Re: DebConf 2 post-mortem
On Wed, 10 Jul 2002 11:03, Andreas Schuldei wrote:
> * Colin Watson (firstname.lastname@example.org) [020710 03:59]:
> > I thoroughly enjoyed this year's event. The talks were all
> > well-delivered and spanned a wide range of topics, and although time
> > management certainly wasn't strictly to the clock I think the relaxed
> > and somewhat freewheeling discussion on several occasions was quite
> > productive. The commercial use / installers discussion on the opening
> > day and the debconf talk were particularly good examples of this.
> I dissagree on that one. The only well delivered talk was the one by
> demian Connway. (and that one was as usefull as a wart, eventhough fun
> for those good at perl.
The only talk that I didn't find useful was the one on obfuscinated Perl,
he's a great speaker (obviously much more skilled than everyone else who gave
a talk) but he wasn't speaking about Linux or Debian.
I suggest that a requirement be that the talks involve Debian in some way
(could just be a Debian spin on a talk that is not Debian specific) and that
all talks are specific to Linux/Hurd/BSD (the Debian OSs). Also the entry
requirements for a BSD talk should be higher than that for a Linux or Hurd
talk because the vast majority of BSD users won't use Debian/BSD and
therefore it needs to be even more relevant.
The talk about Debian/BSD was very suitable as it was all about porting
Debian, but talks about general BSD issues would not be.
> it was a problem that most did not want to admit
> that it went over their hads.) Most speakers made up the talk as they
> went and/or lacked basic skills (involving gestures, looking at the
> audiance (not mostly at the screen/overhead display)).
I have been thinking of running a BOF at a future conference on how to give
good talks. I don't claim to be the guru of public speaking, but I'm sure I
can give some useful advice (and run a forum for discussing other ideas). I
thought about doing it for Linux-Kongress, but I'm already over-committed for
that one. Andreas, if you want to run a BOF for L-K then I'd be happy to
help you out.
> If I am going to
> deliver a talk to 100+ people for 30-60min, then i should have the decency
> to practice that talk a few(!) times beforehand. I think those instances
> which went out of hand (time and schedule wise) are at least partly due to
> that fact. A talk gets (much!) better by rehersing it. Even for nativ
Regarding the length, probably the best thing to do is to have space reserved
for the maximum amount of time a talk will go. It's easy for a talk to vary
in length by ~30% or more depending on the number of questions and the
percieved understanding of the audience.
Also to give a good talk you need to have the environment set up well, things
would have been easier on the speakers if the PA system had been operational
and if things had been ready to start on time.
Sitting around for 30 minutes or more waiting for another talk or a
key-signing to finish really doesn't make it easier to give a good talk.
As for rehearsing a talk, that's often impossible for group talks.
Finally, we should keep in mind that most people don't have the natural
talent for public speaking that people like Rusty seem to have. Many (most?)
people who have the technical skills required to give a good technical talk
on Linux issues have less communication skills than the average for the
population. People who start speaking have to start somewhere, we should set
our criteria for conferences the same as our criteria for Debian development,
technical merit comes first and the UI is a lesser priority, then people can
move up to speaking at OLS after doing a couple of DebConf's.
I do not get viruses because I do not use MS software.
If you use Outlook then please do not put my email address in your
address-book so that WHEN you get a virus it won't use my address in the
To UNSUBSCRIBE, email to email@example.com
with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org