Dear Susanne, and all,
I am sorry that you are disappointed with the way I have organised FEniCS
this year. If you elect not to come, that is of course your choice, however
I would echo what Marie said: your input is valued and you will be missed
personally, so please come. In particular, I take total responsibility for
the organisation of FEniCS this year, so do feel free to take issue with
how I'm doing it, as you have done. Don't feel you have to not come in
order to make that point.
The core of what is going on is that the FEniCS workshop this year has
attracted a lot of interest. There were 19 presentations in 2012, 23 in
2013, and 22 last year. This year 50 abstracts were submitted. This is
probably some combination of the ongoing growth FEniCS usage*, the effect
of increased advertising of the workshop this year, and the effect of
hosting in the UK where there seem to be a lot of FEniCS people (Cambridge
was a big year in terms of attendance too). (*I'd love to say that it's the
explosion of interest in Firedrake, but that's clearly just wishful
thinking!)
I did suspect in advance that this year was likely to be bigger than
previously, and a number of changes were introduced. First, FEniCS this
year is half a day longer than last year. However this doesn't actually
translate into a lot more talk slots (we have 25 instead of 22) for two
related reasons. The first was informal feedback from last year that formal
plenary discussion sessions don't work so well but we need more informal
discussion time, such as over coffee and lunch. The second was my
perception that ramming a large number of talks into each day is
counterproductive: people lose concentration and the speakers don't have
the chance to go into any depth. We therefore have 10 talks per day, which
I believe is about the limit of what one can sensibly concentrate on (it
might already be too much). We also have 2 x 45 minute breaks and 90
minutes for lunch. Actually there isn't a lot of slack in the breaks, since
we may have up to 100 attendees this year and so we'd probably need 30
minute coffee breaks and 90 minutes for lunch just on logistical grounds.
A second way to accommodate more talks would be to extend the meeting, or
have simultaneous sessions. Neither of these is possible at this stage
because they change the budget and affect the number of rooms we need (and
rooms at Imperial are a big deal - lunch is already going to be crowded due
to our limited space allocation). I did already extend by half a day over
last year but it was impossible to guess how much to extend in advance. I
doubt that parallel sessions would be popular at this workshop, but
extending to 3 or 3.5 days is something that next year's organisers may
consider if they think that FEniCS '16 is also going to be large.
The other way that we looked to accommodate a larger meeting is by having a
proper poster session. In contrast to having lots of short talks, my
experience is that a good poster session allows a lot of authors to go into
significant depth with very interested audiences. It allows for a level of
interaction which doesn't really occur in talks. A part of having a good
poster session is the workshop organisers giving it sufficient prominence
and support. Our poster session is scheduled on Monday after only a half
day of talks so everyone should still be fresh. In addition, because we
know how scientists work, we will be providing drinks and canapés at the
poster session. Experience from EGU, AGU and SIAM CSE show that that helps
ensure everyone goes and interacts with the poster presentations fully.
Knowing that we might run out of oral slots meant that there might be
competition for places. This was not a surprise or an emergency measure,
and all authors were advised of this when submitting their abstract. The
exact wording from the abstract submission form was as follows:
"Please indicate whether you wish to be considered for an oral
presentation or for a poster. Authors not selected for oral presentation
may be invited to present a poster."
Except for text box headings, this is the only sentence on the submission
form, so it's not exactly buried in the small print. Interestingly quite a
few presentations were submitted as posters only - so many people chose
this route.
This concept of submitting either for a poster, or asking to be considered
for an oral slot is exactly the format that AGU and EGU have been using for
many years. The difference we made was that at EGU (and, I believe, AGU)
you only get a decision, you don't get to see any reasons. The review
process was that the scientific committee reviewed every abstract which
asked for an oral (except where they were themselves an author). I also
read all the poster submissions, just to make sure none were completely off
topic. This means that I read 45 abstracts and the other two committee
members reviewed between 30 and 40. The abstracts themselves are short so
this is a manageable load, but it does mean that he reviews are short
reasons for decision, they are not a comprehensive analysis of the work
presented. Indeed on the basis of short abstracts, it would be impossible
to do that.
Clearly short abstracts are a limited piece of information, and it's always
possible to disagree with reviewers' judgements (indeed reviewers make bad
calls on important papers, not just on which abstracts should be talks and
which posters at informal workshops). However it is a reasonably efficient
process and given that it has such low consequences: nobody loses a
publication, everyone still gets to present, I think it's an appropriate
one. If the consensus is that people would rather not see the reasons for
decision because they are upsetting, then I'm sure that's something that
future hosts could take on board if they have a similar review process.
Personally, having had EGU decisions just drop out of thin air, I would
rather see some brief reasons, but I concede that's an opinion.
The final point that I would like to make is that the standard of work in
this community is high. The quality of work in the poster abstracts is such
that anyone involved in the FEniCS and Firedrake world should want to see
them. Like Marie, I think the poster session was the highlight of SIAM CSE
15, and it usually is at EGU and AGU too, so I look forward to seeing it at
FEniCS '15.
Regards,
David
PS, In the light of what I've said about popularity, get your fingers out
and register! We may well sell out, and I have no space to allow extra
registrations.
On Mon, 11 May 2015 at 12:36 Marie E. Rognes <[email protected]> wrote:
>
> Dear Susanne,
>
> I am very sorry to hear that you are not planning on attending FEniCS
> '15 this year, I was very much looking forward to seeing you
> again.
>
> At SIAM CSE this year, there was approximately 300 posters and the
> poster sessions was one of the highlights of the conference: they were
> extremely enjoyable, interesting and constructive. I easily believe
> that this also can be the case for the poster session at FEniCS this
> year and in years to come, and I am looking forward to being a part of
> it myself (as a small contributor to the dolfin-adjoint poster). Could
> you somehow be persuaded to change your mind about contributing?
>
> Disclaimer: I have had absolutely nothing to do with the organization
> of FEniCS '15 this year.
>
> Best,
> Marie
>
> --
> Marie E. Rognes, PhD
> Head of Department
> Biomedical Computing Department
> Simula Research Laboratory
> P.0.Box 134, 1325 Lysaker, Norway
> http://home.simula.no/~meg/
> Skype: m.e.rognes
> Cell: +47 45 66 23 96
>
>
>
> On 05/11/2015 01:19 PM, Claus, Susanne wrote:
>
> Hi,
>
> I have long debated with myself if I should say something or if I should
> just not participate at FEniCS 15 but in the end I felt I had to speak up.
>
> I was disappointed last Friday to receive a rejection to give a
> presentation at the FEniCS 15 conference. I always enjoyed this conference
> in the past and felt it was a safe and constructive environment. So I was a
> bit shocked that I was rejected based on reviewer comments stating that my
> work has no influence on the FEniCS/Firedrake technology and being only an
> application built on top of FEniCS or that I had not made enough progress
> in the last year to present. These are unfair and generic comments. You
> could argue this about most developments by claiming that they are “just an
> application” or that not “enough” progress was achieved.
>
> Additionally to these unclear reviewer comments, I did not understand
> that there even was a pending decision on whether the contribution was
> categorised as a poster or as a presentation. This should have been made
> clear from the beginning. Now, it seems more like an emergency solution to
> me.
>
> I understand that it is a challenge to schedule the high number of
> presentations but forcing people into presenting posters with negative
> comments is the wrong way to go about solving this challenge.
>
> What message does this send and what atmosphere does this create?
> Would it not be possible to schedule the presentations in a different way
> (e.g. shorter time-slots per presentation, or different presentation
> time-frames) and give everyone a chance to speak?
>
> Best wishes,
> Susanne Claus
>
>
> _______________________________________________
> fenics mailing
> [email protected]http://fenicsproject.org/mailman/listinfo/fenics
>
>
>
_______________________________________________
fenics mailing list
[email protected]
http://fenicsproject.org/mailman/listinfo/fenics
