At 09:05 AM 2/21/01, Alan Kay wrote:
>Yes, it is still somewhat tough for individuals to get both papers 
>and videos, that is just what university (especially department) 
>libraries are for. I bet CMU's Computer Science library has all (it 
>should, since lots was done there and the tradition there was to 
>share). The Computer History Museum has quite a bit of everything, 
>and this is where university department libraries should be directed 
>if they haven't figured out how to get this stuff.  There are very 
>few other fields that are trying to be serious about what they do 
>that are so indifferent to keeping landmark achievements from the 
>past in view.
>
>So my claim here is that via what libraries are for, it should not be 
>difficult to find out what has existed over the last 40 years, nor 
>should it be difficult to get it. Compare this with other fields, the 
>relatively small amount of stuff done since 1960, the tremendous ease 
>of copying using xerox machines, video tapes, digital media and 
>communication. If you can't find it readily then there is something 
>wrong with the larger university processes re computer "science" (and 
>I suspect there may just be).

My experience with CMU's library (both the CS library and other parts of
it) is that it works better for anything that's been formally published.
So we've got a copy of Ivan Sutherland's thesis, and even a copy of a video
called "Doing with images makes symbols" that some guy with a name
remarkably to yours did.  ;)

Unfortunately, the library is miserable at providing copies of videos that
weren't distributed by a publisher.  In fact, the library is pretty
miserable about providing videos in general.  Consider the last 15-20 years
of SIGGRAPH.  Every year the conference produces both a written proceedings
and a video compendium.  And what do libraries do?  They purchase the
written proceedings but not the video.  There might be a problem in
perception here: the written record is often perceived as the "formal"
record, while the video proceedings is "bonus" footage.  It's a little
added extra, but it's not necessary.

I suspect you could put part of the blame on the limited budgets for
libraries.  If the video proceedings costs extra, and it's supplemental
info only, why spend the cash on it?  But you can also put part of the
blame on both the publication process and the researchers themselves.  When
I defend my thesis I'm going to show videos that will probably never be
published. The problem is that the computer science department has a formal
process for publishing your written thesis, but no process at all for
preserving the video tapes you show at your defense.

This informal video record isn't anything new either. Most of the videos
I've seen from the past (for example, some of the early work on VR done at
NASA Ames) are a completely informal record: you have to know someone who
has a copy to see them at all. And the folks at NASA Ames did publish a
written version of their work, but it's nowhere near as comprehensive as
the video record. Plus it doesn't even mention the video record. So here's
where I blame researchers as well. We've built a tradition of informally
sharing videos of the work we do, and we aren't working to change it.
Informally sharing videos might have worked fine when the field was small
and new, but it doesn't scale at all. We've now got a written record of the
past that's often disconnected from the video record.  If I'm a new VR
research and I hear about this VIEW system that the folks at NASA Ames did
and I try to find out about their work I find the written record because
that's what's publicly available.  However, the written record says nothing
about, for example, how the VIEW system implemented portals into other
spaces. For that you have to see the video tape.  But how do you find out
about the video tape?  There's no record of it that I've ever found.  Even
the literature that the NASA Ames team produced says nothing about it. To
know about the video tape you have to know someone who's seen the video
tape. If we want future researchers to be aware of our work we need to
provide a better method of discovering it than word of mouth. Research is
one part inspiration and one part promotion.

Or, as an information sciences friend of mine would put it, you can have
all the information in the world available to you, but without an index
you're never going to find anything.

>I also suspect that a very large percentage of kids going to college 
>these days concentrating on computers are much more interested in 
>vocational training for eventual jobs than in learning their field 

Well, yes, there's that too.  =)

Jeff