Speaker: Paul Ginsparg (LANL)

Panel Discussion: Current Thinking Panel

I only have one or two sentences and then I'm just going to do a brief demo. In answer to Michael (Peskin)'s question about submitting software of external pointers, needless to say, it's already trivial. All I wanted to comment on what's been said so far today (and this is what I tried also to say last night is what we're really seeing exposed is that formerly we were beholden to the conventional publication companies for the production and distribution, and they've adopted the tedious mechanics of it as their primary focus. But now that we have alternate means of doing everything, we should really be concentrating on the kind of intellectual value-added that Andy (Cohen) was talking about -- I'm hoping that now that we're freed from the tyranny of worrying about all the trivial mechanics, we can really start worrying about the more important intellectual aspects.

Anyway since some people probably haven't seen the WorldWideWeb interface to some of these systems under discussion, I thought I'd just click on a few pages on this little laptop set up here. OK, so it's a little hard to see (projected on the screen). This is to demo, "I'm not only a provider of these systems, I'm also a user."

So here are some of the archives and starting page for xxx.lanl.gov (click on xxx). This isn't complete yet, but you can see the first one, hep-th = high energy physics, theory, that I mentioned started in August of 1991. Earlier someone mentioned hep-ex (= high energy physics, experiment), which just recently started, as you see in April of 1994. Since the issue arose, we can go in and have a look at what the activity is so far this month. (click on xxx form interface .) So this is the so-called "form interface" for the archives, and we click on these menu buttons, then pull down over here to hep-ex, and have a look at this month's listings (click on Oct '94 hep-ex listings). We can see title listings for 9410, that's October of 1994, and so far this month there are ten papers. That's not large by the standards of hep-th or hep-ph, but for the experimental archive it sounds right so far.

If we want to look at the abstract for some paper we click on abs over here (click on hep-ex/9410010) so you see you get all the information. And here's a good example of author-supplied external URL -- this was somebody who submitted a paper, said "OK I've also got my own postscript," and included an http pointer to it after the title, which the interface automatically converts into a clickable hyperlink to the remote site. So this person could have equally well put "the computer program that generated this is available at ..." and provided an http or ftp URL, which the interface here would automatically convert to a network hypertext link, and there you already have automatically the distributed system mentioned earlier. In addition, the other links on the abstract page are to "source" or "postscript"; the source is coming from the archive here and the postscript is what's being posted by the people at slac. Also "refers to" and "is cited by" on this page are links to the databases provided by slac-spires, that provide hyperlinks back to here.

Now let me go back to the form page for a moment (click on xxx form interface), just to give a search example. If you want to search the past year, we can, for example, go into hep-ph, and find everybody's favorite example (click on find peskin 94): And so we see Michael's papers for the past year --- He's got three in here: from March, May, and August. So we look at the abstract from this one of May '94 (click on hep-ph/9405255), and it's a "broad brush introduction to the theory of spin in quantum field theory," a review article of some sort. And that's one of the criteria that Andy (Cohen) mentioned: We'd like to be able to select for review or pedagogic papers, not knowing in advance that Peskin might have written one. It would be so useful to have a means of finding such papers directly if we're interested in reviews on some topic; And not only in fields which we know, but for example I'd like to go into a biophysics database and find things that are useful pedagogic reviews as determined by members of that community, rather than having to search blindly (as is currently the case).

Two other things I wanted to mention briefly (click on blurb). Here I've posted for anyone interested some background blurb. And there's a photo of the machine itself sitting in a corner under the table in my office (click on exciting action photo), and I've got also posted here usage statistics (click on usagr). Here you see the usage statistics for the web interface on this machine, since January '94. (This is historically the second web interface -- the first one was set up by Joe Carlson, also at the lab, and then I set up this one directly on the machine xxx). You can see usage increases quickly with time. This starts at January of this year, so these are months of this year up to October. And a striking aspect of this graph you see is the seven day periodicity, which is evidence that many physicists still do not have ready network access on weekends. But the overall trend is evidently monotonic increase in activity. Pat (Kreitz) mentioned people going on vacation in July but that effect is represented here only as a sort of flat section here over the summer, then September started booming again, and now there are on the order of 11,000 connects per day. As far as the distribution in time on the server, as you see the peak here is between 8:00 and 9:00 in the morning; that's late morning on the east coast, late afternoon in Europe, so that's when you get maximal contributions and requests from all over. And here you see accumulated by weekdays that the weekly access is curiously symmetric about Wednesday. Another useful observation is to note where all these requests are coming from. If you look at the hosts here, you find requests from Estonia, Egypt, Turkey, Iceland, Singapore, Malaysia, Russia, ...; e.g. here from Slovenia 286 requests. So, it's simply misguided when people claim the system is too progressive for the third world. Perhaps true to a certain extent, but many of them are already much further ahead than most people realize (and perhaps much further ahead than many people here), and many report that the system has already been a godsend to them.

One last thing, well still two other things. Much of what I said last night is contained in a published article in Computers and Physics (Vol. 8, No. 4, p. 390, Jul/Aug 1994), and you can also get it directly on-line from here in compressed postscript (click on blurb.ps.Z) and here are the attached figures (click on figures). And continuing through here, a final thing I mentioned (click on pg), first that's me, and then here (click on lh94) are the summer school proceedings. This is a good future model for all conference proceedings, summer schools, and the like. Here is even a book cover (click on lectures, then on cover). Things are still coming in, but here are the lectures so far, and the "still awaiting" is one way as an editor to encourage people to get their contributions in. And as soon as they're in, these titles just become on-line links to the source material on the database. There are also photos (click on photos) scanned in color (another advantage of the electronic version). And all this is an example of an editorial overlay, where these high quality papers are preselected in this case by our having selected all of the lecturers in the first place. But anyone else can compose an overlay to any other subset of selected material from the database, so we have as well a model for refereed highlights selected from a raw database -- the likely wave of the future.