systems.

The reality is, interpreted in the correct format, malformed input is extremely rare. When it does occur, it is nearly always the result of concatenation of raw byte-data that encodes different formats. For example, a source file that has been edited as UTF-8 by one author, and ISO-8859-1 by another, and then merged in a revision control system.

That's something to fix (the guy with the ISO-8859-1 editor needs to fix his setup), not paper over and ignore.

The solution here is simple: If you find you are reading the file in the wrong format, provide a restart to select the correct format, and reread it -- from the beginning, restarting all processing.

In most cases, the correct action will be to convert the input to UTF-8. In this day and age, nobody should be using any lesser character set, nor any other format for data interchange. (UTF-16 or full 32-bit unicode are suitable for internal processing, but will cause problems for interchange as support is limited in most environments).

And yes, I am aware that a lot of tools still default the charset to the platform's encoding, which for legacy reasons often defaults to some national encoding. Eclipse, for example. This is unfortunate, but data corruption is not the solution.

The only exceptions I see to not doing substitution: (1) Providing visual context to the user about where the problem lies, to aid in investigating and correcting the input -- e.g. selecting the right encoding, or re-encoding the input to the expected encoding, or (2) situations where proceeding without failure is vital, and data integrity is secondary to that goal. (Reading source files does not begin to qualify!)

So please, please, do not do this.

--- Argument by Authority section :=) ---

Developer on:

• MacLisp
• Macsyma
• Symbolics (including Japanese support)
• ANSI X3J13
• CLIM
• Various character-set/encoding libraries.
• Emacs's Edict library

My distaste for this strategy has been honed and reinforced over decades of experience with it.

And you have a lot of company in this, so I've had a lot of opportunity to get experience with it, unfortunately.

So please do not do this!

My First Computer

I just discovered ibm1130.org. It seems they have a SIMH-based IBM 1131 emulator available. OK, I should try to get that up on my Nexus One phone, to go with my first timesharing system (the PDP-10-based ITS operating system, from the MIT AI lab).

This is the message I sent them when I signed their guestbook:

The 1130 was my first real computer. I was a high school junior, and it was the closest computer to me -- 30 miles, at a local college (Morningside College, Sioux City, IA). I got a chance to take a Fortran class. I went on to teach myself assembler, and I remember well discovering how the 1132 printer interface worked, and the narrow timing margins for filling in the bitmask for each character. I never made a serious attempt at writing a driver for it, but I did enough playing around to know what I'd have to do.

Eventually, they got a 1403 and retired the 1132; that surely ended any thoughts of writing my own device driver! (I have written device drivers since, and I suppose the experience made me unintimidated of diving into the low levels, but I generally prefer to operate at a much higher level -- such as expert systems, etc.)

Before going off to MIT, I learned that they actually had an 1130, so I brought my decks of cards with me. I never used them, though, because 1) I'd have had to do some finagling to get access, and 2) I discovered timesharing. I only used punch cards for one class (on a 370, some assembler and some PL/1 as I recall).

There were other smart kids taking that Fortran class, but I think I'm the only one who stuck around for the next 18 months, showing up about twice a week, exploring anything I could get my hands on. Some of them actually had been able to take calculus; I remember being amazed that one kid was writing a program to compute pi, using a Taylor-series expansion. I couldn't figure out how I could find such things out. When I got to MIT, I found myself surrounded by people who'd had more opportunity in high school in traditional areas like math. But getting access to that 1130 was a huge advantage for going into computers.

Originally, I was going to do electronics. In fact, the whole reason I applied to MIT was because they had their EE and computer programs in the same department, rather than having computers are part of the math department. (I ended up ONLY applying to MIT -- University of Michigan never replied with an application form, and University of Iowa's application deadline was long after I was accepted at MIT).

I also got my first exposure to electronic simulation -- they had a copy of ECAP (ECAP II?) which I played around with a bit. I had some trouble using it; I think it was problems with numerical stability.

I never got to use the disks, and I was at the mercy of whatever disk people had in there. I could usually do Fortran. I did get to do pretty much everything else, including single-stepping through my programs at the console. If I recall correctly, toggling C000, D000 into the first locations did a load/store that wiped memory, filling it with the D000 value...

I'll have to see if I can get the 1130 simulator up and running on my Nexus One phone. I currently have the PDP-10 OS I used at MIT -- ITS -- running there.

My first open source software

(Via Sam Ramji): Garrett Serack posted his first experience, 25 years ago, with open source software.

He asks:

So, I gotta ask, when was your Open Source moment? What was the first piece of Open Source software you used? Did you play with the source?

Well, I suppose it was MIT's ITS operating system (Incompatible Timesharing System). The code was, in fact, shared elsewhere -- and I have it running today, in a PDP-10 simulator on my Google Nexus One phone!

I first logged into ITS in 1972, though I didn't get seriously involved with it until later.

I don't recall making changes in the actual source. I did, however, patch the running system from time to time, to do various amusing things. And I maintained the command processor/debugger (DDT), and at one point hacked my running DDT to be able to patch other people's running processes. I then transcribed those changes back into the source.

This was useful for a weird purpose. Having no security, we'd occasionally get people who'd come along for the sole purpose of causing trouble -- harassing people, usually, but occasionally actually trying to do damage. (Fortunately, security through obscurity generally actually worked on these guys).

Kind of like the people you see commenting on Ziff-Davis blogs and newspaper sites these days. A lot more power, but the same mind-set.

You could get rid of these people -- but they'd just come back, and take it as a challenge, usually changing identity in the process.

So what I'd do was to patch their command prompt, to give the illusion that the system was flaky and unusable. I'd stick in a sleep, to make it look slow (and to slow them down), and then make it give creative error messages.

I still find this to be a good strategy today, for moderating forums. In the ITS days, we'were pretty happy to have people come and check us out and explore, and many of us spent a lot of our time answering questions. We didn't mind that at all; after all, wed once been guests with questions!

I'm pretty much a free-speech guy. I don't have to agree with someone to support their speaking, nor to engage in discussions with them. But some people use speech as a weapon to interfere with free speech by others. What do you do about it?

In my experience, banning them doesn't really work. They just pick up the challenge under a new identity, laughing at you.

What DOES work -- every time in my experience -- is to set things up so only THEY see their posts. Nobody ever responds to them. Nothing has been done to them, that they can see -- it's just a boring experience.

Anyway, that's when and where I first experienced the power of open communities. It's not just the software, it's the people, and the opportunities.

Back around 1976 or so, I created Teach Lisp -- what I believe to be the first open online network learning environment (as opposed to Plato, which back then required specialized hardware, and was only experimentally networked). This packaged a set of Lisp courseware, written by myself and Kent Pittman (later editor of the X3J13 Common Lisp standard), and allowed people to log on and  type :TEACH and be presented with a text-based menu of units, experiment interactively with the various data an programs set up by the course -- and ask for help.

This help could be interactive, if any of us were available; if not, it would fall back to email. It would keep logs of the sessions, so we could see what difficulty they were having. (We had a number of volunteer aides).

Today, those logs or our email discussions occasionally crop up in Google searches (though none seem to be googllable at the moment, they may show up yet again).

I am still in touch with some of the people who used the program. 35 years later. They covered quite a range, from high school students to physicists.

To complete the arc of history, thanks to a Facebook query, I recently received snapshots of the relevant directories, so I should be able to get this running again -- on my phone.

Regarding editors -- back in the day, a lot of us came up with lots of TECO macros to make our text editing better. Eventually, we started pooling our efforts. I'm not sure exactly how it came to pass that Richard Stallman took charge of all this, but the result was ITS Emacs. From that, of course, came GNU Emacs and the GNU project and the whole of what we know as open source software.

Of course, RMS contributed enormously to the creation of ITS Emacs; I believe many of the core ideas that made it so practical as a shared platform were uniquely his. But it was also a platform for a community effort, and was ported to Twenex (PDP-20), and various macro packages for other editors were created to at least give a degree of editing compatibility

I still have Emacs command strokes wired into my fingers.. I have Gnu Emacs running pretty much everywhere (except my phones!).

Back around 1980, I worked with Per Hammarlund, then at the Royal Instutitue of Technology in Stockholm, to add to the Japanese/English dictionary lookup package he created, a reverse morphology capability,so it could take Japanese text, turn it back into a series of candidate root forms, and look up the possible translations. Later, after I met my wife, this became even more useful...

To me, open source has been there my entire adult life.  But I hardly ever think about it that way, unless I'm thinking about whether I can use it in a product and what license is it under?

To me, it's about community and sharing and cooperation, and software is just one product. Even more valuable than the software is the vast sharing of knowledge. If I have a question about some problem -- whether it be with an error message from a piece of software, or a crazy decision by a government bureaucracy, my first step is generally to google it. I don't even consider asking it on a mailing list anymore -- because someone surely has done that.

These days, programing in Java, if it's an error message, as likely as not, my search will turn up the actual code that issued the error message.

I have never subscribed to RMS's dictum that "Software wants to be free". But I do firmly believe, "People want to share", and we're all enriched in the process.

Beyond Googling for Knowledge

One of my goals in writing this blog is to bring to my readers a better sense of how we have gotten to this point in time, with computers and the internet and Google being ubiquitous. So many people are surprised or even disbelieving to learn I've been using the internet since 1972 (called the ARPAnet back then). I think the human journey to this point is fascinating, from the development of the Roman system of roadways to a future we cannot imagine.

I have a host of topics in store, but today I'm going to jump ahead to a piece of breaking news, which fits into that picture somehow. I can't say "this is the future" -- but I do believe, this points to the future in a way that is both clear, and yet familiar, for it refers to capabilities we have long dreamed of for our computer assistants.

Stephen Wolfram, the guy who most successfully brought Symbolic Mathematic Software to the commercial market with Mathematica, and who sponsored MathWorld, is now about to unleash a new effort, which he expects to be even grander.

Wolfram|Alpha Is Coming!

Of course, like any grand idea, it could be a bust -- at least for now. But if a simple idea like Google can change the world, as it surely has, this is surely worth a deeper look.

I've worked on Macsyma (a predecessor and at times perhaps a rival to Mathematica, a version of which is freely available as Maxima and commercially from Symbolics). And I've worked on Cyc -- Doug Lenat's massive effort to encode all of human common-sense reasoning into a knowledge base. This gives me an appreciation for the scope of what he's trying to do.

The opportunities are there, and growing all the time, as more and more knowledge becomes online, and in better and better forms for automated reasoning. Computers have long been able to do things that surprise and surpass humans. But the scope and variety of what humans do, and want help with, always is a challenge. One of the biggest issues with both Cyc and Macsyma was just the huge variety of types of reasoning humans bring to bear, and expect their computers to bring to bear on their behalf.

And the huge amount of knowledge required to be effective for Cyc made knowledge entry a Herculean task. Yet it would appear that Wolfram|Alpha manages to find a methodolgy and compromise that allows for curated knowledge -- separating the wheat from the chaff, without impassible economic barriers to encoding and capturing that knowledge. I am intensely curious about both the choices made in the approach, and the economics of those decisions, and the domains over which it will be effective.

So far, I have yet to see this, and I have a million questions. Rather than speculate further, I refer you to someone who has apparently seen it, Nova Spivack: Wolfram Alpha is Coming -- and It Could be as Important as Google.

For the moment, there's not much there at the WolframAlpha site -- just a couple tantilizing links, including one to request early access. It's expected to launch in May, 2009. 

In the credit-where-credit-is-due department, I came upon this via Matt Marshall's blog entry Wolfram Alpha — it’s like plugging into an electronic brain, which was called to my attention by Don Hopkins on Facebook.

[Addendum: One reason I believe the future lies in this direction (in terms of capabilities) is that my 8-year-old daughter already expects Google to do this, because she's seen simple examples. I myself try it from time to time, knowing it probably won't do anything but a query, but accepting that as a fallback.]

Getting it wrong -- the oldest software bug?

Tom Van Vleck tells a simple story of a centuries-old algorithm gone astray:

In 1800, mathematician Carl Frederich Gauss presented an algorithm for calculating the date of Easter. He got it wrong. Repeatedly. But he fixed it up each time, and it has since been widely published.

In 1971, Tom Van Vleck wrote a calendar program for Multics that used it, figuring that would be the most authoratative way to get it right.

After being right for 180 years, it failed in 10.

Ooops!  Read all about it. Can anyone top this for an old, old bug?