Carl Malamud: Internet Talk Radio, flame of the Internet. 

This is Carl Malamud. Welcome to Geek of the Week. We’re talk­ing today to Brian Carpenter, who’s the group leader for the Communication Systems Group at CERN. CERN is the European lab­o­ra­to­ry for par­ti­cle physics. Welcome to Geek of the Week, Brian. 

Brian Carpenter: Well hello.

Malamud: I guess I’d first like to ask you about CERN’s role in the Internet. CERN is one of the kind of mag­nets. It pro­vides con­nec­tiv­i­ty to much of the world. Why is a physics lab­o­ra­to­ry so influ­en­tial in pro­vid­ing gen­er­al connectivity?

Carpenter: I think it’s because our main goal is to sup­port remote users. Our main users are physi­cists who have to work most of the year in their own uni­ver­si­ty depart­ments. They have to teach their stu­dents. And once in awhile they would like to do some research. And the exper­i­ments have to be done in a cen­tral loca­tion, which in Europe hap­pens to be at CERN, which is near Geneva. So, when these peo­ple are back home, they need access to their data, they need access to the exper­i­ments. If the exper­i­ment is run­ning they want to be able to log in and check things from home. And that’s why it just became a pri­or­i­ty to us to have net­work con­nec­tiv­i­ty. And the best way to get con­nec­tiv­i­ty was to be a hub.

Malamud: And you were a hub at first for HEPnet, the High-Energy Physics Network, which was based on DECnet.

Carpenter: Well his­tor­i­cal­ly I would say that that’s prob­a­bly the first thing. You know, the Italian physics com­mu­ni­ty real­ly start­ed using DECnet in Europe, and then it migrat­ed to CERN, and grew out­wards from CERN. And more or less at the same time we became one of the main nodes in the EARN net­work, which was and still is the European BITNET. So those two bits of con­nec­tiv­i­ty grew up in the ear­ly 80s, I would say. Early to mid-80s. And that’s real­ly how we got into net­work­ing as opposed to a few point-to-point links to exter­nal laboratories.

Malamud: And that net­work­ing in the HEPnet world was DECnet. You were one of the larg­er DECnet Phase IV sites, I believe?

Carpenter: Yes. It was a com­bi­na­tion of DECnet and X25—X dot 25 in America, used as both as a base pro­to­col and as a direct sup­port for remote login. So that was a his­tor­i­cal sit­u­a­tion. And when the BITNET tech­nol­o­gy arrived, we start­ed run­ning RSCS some of the lines as well. So we’ve always been a multi-protocol site, but for a long time DECnet was prob­a­bly the high­est bitrate, if you like.

Malamud: Are you still a large DECnet site? 

Carpenter: We’re still very depen­dent on DECnet because the exper­i­ments that’re now run­ning, these are exper­i­ments were designed in the ear­ly 80s. They com­mit­ted to com­put­ers from Digital Equipment Corporation; this is not a com­mer­cial, it’s a state­ment of fact.

Malamud: Physicists have often used VAXes, VAXes and Fortran.

Carpenter: Absolutely. VAXes and Fortran. And so these peo­ple found DECnet just came in the box with their com­put­er, so they did­n’t have too much choice. And that’s how we got into DECnet. And we’re clear­ly going to stay with DECnet as long as our users say they need it. Because we’re user-driven.

Malamud: Are you run­ning DECnet Phase V or Phave IV?

Carpenter: We’re run­ning both. As you prob­a­bly know, some com­po­nents of of DECnet Phase V are released. And we’re run­ning them. We’re plan­ning to migrate back­bone rout­ing to DECnet Phase V next month—that’s August. And not all the sites are going to migrate. This is a par­tial migra­tion of the back­bone. And you know, we have to do this because this is the cur­rent prod­uct, and you have to fol­low the cur­rent prod­uct. So you know, we think we’re going to be with DECnet for quite a while.

Malamud: Do you also run TCP/IP services?

Carpenter: I would say we do run a few TCP/IP ser­vices. This has grown at incred­i­ble speed. I mean, we start­ed run­ning TCP/IP on the LAN in the 1980s. And at a cer­tain point, it became obvi­ous that the Internet was going to grow in Europe. And this was a sort of spon­ta­neous growth, and it became pret­ty obvi­ous that we were not going to be able to deny our users this ser­vice. So at a cer­tain point we got a col­lec­tive deci­sion in the HEPnet com­mu­ni­ty to start sup­port­ing TCP/IP ser­vices. And hav­ing tak­en that deci­sion, which was a polit­i­cal deci­sion, we then start­ed imple­ment­ing it. Which meant buy­ing routers, get­ting to know the tech­nolo­gies, set­ting up name­servers, and putting IP traf­fic in par­al­lel to DECnet on quite a few of our leased lines, using mul­ti­plex­es, while we got the expe­ri­ence. What’s hap­pened is the growth rate in IP traf­fic has been quite fan­tas­tic, as the growth rate in a num­ber of Unix sys­tems has been fan­tas­tic. I think those two things clear­ly go togeth­er. And now it’s dom­i­nat­ing our traf­fic by a large margin. 

Malamud: In Europe, there has been his­tor­i­cal­ly a strong pres­sure to move towards OSI pro­to­cols. The DECnet Phase V world is a mod­i­fi­ca­tion or imple­men­ta­tion of the OSI pro­to­cols yet you’re also run­ning TCP/IP. Do you feel caught in the mid­dle there? Can you run a multi-protocol net effectively?

Carpenter: Well this is a very embar­rass­ing ques­tion for me because I took over the group in the begin­ning of 1985, and I remem­ber very well in May 1985 I came to a con­fer­ence in Amsterdam, where we are today, and I gave a talk on the OSI net­work­ing pol­i­cy of CERN. And I had to give this talk sev­er­al times because I was new in the job and my boss­es want­ed to know what the new guy was going to do. So I got quite a rep­u­ta­tion for talk­ing about OSI.

Now, you know, it’s very dif­fi­cult when you find egg on your face to wipe it off and still look sen­si­ble. So, Carl is smil­ing because I was wip­ing egg off my face in front of the micro­phone. It’s a real shame we don’t have video as well as audio.

Malamud: We’re work­ing on that.

Carpenter: Okay. That’s good.

So you know, by ​’88 or so it became obvi­ous that OSI was just not com­ing fast enough, if it was com­ing at all. But, there was a lot of pol­i­tics around that in Europe. There was some peo­ple who very hon­est­ly believed that OSI was going to be the solu­tion to all our prob­lems, and who did­n’t mod­i­fy that view as quick­ly as I did in the light of events, with the way that the Internet Protocol suite became avail­able on pret­ty well every­thing you could buy and OSI was still theory. 

So we went through a peri­od which was polit­i­cal­ly quite del­i­cate, I would say, of per­suad­ing our man­age­ment that actu­al­ly we’d been wrong by telling them OSI was the future; telling them IP was the future, at least for a while. And we can talk about how long that is. And so this is a sit­u­a­tion we got into it. 

Now, we’ve always con­tin­ued some effort on OSI at CERN. We’ve par­tic­i­pat­ed in the RARE CLNS pilot. That’s the pilot project for using the OSI con­nec­tion­less lev­el 3. And in fact that’s still run­ning; it’s even sup­port­ing part of the demon­stra­tions here at the IETF. And you know, OSI has not gone away, but it has­n’t arrived either.

Malamud: It is going to arrive?

Carpenter: I would like to quote an offi­cial of the European Commission who showed a trans­paren­cy once which said, ​“OSI has a con­stant prop­er­ty of being just around the cor­ner.” And I think that’s absolute­ly true. You know, we’ve just been wait­ing too long. And I think this is why peo­ple’s patience in Europe snapped and every­body, with very few excep­tions at a cer­tain point real­ized that to deliv­er ser­vices to the users, there’s no alter­na­tive to the Internet pro­to­col suite. 

Malamud: Is OSI going to hap­pen, or has it lost its win­dow of opportunity?

Carpenter: I think that’s still a very inter­est­ing ques­tion, actu­al­ly. I think it’s still a rel­e­vant ques­tion. My per­son­al opin­ion is that OSI as the grand plan has lost its win­dow of oppor­tu­ni­ty, that we will see com­po­nents of OSI which will be useful. 

Malamud: Which of the com­po­nents do you have in mind?

Carpenter: Well, I think there is still some chance that X.500 will be use­ful, the direc­tor ser­vice, because I think the Internet DNS has lim­i­ta­tions. I’ll prob­a­bly be shot down if any­body hears me say­ing this, but I think…personally I think one of the options which is viable for the next gen­er­a­tion of IP is the TUBA option, which uses com­po­nents of OSI. I have to say that care­ful­ly because that’s not the only option, and it may not be the one that’s select­ed. And I think there are lit­tle com­po­nents like that, you know. ASN.1, the Abstract Syntax Notation from OSI is an exten­sive use in the Internet.

Malamud: Are these pieces going to be inte­grat­ed into the Internet pro­to­col suite or will they be sep­a­rate? Is this ships in the night?

Carpenter: I think it’s going to be more than ships in the night. I think there’s going to be a degree of inte­gra­tion in the sense that the Internet peo­ple have the habit of tak­ing what’s good from the oth­er stan­dard­iza­tion bod­ies. I mean they take what’s good for the ITU stan­dard­iza­tion and what’s good from the IEEE and the ANSI. And they’ll take what’s good from the ISO process. But that’s not to say that they’re tak­ing OSI, that they’re buy­ing the whole bun­dle. It means their tak­ing the good com­po­nents. And you know, there are peo­ple who have very emo­tion­al reac­tions on this top­ic and say ISO is bad, you know, and don’t want to hear about it. But I think that’s not going to pre­vail. I think what’s going to pre­vail is a prag­mat­ic approach, tak­ing what’s good, and ignor­ing what’s bad. 

Malamud: There’s been a move by the Internet Society to estab­lish for­mal liai­son with ISO and with the ITU, and to do that as a for­mal liai­son in which rep­re­sen­ta­tives are sent back and forth and papers are exchanged. Do you think that’s a healthy move?

Carpenter: I think that depends entire­ly on the deal. It depends on the deal because if the deal is one that con­strains the IETF process…that’s to say the way the Internet Engineering Task Force gets to rough con­sen­sus and run­ning code, which is the cri­te­ri­on for some­thing being accepted—it’s a very prag­mat­ic approach to stan­dard­iza­tion. If that process is con­strained by liai­son with the oth­er orga­ni­za­tions then I think it’s a bad thing, because that is the real out­stand­ing qual­i­ty of the Internet stan­dard­iza­tion process. If the effect of liai­son is actu­al­ly to feed the oth­er way such that our good ideas get tak­en up by the offi­cial stan­dard­iza­tion bod­ies, I don’t see how it can be harm­ful, so long as we keep change con­trol of what we’re doing.

Malamud: Do you think an ISO stamp on the Internet stan­dards is a use­ful thing in some countries?

Carpenter: I think to the extent that you can not real­ly expect politi­cians to ful­ly under­stand tech­ni­cal issues, I’m afraid it may be that in some coun­tries the rub­ber stamp would make a dif­fer­ence. It would change the per­cep­tion of the peo­ple who are pro­vid­ing the mon­ey, and in the end we need the mon­ey to do things. But I don’t think indus­try’s going to care about that very much. I think it’s only government-funded activ­i­ties where this is real­ly a consideration.

Malamud: Now the International Organization for Standardization, ISO, is a pri­vate body. As is the Internet Society. Could the Internet Society pro­vide that stamp by itself or does it need that ISO liaison?

Carpenter: I think that’s a pre­sen­ta­tion issue. I think the ISOC is too young, it has­n’t got an image at a polit­i­cal lev­el yet. ISO has got an image at the polit­i­cal lev­el, and the bod­ies that are mem­bers of ISO are things like the nation­al stan­dards orga­ni­za­tions of the var­i­ous coun­tries. So, although ISO is not a gov­ern­ment orga­ni­za­tion, it has mem­bers that are gov­ern­ment orga­ni­za­tions. So I think there is an issue there. I think the ISO rub­ber stamp is prob­a­bly still more pow­er­ful, polit­i­cal­ly, than the ISOC rub­ber stamp; that the ISOC process, the IETF stan­dard­iza­tion process, is more pow­er­ful tech­ni­cal­ly than the ISO process. So that’s a dilemma.

Malamud: For a long time, CERN was the back­bone of Europe, the de fac­to back­bone. There were a few oth­er sites; there was University College London, there were some oth­ers that were pro­vid­ing con­nec­tiv­i­ty. But over the last cou­ple years, some­thing called the EBONE has come togeth­er. How did that hap­pened at the cen­ter shift­ed from CERN over to this broad­er consortium?

Carpenter: Well, I’m not entire­ly sure we were ever the only cen­ter in Europe. I think that’s quite unfair on some of our col­leagues in oth­er coun­tries. For exam­ple since we’re in Amsterdam I think it’s fair to say that the Amsterdam people—how to say, the com­plex of uni­ver­si­ty depart­ments here have been very much in the fore­front of net­work­ing. And as I said, so have our Italian col­leagues. And of course so has the JANET net­work in the UK. But yeah, we were clear­ly a major cen­ter in terms of where the band­width ter­mi­nat­ed. And this is less true now. And the rea­son that is, I think that it clear­ly became obvi­ous to many peo­ple that it’s not actu­al­ly the job of a physics lab to pro­vide general-purpose con­nec­tiv­i­ty. You know, we’re not fund­ed to do that, and we nev­er spent any of our physics mon­ey on doing that. We just…offered floor space, essen­tial­ly, to oth­er peo­ple. And it was a sort of evo­lu­tion­ary process. And when peo­ple began to think about this as a man­age­r­i­al issue, they real­ized this was not real­ly the right way to do it. 

It’s my job to make sure our users have the best pos­si­ble con­nec­tiv­i­ty, and one way of doing that is hav­ing all the net­work links in Europe ter­mi­nate in our com­put­er cen­ter, but that is real­ly not the right topol­o­gy. It’s not well engi­neered. And it’s not right polit­i­cal­ly because we’re not there for that, we’re there to do physics.

So this is a very per­son­al opin­ion. I think it’s a nat­ur­al devel­op­ment that this general-purpose infra­struc­ture has devel­oped so much in Europe in the last cou­ple of years, such that the general-purpose users are no longer depen­dent on a dis­ci­pli­nary network. 

Malamud: What changed two years ago that would allow some­thing like the EBONE to be born? Or why was­n’t it born four years ago?

Carpenter: Well, as you know of course, these things always hap­pen in Amsterdam because I hap­pen to remem­ber that you were present at the meet­ing in the Amsterdam zoo—this is true, lis­ten­ers— in the Amsterdam zoo, which was the for­ma­tive meet­ing for the EBONE net­work. I think what changed actu­al­ly was that around that time, peo­ple got fed up with wait­ing for the so-called offi­cial solu­tion to appear. And EBONE was a con­sor­tium of peo­ple with com­mon inter­ests. It was not an offi­cial pro­pos­al. There is an offi­cial net­work now also in Europe, also oper­a­tional or near­ly oper­a­tional called EuropaNET, and it’s going to be very inter­est­ing to see over the next twelve months how much of the traf­fic migrates from the EBONE infra­struc­ture to the EuropaNET infrastructure.

Malamud: Well let’s com­pare those two net­works. How do the net­works dif­fer for exam­ple in who their mem­bers are? Who joins EBONE, who joins EuropaNET?

Carpenter: Well EBONE is a club, real­ly. And like any club you know, you pay your sub­scrip­tion and then you get the same rights as all the oth­er mem­bers. So any­body can join EBONE, as far as I know. I can’t real­ly speak for EBONE, you need to inter­view Kees Neggers for that. But my under­stand­ing is that there is no accept­able use pol­i­cy which would cause any­body any prob­lems on EBONE. If you’re pre­pared to pay for your access point, then you can get access. 

EuropaNET is spon­sored ini­tial­ly by the European Commission. It’s been cre­at­ed in the con­text of RARE, which is the asso­ci­a­tion of the European research net­works, and specif­i­cal­ly of the nation­al networks—that’s to say the equiv­a­lents in each coun­try of the NSFNET in the States—and it’s offered by a com­mer­cial ser­vice provider rather than by a con­sor­tium of indi­vid­ual sites.

But again, as far as I know any­body can nego­ti­ate for a EuropaNET access if they want one. And at some point it will be ful­ly fund­ed by user pay­ments. Although I think there is some start­up fund­ing, actu­al­ly I’m not too clear on the start­up fund­ing for EuropaNET.

Malamud: Now they’re both tran­sit net­works. EBONE is based on two megabit leased lines. What is EuropaNET based on?

Carpenter: As far as I know it’s based on a hub struc­ture, with a cou­ple of hubs. I’m not the real expert on this so I can’t tell you exact­ly how many hubs and lines out to the access points from those hubs. Whereas EBONE is actu­al­ly a ring which con­sti­tutes the back­bone with lines into that ring from exter­nal back­bone systems.

Malamud: EuropaNET as I under­stand it is based on two megabit X.25?

Carpenter: Well EuropaNET is sup­posed to be migrat­ing to a multi-protocol, with gen­uine native IP as well as X.25. So I think at the moment it’s in a bit of an inter­me­di­ate state but it’s sup­posed to come up lat­er this year, if I have the dates right, with native IP. So that’s…you know, that’s progress in terms of tech­ni­cal flexibility.

Malamud: Can Europe have two back­bones? Is that a good thing to have?

Carpenter: Well we’ve got at least two because there are oth­er net­works around like some real com­mer­cial ser­vice providers of IP net­work­ing. So…I don’t see why not. I mean, we’re sup­posed to live in a cap­i­tal­ist soci­ety in Western Europe these days so, why not have two? Why not let the user sites choose?

Malamud: There’s a tremen­dous amount of activ­i­ty and noise in Europe now about ATM. Do you see that as a major ser­vice in the next year or two?

Carpenter: Certainly not in the next year or two. I think it’s going to take a long time to see ATM deployed in a gen­er­al sense in Europe. What’s hap­pen­ing at the moment is a large num­ber of the oper­a­tors, I think it’s almost twen­ty now, signed a mem­o­ran­dum of under­stand­ing about the deploy­ment of ATM on a pilot basis across Europe. That’s a very inter­est­ing devel­op­ment, but one thing I have learned in deal­ing with the PTTs is if some­thing is called an experiment…they might just give it to you free. If it’s called a pilot, you’re going to pay for it. Not at full tar­iff, but you’re going to pay for it. So you know, even at this pilot phase there is a prob­a­bil­i­ty that peo­ple will have to pay to get inter­na­tion­al ATM con­nec­tiv­i­ty. It’s not a cer­tain­ty and it will vary from one coun­try to another. 

And I’ve said this before in a talk in London last year, I think the crit­i­cal issue for ATM is not tech­nol­o­gy. We’ve got tech­nol­o­gy com­ing out of our years. The crit­i­cal issue is will the price per­for­mance of ATM be inter­est­ing to the big pay­ing cus­tomers. Will the banks, the insur­ance com­pa­nies, the air­lines, the peo­ple try to do remote CAD like the car com­pa­nies, will all these peo­ple want to pay for ATM? Will it be cheap enough to inter­est them. It’s clear the aca­d­e­m­ic mar­ket is not going to sup­port ATM.

Malamud: Many peo­ple have said that ATM pro­vides general-purpose con­nec­tiv­i­ty. It pro­vides that cloud into the world that lets you talk to any­body; that does­n’t mean they have to talk back. But it pro­vides that same general-purpose con­nec­tiv­i­ty that the Internet does. Does ATM make the Internet irrelevant?

Carpenter: Absolutely not. I mean, I think the ATM con­nec­tiv­i­ty, if it real­ly appears as the uni­ver­sal con­nec­tiv­i­ty, will be a very inter­est­ing way to car­ry Internet traf­fic. I think it’s just anoth­er tech­nol­o­gy. I think it’s not the uni­ver­sal solu­tion. I don’t believe I’m going to have an ATM con­nec­tion to my toast­er. I think you know, there’s still going to be sub­nets, if you want to get tech­ni­cal, which are non-ATM for twen­ty, thir­ty, fifty years. I mean, as long as you like. And I don’t believe I’m going to get ATM to North Africa or Eastern Asia that quick­ly. And I think that’s very impor­tant. The Internet is real­ly a glob­al network.

Malamud: You men­tioned your toast­er. How’re we going to con­nect your toast­er to the Internet? And are we?

Carpenter: Well. I mean, let me tell you a sto­ry. A cou­ple of weeks ago there was the storm of the cen­tu­ry in the par­tic­u­lar area where I live, just local­ized. And you know, the vol­un­teer fire­men were out all night pump­ing our peo­ples cel­lars. And my neigh­bors hap­pened to be out at a grad­u­a­tion din­ner for their son. And they got home at mid­night to find two feet of water—about six­ty cen­time­ters since we’re in Europe—of water in their cel­lar. And they were a bit annoyed about this because they were mov­ing out two weeks lat­er and had to hand the house back to the landlord. 

Now, it occurs to me that it would have been per­haps quite con­ve­nient if their beep­er had gone off when they were down at this grad­u­a­tion din­ner say­ing you know, ​“There’s a prob­lem in your cel­lar.” And they could have done some­thing about it and maybe got the fire­men around a bit soon­er and pumped it our before all the rugs were ruined. So I think instru­ment­ing your house is not com­plete­ly friv­o­lous. If we are in an afflu­ent soci­ety you know, this is the sort of thing we might want to do. I’m not too sure about the toast­er; that’s…become an Internet joke. But you know, instru­ment­ing my house, yes why not.

We have on the table at least four pro­pos­als for the next gen­er­a­tion of IP. They will have their tech­ni­cal fea­tures. I’m sure they will all work with­in their own terms of ref­er­ence. I’m not sure yet that any of them actu­al­ly solves this prob­lem of instru­ment­ing every cel­lar in every house in every coun­try. You know, it’s not obvi­ous to me that any of these solu­tions will real­ly go to that lev­el. Maybe it does­n’t mat­ter. Maybe that’s the wrong approach to solv­ing that prob­lem; I don’t know.

Malamud: There’s two very large sets of activ­i­ty in the Internet—there’s many oth­ers, but one is focus­ing on the ques­tion of the next lev­el of the net­work lay­er. There’s anoth­er group of peo­ple, includ­ing Tim Berners-Lee at CERN that are look­ing at much high­er lev­el issues of how to orga­nize infor­ma­tion, and find infor­ma­tion. Are these two prob­lems related?

Carpenter: Well, in a strict sense, in a math­e­mat­i­cal sense, of course they’re not relat­ed because they’re occur­ring at dif­fer­ent lay­ers of the OSI mod­el. And the OSI mod­el is use­ful for think­ing about these things. So you can argue that they’re not relat­ed. But I’m afraid they are relat­ed because of course if you start build­ing things like the World Wide Web, this pre­sumes that the net­work is ful­ly con­nect­ed. And when I find a point­er in a hyper­text doc­u­ment that points to anoth­er doc­u­ment that is in Australia, the World Wide Web mod­el pre­sumes that I can go look at that doc­u­ment and get it in a sec­ond or two of real time. And that’s a very big pre­sump­tion about the under­ly­ing infra­struc­ture. Because we’re all used to the Internet being par­tial­ly bro­ken. I mean, every day the Internet is par­tial­ly bro­ken. There are con­nec­tiv­i­ty holes for one rea­son or anoth­er. And if you’re putting higher-level appli­ca­tions on the Internet, and these are very impor­tant appli­ca­tions, that assume the net­work is not bro­ken, then you’re going to have very frus­trat­ed users—much more than today.

If I do tel­net to a machine in Australia and it does­n’t work, then I try again lat­er. If I’m in the mid­dle of a World Wide Web ses­sion and I hap­pen to click on a but­ton that fol­lows a point­er to Australia and it does­n’t work and I get very annoyed and I keep click­ing on that but­ton until the mouse is worn out.

Malamud: So we need to be think­ing in terms of robust­ness and the pro­duc­tion qual­i­ty of the Internet. 

Carpenter: I think that— That’s right. I think this new gen­er­a­tion of appli­ca­tions, and it goes for any dis­trib­uted com­put­ing appli­ca­tion, assume a robust­ness of the under­ly­ing lay­ers that is not nec­es­sar­i­ly the case today. 

Malamud: So should the PTTs then take over those under­ly­ing layers?

Carpenter: Uh, am I allowed to say ​“no com­ment” in this interview?

Malamud: You sure are. Thank you very much, Brian. This has been Brian Carpenter. He’s at CERN. He’s the group leader for the Communication Systems Group which runs the CERN net­work. This has been Geek of the Week.

This is Internet Talk Radio, flame of the Internet. You’ve been lis­ten­ing to Geek of the Week. You may copy this pro­gram to any medi­um, and change the encod­ing, but may not alter the data or sell the con­tents. To pur­chase an audio cas­sette of this pro­gram, send mail to radio@​ora.​com.

Support for Geek of the Week comes from Sun Microsystems. Sun, the net­work is the com­put­er. Support for Geek of the Week also comes from O’Reilly & Associates, pub­lish­ers of the Global Network Navigator, your online hyper­text mag­a­zine. For more infor­ma­tion, send mail to info@​gnn.​com. Network con­nec­tiv­i­ty for the Internet Multicasting Service is pro­vid­ed by MFS DataNet and by UUNET Technologies.

Executive Producer for Geek of the Week is Martin Lucas. Production Manager is James Roland. Rick Dunbar and Curtis Generous are the sysad­mins. This is Carl Malamud for the Internet Multicasting Service, town crier to the glob­al village.