devine@cookie.dec.com (Bob Devine) (02/09/89)
I just got home from the San Diego conference (and was greeted by below-zero temps and lotsa snow). First I want to thank the organisers for a well-run conference. Handling ~2000 people and dozens of sessions is not an easy job. The only complaints I have are that speakers were hard to locate after they finished their session (perhaps a room devoted to after-session questions?) and the mid-morning danish always dissappeared too soon ;-) Other suggestions are: 1. Multiple tracks for the main session. I was as the first part of the discussion of this issue at the Board of Directors BOF, but I couldn't stay around long enough to give my opinion. It seems silly that there are multiple, parallel tutorials and multiple BOFs but only a single thread for the main session. I would prefer 2-hour blocks where there are 3 speakers of 1/2 hour per block. The scheduling of attendees is by self-selection at the breaks. There could be system administrator thread, a new features thread, a potpourri thread, etc. 2. Use the conference as more than a series of meetings. Gathering 2000 Unix devotees in one spot and then not making use of that concentrated talent seems wasteful. - Get a "pulse of the crowd" by doing a poll on what is needed in computing. Or a poll on what would be a really neat application to have. Any information gathered is valuable for its own sake and can be used as a news item. The American Bar Assoc, for example, gets national attention for its conference when it releases opinion surveys. - Allow people to bring along and demo their work in progress. A description of a user interface is boring without the actual thing to see/play with. Granted, applications that run on a Cray could be difficult to demo... - Use the conference as a way of telling people about Unix. What is in the weekly computer newpapers is a lot of vendor press releases. Usenix could be publicized by press releases that highlight interesting papers. 3. Add more features. - Add some contests. Something like having a computer trivia quiz as described in this month's ACM magazine would be fun. - Have an equipment demo room. A computer conference without computer is sort of like a automobile show without cars. - More panel discussions. There are many areas where there is One True Way. Opinions matter! Put several knowledgable folks together to see if heat or light can be generated. 4. Open up the conferences to more than Unix. This might be a heretical statement. But, many sites have quite a mixture of systems. Topics could be on how to manage mixed-computer sites. Or on how run different OSs on the same network. The point is that there can be a fruitful exchange of information for this area. If Usenix doesn't do it, who will? Bob Devine
nts0302@dsacg3.UUCP (Bob Fisher) (02/09/89)
From article <8902090223.AA01916@decwrl.dec.com>, by devine@cookie.dec.com (Bob Devine): > > - Get a "pulse of the crowd" by doing a poll on what > is needed in computing. Or a poll on what would be a > really neat application to have. Any information gathered > is valuable for its own sake and can be used as a news item. > The American Bar Assoc, for example, gets national attention > for its conference when it releases opinion surveys. Sounds good, but I wonder which pulse we really need to take - the end user, management or conference attendees (who seem to be mostly hackers). Perhaps a questionaire sent out in advance of the conference to be completed by management (maybe after a poll of end users). Presumably, managers (and dedicated individuals) who are willing to put up the money to attend the conference are somewhat enlightened and would provide good input to a questionaire. > - Use the conference as a way of telling people about Unix. > What is in the weekly computer newpapers is a lot of vendor > press releases. Usenix could be publicized by press releases > that highlight interesting papers. Hear! Hear! Shout it out before the conference begins about what is planned and again after the conference is over. Announce the results of the poll. Perhaps someone will volunteer to write a regular news article for the papers/magazines. If it were provided as a *free* press release from USENIX, it might get published more easily and possibly by more than one publisher. > - Have an equipment demo room. A computer conference without > computer is sort of like a automobile show without cars. I have reservations about this. Setting up a demo room is ticklish. How much space is needed? How do we get the vendors to pay for it so that it doesn't become an extra expense for USENIX. It may cost vendors more to rent space in a central area than to rent a room of their own. A central demo area needs open space to provide traffic flow. The vendors may prefer to rent hotel rooms and show their stuff privately with fewer distractions. Of course they would have move visitors in a concentrated area. > 4. Open up the conferences to more than Unix. This might be a > heretical statement. But, many sites have quite a mixture of > systems. Topics could be on how to manage mixed-computer sites. > Or on how run different OSs on the same network. The point is > that there can be a fruitful exchange of information for this > area. If Usenix doesn't do it, who will? PCs as well as mainframes. This was my first USENIX. I understand that the summer conferences are somewhat in the nature of an exposition, so some of what I said about demo space may be wrong or already considered. Anyhow, THANKS FOR A GREAT CONFERENCE IN SAN DIEGO! -- Bob Fisher (osu-cis!dsacg1!bfisher) 614-238-9071 (Autovon 850-9071) From the Internet: bfisher%dsacg1.uucp@daitc.arpa US Defense Logistics Agency Systems Automation Center DSAC-TSX, Box 1605, Columbus, OH 43216-5002
peter@usenix.UUCP (Peter H. Salus) (02/10/89)
In article <8902090223.AA01916@decwrl.dec.com>, devine@cookie.dec.com (Bob Devine) writes: > > 1. Multiple tracks for the main session. > Over the several years that I have been Executive Director, the various Program Chairs have attempted both Parallel and Single Track Sessions. In every case the option selected was criticized by several attendees. My conclusion a year or so ago was to leave the Chairs alone as the critics would be there no matter what one did. We even had critics of the procedure when we had the WiPs run in parallel to main sessions. Clearly, anything the Association and the Program Chairs do will make some subset unhappy. We even get flak on which tutorials are simultaneous: with 23 tutorials over two days, there's no way to avoid conflict. > 2. Use the conference as more than a series of meetings. > - Get a "pulse of the crowd" by doing a poll on what > is needed in computing. Why not? We ran a member survey in 1986 and will be doing another in March. The response forms are carefully read: the increased number of Workshops, the WiPs, and COMPUTING SYSTEMS all developed from member responses. But as I pointed out in both SF and SD, the Association runs on volunteers -- if Bob Devine wants to run a survey on "what is needed," I think the Association would be happy about it. > - Allow people to bring along and demo their work in progress. I have seen folks do this at Graphics Workshops; we don't prevent anyone from doing this in any way. (One of the papers in the first session in SD had a demo on a Mac SE.) > - Use the conference as a way of telling people about Unix. > We do so. The Association has a Press Representative and a Press Room. It issues releases. Exactly what the weeklies pick up on cannot be controlled; the monthlies like UNIX/ World and UNIX Review; papers like UNIX Today!; and newsletters like Unique and UNIX in the Office seem to do well by us. > 3. Add more features. > - Add some contests. Among the Board members, both Rob Kolstad and John Quarterman have been discussing this. Write kolstad@usenix.org. > - Have an equipment demo room. In general, USENIX has a vendor exhibit at each summer technical conference. It has none at its Winter Conferences because these are generally in the same place as UniForum. This year UniForum is in March, which is too late for our members, so there was no simultaneity. Next January (1990), both will be in DC [for the traditional blizzard]; in 1991, both will be in Dallas; in 1992, both will be in SF; but in 1993, we will be separate again, as UniForum will be in March in DC [USENIX has not yet decided on its location]. Anyway, I was a the American Physical Society meetings last month, and no one bemoaned the absence of a cyclotron, a linear accelerator, or a photospectrometer. > - More panel discussions. > Again: these are at the option of the Program Committees; if you want something, propose something. Just posting a long piece on the net doesn't help. Yesterday was the deadline for papers for Baltimore: make a proposal for a panel discussion! > 4. Open up the conferences to more than Unix. > They are open to "more than UNIX." Surely no one thinks that C++ or Eiffel or Mach etc. is identical to UNIX. Read the blurbs: the Association is serious about advanced systems and languages. While the Baltimore CFP in ;login: talks about "work related to or based on the UNIX operating system," a quick scan of tutorials and papers over the past three conferences shows how "liberally" the committees have seen that. (If you look at the past three keynoters, it's hard to see either Steve Jobs or Adele Goldberg as a devout UNIX person.) Peter H. Salus Executive Director
mark@drd.UUCP (Mark Lawrence) (02/10/89)
devine@cookie.dec.com (Bob Devine) wrote: [some suggestions] } 1. Multiple tracks for the main session.... I feel for the Board. Damned if ya do and damned if ya don't. I *like* the single track main sessions. I think that the Board has responded well to past criticism about multi-tracked main sessions (which were tried). I go with the faction that says I don't want to have to make choices among two talks that happen to be presented simultaneously. } 2. Use the conference as more than a series of meetings. ... I like this idea if the opinion monitored is something of substance. } 3. Add more features. ... I think that this happens, but more or less as people WANT to devote energy and effort to make it happen. The Board has demonstrated openness to such activities in the past and my guess is that they probably will continue to do so *IF* people volunteer to do the work. } 4. Open up the conferences to more than Unix. This might be a } heretical statement. *NO* and Yes, it is. I joined Usenix for a reason. It's called Usenix *for* a reason. "Keep them ferin' operatin' systems *out* of my association" (children and grandchildren of UNIX excepted -- e.g. Mach) } Bob Devine Mark Lawrence
reggie@pdn.nm.paradyne.com (George W. Leach) (02/10/89)
In article <8902090223.AA01916@decwrl.dec.com> devine@cookie.dec.com (Bob Devine) writes: > I just got home from the San Diego conference (and was greeted >by below-zero temps and lotsa snow). First I want to thank the >organisers for a well-run conference. Handling ~2000 people and >dozens of sessions is not an easy job. I second that! In fact, one rarely gets to appreciate all of the work and effort put into planning the event itself. Peter, Judy, and the rest of the USENIX staff should be thanked for all of the legwork involved in setting up the environment in which the technical presentations were made. >The only complaints I >have are that speakers were hard to locate after they finished >their session (perhaps a room devoted to after-session questions?) >and the mid-morning danish always dissappeared too soon ;-) I had little problem in locating them. They are right up front and can be assaulted at the break. The *real* problem is due to the limited amount of time between sessions and the number of people who want to talk to the speakers. It is extremely difficult to just happen to bumb into people in a crowd of 2000! Would contacting a speaker via their hotel room phone be considered appropriate? > Other suggestions are: >1. Multiple tracks for the main session. I don't like this suggestion > I was at the first part of the discussion of this issue at the > Board of Directors BOF, but I couldn't stay around long enough > to give my opinion. It seems silly that there are multiple, > parallel tutorials and multiple BOFs but only a single > thread for the main session. I would prefer 2-hour blocks > where there are 3 speakers of 1/2 hour per block. The > scheduling of attendees is by self-selection at the breaks. > There could be system administrator thread, a new features > thread, a potpourri thread, etc. There is a problem with both approaches. At the Winter 1986 Conference in Denver there were three technical sessions, one each day. The first was great, in my opinion, it was devoted to Windowing Systems. However, the last two were not of great interest to me: UNIX on the Big Iron and then ADA and the UNIX System. So for me, two days were a waste. However, the San Diego Conference has smaller sessions on various topics. I was able to attend those of interest and not attend others. The Winter 1988 Conference in Dallas utilized the multiple technical session approach. I either missed certain talks due to overlap or I jumped back and forth between sessions to catch parts of various talks. It was not a lot of fun. I see no problem with the format as it was utilized at the San Diego Conference. As long as we don't have one, day long session on the same topic I think it is fine. When there is a block of talks that I am not intersted in, I can find plenty of other things to do. > - Get a "pulse of the crowd" by doing a poll on what > is needed in computing. Or a poll on what would be a > really neat application to have. Any information gathered > is valuable for its own sake and can be used as a news item. > The American Bar Assoc, for example, gets national attention > for its conference when it releases opinion surveys. A much wider sample could be obtained over the net. Brian Reid has experience with conducting surveys over the net. But it may be worthwhile to check the USENIX attendees, many of which probably don't even belong to the Association. > - Allow people to bring along and demo their work in progress. > A description of a user interface is boring without the > actual thing to see/play with. Granted, applications that > run on a Cray could be difficult to demo... How about films of the system instead? I saw such a film at the Winter 1986 Denver Conference on GLO (I believe). It certainly would be less costly. Vendors can afford to do such a thing for products because they will hopefully recoupe the money spent in any sales that are generated by the demo. However, the cost of setting up such a demo for a research prototype may not be worthwhile. Perhaps those who attend SIGGRAPH Conferences can shed some light on this area. There is certainly lots there that is better seen than heard about. > - Use the conference as a way of telling people about Unix. > What is in the weekly computer newpapers is a lot of vendor > press releases. Usenix could be publicized by press releases > that highlight interesting papers. Isn't that the /usr/group crowd does with the UniForum Circus :-) >3. Add more features. > - Add some contests. Something like having a computer > trivia quiz as described in this month's ACM magazine > would be fun. Am I the only person on the planet that thought that trivia contest was a waste of printed space in CACM (a magazine????)? Frankly, I find such contests just as boring as the TV gameshows. > - Have an equipment demo room. A computer conference without > computer is sort of like a automobile show without cars. Demos take place in hotel suites. This past week DEC was showing a new product and NCD was showing their X Terminal. At the Summer Conferences there are far more vendors at USENIX. This time of year they are gearing up for UniForum. > - More panel discussions. There are many areas where there > is One True Way. Opinions matter! Put several knowledgable > folks together to see if heat or light can be generated. I have seen some panel sessions at other conferences produce some truly useful information, while others are a total waste of time. But it should be looked into as something that could be set up. >4. Open up the conferences to more than Unix. This might be a > heretical statement. But, many sites have quite a mixture of > systems. Topics could be on how to manage mixed-computer sites. > Or on how run different OSs on the same network. The point is > that there can be a fruitful exchange of information for this > area. If Usenix doesn't do it, who will? Quick, take away this man's source license :-) May you be condemned to a life of assembly programming on an IBM 370 in the hell of MISland!!! I think that any topic that is somewhat related to UNIX is fair game for the conference, eg. ADA and UNIX session at the Denver Conference in 86. What I would like to see is a wider availability of the Tutorial Notes. I know that limited quantities are sold after the tutorial sessions are over with. However, they sell out fast. If you don't get there in time, you will have to wait until you get to go to another conference. Perhaps USENIX could sell them as they do the proceedings, from the office in Berkeley. -- George W. Leach Paradyne Corporation ..!uunet!pdn!reggie Mail stop LG-129 reggie@pdn.nm.paradyne.com P.O. Box 2826 Phone: (813) 530-2376 Largo, FL USA 34649-2826
henry@utzoo.uucp (Henry Spencer) (02/10/89)
In article <8902090223.AA01916@decwrl.dec.com> devine@cookie.dec.com (Bob Devine) writes: >1. Multiple tracks for the main session. > I was as the first part of the discussion of this issue at the > Board of Directors BOF, but I couldn't stay around long enough > to give my opinion. It seems silly that there are multiple, > parallel tutorials and multiple BOFs but only a single > thread for the main session... I didn't make the Board meeting at all, but speaking as a member of the Program Committee, we'd have been happy to go to parallel tracks if there had been enough decent submissions. There weren't; there were barely enough for a single track. >4. Open up the conferences to more than Unix. This might be a > heretical statement. But, many sites have quite a mixture of > systems. Topics could be on how to manage mixed-computer sites. > Or on how run different OSs on the same network... Again, speaking as a Program Committee member, we only rejected one otherwise-decent paper for having absolutely nothing to do with Unix. We'd have been happy to see papers on mixed environments. Usenix can't put such papers in the program unless people submit them. -- Allegedly heard aboard Mir: "A | Henry Spencer at U of Toronto Zoology toast to comrade Van Allen!!" | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
muller@sdcc7.ucsd.EDU (Keith Muller) (02/12/89)
The nature of the technical session is clearly driven by the papers that get submitted. I spents months calling around and sending letters to "coerce" people into submitting papers. In several cases this worked as couple papers in the conference only appeared after these kinds of "strong armed" efforts were executed. Admittedly some of the flavor might be blamed on who I tried to tap for papers, but I was limited to those areas of research which I was familar with. I also suspect that the members of the program committee acted in a similar way. The actual format of the conference was chosen AFTER the papers were reviewed, not before. So it was the content of the papers that determined the conference, not any decree made by USENIX or the program committee. As you see, people should not feel that they have little control of the flavor of a conference. Through papers that they submit or through people they encourage to submit papers, the conference is formed. The bottom line here is that USENIX members should encourage anyone they know who is doing work that might of interest to the USENIX community to submit papers to future conferences. Without these papers, the conference will become just another trade show. Keith Muller Co-Chairman Winter 1989 USENIX
henry@utzoo.uucp (Henry Spencer) (02/12/89)
In article <5586@pdn.nm.paradyne.com> reggie@pdn.nm.paradyne.com (George W. Leach) writes: >... Would contacting a speaker via their hotel >room phone be considered appropriate? Leaving a message for them with the hotel is less likely to disturb them and more likely to reach them (many Usenix attendees are found in their rooms only when asleep). >... However, the >San Diego Conference has smaller sessions on various topics... For those who don't understand how these things get set up, I should observe that in a conference like the San Diego one, which didn't advertise any specific choice of topics in advance, the session topics get picked to fit the papers, not vice-versa. That is, after the accept/reject decisions on individual papers are pretty much complete, one starts trying to group them into coherent sessions. This sometimes works and sometimes doesn't. > [interactive demonstrations] ... the cost of setting >up such a demo for a research prototype may not be worthwhile. Perhaps >those who attend SIGGRAPH Conferences can shed some light on this area. Here again I can comment with some authority. I was co-chair of the demos track at CHI+GI 87 (joint ACM-SIGCHI and Graphics Interface conference). Getting good demos is very tricky; CHI+GI had quite a mixed bag. We didn't have too much trouble with marketing hype, partly because we'd specifically indicated that systems ought to be demonstrated by their authors. However, it is *very* difficult to tell whether a demo is going to be worth seeing without seeing it. Kate Ehrlich, the other demos co-chair, tried hard to pick good ones based on written descriptions, and concluded afterward that this approach was basically a failure. Also, it is a lot of work running a good demos setup. I saw very little of CHI+GI except the inside of the demos room. We needed a lot of AV gear (it is impossible to do effective demos for a substantial audience without projection video, for example) and connecting 57 different kinds of computers to video hardware can be a serious headache. Showing demo videotapes, instead of live demos, might be a good way to try to assess interest without getting into all the complications. > What I would like to see is a wider availability of the Tutorial Notes. >I know that limited quantities are sold after the tutorial sessions are over >with. However, they sell out fast. If you don't get there in time, you will >have to wait until you get to go to another conference. Perhaps USENIX could >sell them as they do the proceedings, from the office in Berkeley. The availability of tutorial notes is not necessarily under Usenix's control; the authors have a large say in it, and often don't want unlimited quantities distributed. -- The Earth is our mother; | Henry Spencer at U of Toronto Zoology our nine months are up. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
smb@ulysses.homer.nj.att.com (Steven M. Bellovin) (02/13/89)
As other responders have said, the content of the conference is largely
determined by the papers submitted. For example, for Baltimore Usenix
about 10% of the papers submitted appear (from their titles) to be
related to security. That more or less guarantees some presentations
on security topics, unless they're all turkeys. On the other hand, I
don't see any papers on neural networks (though they were explicitly
solicited in the Call for Papers), nor anything on troff and its
friends.
I was on the program committee for the Salt Lake City Usenix (June '84);
let me describe what we did. We sorted through the papers, and
performed a rough triage: great paper, might be usable, and obvious
turkey. We then made another pass on the middle pile, sorting it
further. Then we looked at what themes we had, and started assigning
papers to different sessions. We did have two tracks, one in a large
auditorium, and one in a smaller ballroom, so we had to guess which
would be popular vs. which would require interactions. Panel
discussions, for example, were in the ballroom, so the audience could
heckle (I mean comment) better. Finally, we filled in the holes with 1
or 2 papers that we felt were good enough, and complemented the other
papers in the session. We did not accept papers that didn't meet our
standards, though we were hindered in our judgements because
submissions at the time were of abstracts only, not complete papers.
We also got a lot of flack for the track assignments, from authors who
felt they'd been slighted, or from attendees who thought that two
sessions clashed.
Panel discussions are a tricky matter to organize because you want
topics (and speakers) who will disagree, disagree loudly enough to make
it interesting, but politely enough that the conference doesn't start
to resemble alt.flame. Some people do not like panel discussions
because they leave no permanent record; there's nothing that can be
cited in a later paper. But they're a good way to present current
opinions.
The committee is currently reviewing the submissions for the Baltimore
Usenix; we'll make our decisions in early March. If you have any
concrete suggestions (i.e., topics you'd like to see panel discussions
on), please mail them to the committee before March 1:
usenet: {ucbvax,decvax,decwrl,seismo}!sun!balt-usenix
internet: balt-usenix@sun.com
If you can suggest moderators or panel members, so much the better.
--Steve Bellovindavid@dhw68k.cts.com (David H. Wolfskill) (02/13/89)
In article <5586@pdn.nm.paradyne.com> reggie@pdn.nm.paradyne.com (George W. Leach) writes: >In article <8902090223.AA01916@decwrl.dec.com> devine@cookie.dec.com (Bob Devine) writes: >> Other suggestions are: >>4. Open up the conferences to more than Unix.... > Quick, take away this man's source license :-) May you be condemned >to a life of assembly programming on an IBM 370 in the hell of MISland!!! Ummm.... I think -- as one who seems to have received such a condemnation -- that such a punishment is a bit on the harsh side.... (Yes, it's a daily pair of culture-shocks to go to work where I'm a systems programmer on an IBM machine running MVS/XA, then come back to this UNIX machine at home. Of course, that's one reason I give away logins on the machine (especially to my colleagues at work) -- to show them the Right Way.... Not sure it really does a lot of good, though -- management seems to think the whole group (systems programmers... nearly all of whom have logins here) are a bunch of trouble-makers.... :-) Perhaps just a couple of months would be adequate to convince him of the error of his ways.... :-) :-) Let us not be harsh, david -- David H. Wolfskill uucp: ...{spsd,zardoz,felix}!dhw68k!david InterNet: david@dhw68k.cts.com
root@helios.toronto.edu (Operator) (02/17/89)
In article <1989Feb9.210123.19047@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: >In article <8902090223.AA01916@decwrl.dec.com> devine@cookie.dec.com (Bob Devine) writes: >>1. Multiple tracks for the main session. > >we'd have been happy to go to parallel tracks if there >had been enough decent submissions. There weren't; there were barely >enough for a single track. Why not open USENIX up for more than the very-specialised, very-advanced topics that seem to make up the sessions now? I'm sure there are lots of people out there, sysadmins for instance, who might have something to say that would help make other people better informed about UNIX, but who wouldn't dream of submitting a talk of a general nature or at a novice level to USENIX. As an example of this, Andrew Hume mentioned that he gave a session at a DECUS Symposium on grep and awk, and about 100 people, who were genuinely interested, turned up. But he said he didn't even consider submitting something like that to USENIX because it wouldn't be accepted. Surely if the alternate tracks dealt with things from a different level and/or a different perspective, you'd be less likely to get people complaining about conflicts? In any case, having two or more streams (now there's an appropriate term :-) ) can only let people get more out of the conference. Maybe they can't go to two at once, but if there had only been one stream, one session or the other, or both, might not have been offered. And in the meantime a lot more people are probably able to find something to go to most, if not all, of the time. I understand that USENIX is a *technical* conference, and I like that, but speaking as a first-time USENIX attendee at San Diego, there were an awful lot of the sessions that held neither the faintest interest nor relevance for me or for my work (and the room was never again as full as it was for the keynote). Novice does not necessarily equal non-technical, and with UNIX moving into more areas where the people who have to work on/look after the computers know little or nothing about UNIX, more information is needed on details which don't fall within the scope of, say, the tutorials on BSD Internals (*please* don't schedule this parallel with SysV next time, some of us have to use both) or System Administration. I don't intend this as criticism, just a suggestion that accepting more general topics *as*well*as* the papers that you accept now might be beneficial to everyone involved with USENIX, UNIX and other related things. I apologise if this point of view has been aired already in this forum; I haven't finished reading all the postings here (I'm still catching up on mail and news that accumulated while I was at USENIX :-) ). It is a serious suggestion; please accept it as such. -- Ruth Milner UUCP - {uunet,pyramid}!utai!helios.physics!sysruth Systems Manager BITNET - sysruth@utorphys U. of Toronto INTERNET - sysruth@helios.physics.utoronto.ca Physics/Astronomy/CITA Computing Consortium
rcd@ico.ISC.COM (Dick Dunn) (02/22/89)
> Why not open USENIX up for more than the very-specialised, very-advanced > topics that seem to make up the sessions now? I'm sure there are lots of > people out there, sysadmins for instance, who might have something > to say that would help make other people better informed about UNIX... There's a problem with this suggestion: Although it might increase the number of submissions, it wouldn't increase the number of submissions of the type which interest most of the people who attend now. What it would do is increase the number of people who want to attend...and I think there is already a far-more-than-optimal number of people attending. I would also disagree that the set of topics is very specialized or advanced... > I understand that USENIX is a *technical* conference,... ...which is just the point. The criteria that Henry indicated for papers seem to have been generously applied. When the topics are sufficiently advanced to be of interest to people of moderate sophistication, they have to be somewhat specialized--you cannot examine anything very large in detail in 20 minutes. >...and I like that, but > speaking as a first-time USENIX attendee at San Diego, there were an awful > lot of the sessions that held neither the faintest interest nor relevance for > me or for my work... I won't try to dispute your experience, but I found a lot of papers which, while not relevant to my work or my particular areas of interest, were nevertheless both interesting and understandable. I think there were 34 papers presented if I count correctly; there were about 5 or 6 that, even with hindsight, I would have skipped. >...(and the room was never again as full as it was for the > keynote)... Sure, but you expect a general decline in attendance across the sessions, as people meet other people and get diverted to matters of particular interest. (The presentations are NOT the only thing going on.) Besides, I think a lot of people go to keynote addresses to see "who is this guy, any- way???" Certainly that's the case with someone in charge of such a big chunk of AT&T. I made sure I was there because I wanted to find out whether he would explain the behavior of AT&T in the past year or so that could only with considerable charity be described as "egregiously unusual" and managed to bring IBM and DEC together in OSF. (Alas, I was disappointed; O'Shea was altogether too sharp to explain any of it...but I digress, and #include <stddisclaimer.h> anyway...) > I don't intend this as criticism, just a suggestion that accepting more > general topics *as*well*as* the papers that you accept now might be beneficial > to everyone involved with USENIX, UNIX and other related things. I disagree that the technical sessions should be so redirected, but I do think there's something else here for USENIX to think about. There's a tremendous demand for education on UNIX-related topics, and USENIX is in a good position to capitalize on it because it has some built-in filtering of the high-level bogosity that comes in a lot of traveling mistrel-and-$600- seminar shows. -- Dick Dunn UUCP: {ncar,nbires}!ico!rcd (303)449-2870 ...Just say no to mindless dogma.