This is a great idea. I am for it.
From: eclipse.org-eclipsecon-program-committee-bounces@xxxxxxxxxxx
[mailto:eclipse.org-eclipsecon-program-committee-bounces@xxxxxxxxxxx] On Behalf Of Ed Burnette
Sent: Thursday, December 15, 2005
7:20 AM
To: Eclipsecon Program Committee
list
Subject: RE:
[eclipse.org-eclipsecon-program-committee]Longtalkbreakdownby projects and the
two thematicvirtualtracks(embedded and RCP)
There's always going to be more
presentations than slots to give them. So how about this idea:
1. Accept an unlimited number of quality
presentations. Put all of them on the web site[1].
2. Give a limited number of the
presentations live at the conference. The live ones would be specially marked
and schedulable of course, but other than that they would be equal to the
non-live ones.
So what we're currently doing is figuring
out which ones should be given live. But instead of "declining" all
the rest, we could have an intermediate kind of "accept" that means
we want them to be part of the conference knowledgebase, just not given
live.
Let's face it, the number of people
physically attending a conference is *far* outweighed by the number of people
who are interested in the material and read / listen / view it online. I would
guess the ratio is over 100 to 1.
With this idea the "accept for live
presentation" decision is more like deciding whether you want to go see a
movie vs. renting it and watching it home later. Do you get something from the
big screen experience? Similarly, which presentations would benefit the
listeners most who are there live?
[1]: "On the web site" means at
a minimum there is a downloadable slideshow. But richer media is an option if
the presenter wants to supply it, such as audio, screencast, or video.
From: eclipse.org-eclipsecon-program-committee-bounces@xxxxxxxxxxx
[mailto:eclipse.org-eclipsecon-program-committee-bounces@xxxxxxxxxxx] On Behalf Of Tim Wagner
Sent: Wednesday, December 14, 2005
1:09 AM
To: Eclipsecon Program Committee
list
Subject: RE: [eclipse.org-eclipsecon-program-committee]
Longtalkbreakdownby projects and the two thematic virtualtracks(embedded and
RCP)
+1 on 60à45min for long talks so
that we can accept more.
It’s a tough question –
there’s a huge amount of value in keeping a single, yearly conference,
but it necessarily means that we’re bringing together some fairly
disparate interests, and all of them require some minimal level of
representation. FWIW, I think a tightly packed program is still better than the
alternatives so long as it’s possible to attend what you’re
interested in. (My big complaint about JavaOne isn’t that it’s a
large program, but that they make it hard to read the program (and thus find
what you’re interested in seeing) and then you often can’t get into
the popular talks anyway without lining up an hour or more in advance. I hope
we can avoid those problems in EclipseCon…)
In terms of future breakdowns, I’m a
fan of domains for several reasons: presenters usually know what area
they’re in, attendees often come with a list of topics they want to learn
more about, and PC members can rely on top-level projects or other domain-based
groups to help provide expert advice within such buckets. Some special
categories operate similarly – for instance, we had the RCP and embedded
“themes” this year, business is a special case, and
“newbie” might be as well. It’s also an easier way to make
progress as a program committee – reading every submission is
increasingly daunting, so breaking down into separate groups (assuming the allocation
can be worked out) feels more scalable as a process.
