[comp.org.usenix] Presentations

dvk@sei.cmu.edu (Daniel Klein) (07/13/89)

In article <15901@vail.ICO.ISC.COM> rcd@ico.ISC.COM (Dick Dunn) writes:
>	- It takes skilled people and special equipment to produce good
>	  slides.  Many smaller organizations don't have it.  Good overhead
>	  transparencies can be produced on a < $4k laser printer on a PC.
>	  Smaller organizations DO have this sort of equipment because the
>	  dog-and-pony shows they're doing all the time need overheads.

Good slides can be made just as easily as good transparencies.  If you don't
want to go into all the glitz of color, you can go to your local photo shop
and have them create slides from paper copy for around $10 for a 24 slides.
Since 24 is too many for a 20 minute talk - you win.  Slides are cheap,
especially when you figure that overheads typically cost $0.50 each anyway.

>	- The overhead format is more convenient and familiar to many
>	  folk.

Yeah, we've always done it that way, and we don't want to change, even if
it's better another way.

>	- When you're doing that one last proofreading on Friday afternoon
>	  before you go to the conference, and you find that you wrote
>	  "Denise Ritchie", you can run off a new overhead yourself in a
>	  couple of minutes with nobody else's help.
>	- When you're sitting in the hotel room going over the talk the night
>	  before, and you discover that LR(k) somehow became LURK, you can
>	  patch over it with a marker.

This is one of the biggest problems at USENIX conferences - the last-minute
mentality.  If you are going to be making a presentation to over a thousand
people, you simply cannot make your slides at the last minute.  You cannot
prepare your talk on the Friday before.  It is absolutely unconscionable to
waste that many people's time by having an ill prepared talk.  Quite simply,
PLAN AHEAD.  Have your talk ready a month in advance.  Have your slides ready
a month in advance.  Don't complain about it - just do it!  For the D.C.
conference (coming up), all authors will have 2 months from notification of
acceptance until final paper deadline, and then another two months to get
their talk together and prepare slides.  It is unthinkable that anyone should
wait until the last minute to get their shit together, just so that they can
bore a huge number of people with bad slides and a choppy talk.

>	- When you're answering a question about "that slide about work-
>	  stations promoting Ferrari ownership" you can go back to it
>	  directly (by calling for the overhead) instead of subjecting a
>	  thousand people to the dizzying process of cycling backward
>	  through two dozen slides in an automatic changer.  (In Baltimore,
>	  the people who used slides generally had trouble backing up at
>	  all.  This is a common phenomenon.)

A 20 minute talk should have between 10 and 15 slides, maximum.  Flipping
back through 15 slides won't hurt anyone, and a well prepared speaker will
know the order of their slides ahead of time, so there should be no "hunting"
for that correct slide.  Calling for the overhead is just as time consuming,
because more often than not, the overhead turner does not know the talk, and
is only there to flip pages.  Use slides!  They are better, and as cheap.

>I'm curious about just how many speakers or potential speakers really have
>access to such facilities.  It must be very convenient for it to be useful,
>for the reasons I mentioned above.  Obviously the AT&T or DEC sized places
>can do it.  At the size of a company like Interactive (a few hundred
>people) the art dept exists but may not be geared to producing material for
>technical presentations.  Where's the breakpoint in the commercial world--
>perhaps a Kperson?  How are universities set for this?  Note that it's more
>than just having an art dept; they have to be able to deal with your
>material, which may contain tricky technical stuff (and they must be
>trained *not* to fix things which look obviously wrong or silly!:-), and
>they have to be able to turn it around pretty fast.

The SEI has 150 employees.  We have an art department.  Smaller organizations
also have them - the key is to PLAN AHEAD.  If you get in on the art department
queue EARLY, you can get your artwork out in time.  Typically, our art
department has a 2-3 week waiting list.  All I have to do is not wait until
the last minute, and all is fine.
-- 
-- =============--=============--=============--=============--=============--
"The only thing that separates us from the animals is superstition
and mindless rituals".          Daniel Klein   CMU-SEI   +1 412/268-7791
				dvk@sei.cmu.edu    uunet!sei.cmu.edu!dvk

rcd@ico.ISC.COM (Dick Dunn) (07/14/89)

I seem to have hit a nerve with both Steve Bellovin and Dan Klein--both of
whom have presented (good) papers at recent USENIXes and both of whom were
on the program committee for Baltimore.

They may be pretty sensitive about quality of presentations, and I probably
didn't convey quite what I meant.  I was not trying to say that "overheads"
are superior to slides--but only that each format has some advantages over
the other.  I'd like to see good equipment which can accommodate either
format.

In article <3624@ap.sei.cmu.edu>, dvk@sei.cmu.edu (Daniel Klein) writes:

> >	- The overhead format is more convenient and familiar to many
> >	  folk.
> 
> Yeah, we've always done it that way, and we don't want to change, even if
> it's better another way.

I think this is a cheap shot.  I don't agree that slides are better in all
respects, and I think that if you're standing in front of a thousand people
trying to do a smooth presentation, there's a distinct advantage to working
with a familiar format.  People *do* fumble with slide projectors!  They
shouldn't--in theory a slide projector is trivial to operate--but theory
doesn't matter unless practice obliges.

The main objection seems to be to my suggestion that overheads are much
easier to fix at the last minute, either by running off a new one just
before you leave for the conference, or by manual correction with a
marker.

> This is one of the biggest problems at USENIX conferences - the last-minute
> mentality.  If you are going to be making a presentation to over a thousand
> people, you simply cannot make your slides at the last minute...

I agree...but that's not anything like what I was suggesting.  I was
suggesting that you might find a mistake at the very last minute and want
to correct it.

Do better proofreading, you say?  Fine, that will reduce the chance for
error.  With the small number of slides you need for a 20-minute talk, the
errors should be down to almost zero.  Yeah, almost.  If you don't think
you can prepare a talk with incredible care, go over it repeatedly,
study every detail, and still miss an error, either you haven't given
enough talks or you lead a charmed, Murphy-free existence.  Chances are you
won't have anything you need to correct if you're even reasonably careful--
except that (Murphy) the harder it is to make a correction, the more likely
you are to find one at the last minute!

Even if you don't have any errors, there's always the chance that you'll
get that last minute call or email that says, "Hey Billybob, y'know that
test you wanted me to run on the Frobozz/2?  I finally ran it, and you
won't believe the numbers!"  Of course, you can ignore the last-minute
info, but unless it's really too late to incorporate it, that's a cop-out.
(USENIX is supposed to be timely, remember?)  If you have to put up a
slide you couldn't correct, and point out the change, it will break your
stride in the talk, waste some of your valuable time, and confuse the
people whose attention was drifting just then.

It's THESE things that you can avoid by being able to make a last-minute
minor correction.  You'd better have your talk all ready to go well in
advance, as both Klein and Bellovin said.  In no way do I advocate doing
the preparation at the last minute, but I might like the option of being
enough of a perfectionist to make a small correction at the last minute,
without the constraints of the visual medium being in my way.

> A 20 minute talk should have between 10 and 15 slides, maximum.  Flipping
> back through 15 slides won't hurt anyone...

...IF the speaker knows how to use the projector in reverse!  (Again, that
*shouldn't* be hard, but it seems to be.)  It is disruptive to have to
watch the flash/flash/flash, but I guess it's rare enough that it's not too
bad.

> ...Calling for the overhead is just as time consuming,
> because more often than not, the overhead turner does not know the talk, and
> is only there to flip pages...

Why is that?  By earlier arguments, if you're presenting a talk to this
many people, you'd better have rehearsed it.  I hope it's in front of
people who can understand and critique it!  So rehearse it in front of the
person who's going to turn the overheads for you.  non-problem.

Again, I'm not trying to assert superiority of overheads; I'm just trying
to put them at parity with slides.  I've seen enough overheads done right
that I know they can work - and work well enough that other factors in the
talk are more important.  They have advantages as well as the disadvantages
that others have mentioned.

There are even problems with both media--such as trying to anticipate the
readability in a room of > 10^3 people.  There were some poor slides at
Baltimore--one set in particular was just too dark to read.
-- 
Dick Dunn     rcd@ico.isc.com    uucp: {ncar,nbires}!ico!rcd     (303)449-2870
   ...Simpler is better.