Slides 7th GhentFPG meeting & meeting format

[Jasper Van der Jeugt]

Dear all,

I have attached the slides used in the 7th GhentFPG meeting. I don't
have Pieter Audenaert's slides yet, due to a small technical problem,
but they will also be posted here once they are available.

Since there seemed to be some sort of interest during the meeting for
some problem solving, I was thinking of changing the format of the
next meeting (which will be held at the end of June) a bit: we could
first have a talk, then do some collaborative problem solving (e.g.
one problem, 40 mins), and then have another talk. Any thoughts on
this?

Cheers,
Jasper

[Kenneth Hoste]

Sounds great to me (should be no surprise since I made a suggestion in
this direction during the meeting ;-)).

Maybe problems from previous VPW programming contests are good for this?
Or Euler problems?
Do we want to focus on training for VPW-2012?

Maybe it's a good idea to have two talks first, and then the problem
solving part? People who are not interested in the latter can leave
earlier, and then we don't needlessly postpone the 2nd talk?

K.

[Nicolas Trangez]

On Thu, 2011-04-28 at 10:57 +0200, Kenneth Hoste wrote:
> Maybe it's a good idea to have two talks first, and then the problem
> solving part? People who are not interested in the latter can leave
> earlier, and then we don't needlessly postpone the 2nd talk?

Even though I am interested in the "problem solving" part, I do think
this is a valid remark.

Nicolas

[Jasper Van der Jeugt]

>> Maybe it's a good idea to have two talks first, and then the problem
>> solving part? People who are not interested in the latter can leave
>> earlier, and then we don't needlessly postpone the 2nd talk?
>
> Even though I am interested in the "problem solving" part, I do think
> this is a valid remark.

OK, that's certainly fine for me as well. I was more thinking of the
problem solving part as a sort of "break" between the talks, but I
understand this would be unpleasant for those who are not interested
in this.

Cheers,
Jasper

[Jeroen Janssen]

Only if they are short talks. There are quite some people that have to get up early the next morning or that have to catch their trains etc.

I would propose to have it start from 19u and have two talks of half an hour each at the maximum then.

[Kenneth Hoste]

I just want to avoid that we need to cut off the problem solving part
when it gets interesting, to make sure the 2nd speaker gets his part of
the cake.
Also, I don't want to scare people off because of the problem solving
part in the middle in which they may not be interested (although it may
also attract more people).

I was just adding to the discussion, I'm not saying I strongly feel the
problem solving part should be at the end. :)

K.

[Andy Georges]

Hi all,

On Apr 28, 2011, at 10:57 AM, Kenneth Hoste wrote:

> On 04/28/2011 10:51 AM, Jasper Van der Jeugt wrote:
>> Dear all,
>>
>> I have attached the slides used in the 7th GhentFPG meeting. I don't
>> have Pieter Audenaert's slides yet, due to a small technical problem,
>> but they will also be posted here once they are available.
>>
>> Since there seemed to be some sort of interest during the meeting for
>> some problem solving, I was thinking of changing the format of the
>> next meeting (which will be held at the end of June) a bit: we could
>> first have a talk, then do some collaborative problem solving (e.g.
>> one problem, 40 mins), and then have another talk. Any thoughts on
>> this?
>
> Sounds great to me (should be no surprise since I made a suggestion in
> this direction during the meeting ;-)).
>
> Maybe problems from previous VPW programming contests are good for this?
> Or Euler problems?

Both seem good proposals. Alternatively there's SPOJ. We should also consider real-world problems. For example, how do I easily script some tasks, etc. Ideally, we would have some larger project with a number of smaller parts that can be solved in a very limited time.

> Do we want to focus on training for VPW-2012?

Some of us will be in the jury, so there might be a problem with that, since we cannot advise on which problems to tackle then. If the chosen problems are too close to those for the VPW, it is unfair towards other VPW participants, and if they're totally dissimilar, the usefulness is somewhat diminished.

> Maybe it's a good idea to have two talks first, and then the problem
> solving part? People who are not interested in the latter can leave
> earlier, and then we don't needlessly postpone the 2nd talk?

I have pondered about this. There are pros and cons for both approaches. Having the solving at the end might mean fewer people stay to do it, and it would make it harder to limit the time spent on the problem. OTOH, people are not forced to sit through 40 minutes of doing nothing if they're not interested. Having it between talks means that we could set a hard limit on solving time, but we force people to sit through it. It's not because we think it is great fun (it is, really!) that all attendants think so. However, I would like to see more students getting involved and solving problems might be a way to get there. After all, the first meeting had a pretty large student attendance, maybe we ought to get back to that, without rejecting the other people who usually travel for quite some time to attend the meetings.

-- Andy

[Tom Schrijvers]

I am in favor of 2 talks and a programming session.  The programming
session gives us more opportunity for informal discussion after the
talks.

Kenneths proposal is a good one: the talks first, then the programming
which is more open ended.

> Maybe problems from previous VPW programming contests are good for this?
> Or Euler problems?
> Do we want to focus on training for VPW-2012?

We can have a mix of problems, ones that don't take too much time to
solve, leaving room to compare alternative solutions.

Like inviting a speaker for a talk we can invite someone
to select a problem and briefly present it.

[Jeroen Janssen]

Indeed, I definitely would not put the problem-solving in between talks as was proposed by someone. You can't force a person to stop thinking about a problem after half an hour, so I am afraid the speaker that comes after the problem solving would not get that much attention.

So, can we agree on two short talks (an hour max) and a programming session?

[Andy Georges]

Hi all,

On Apr 28, 2011, at 11:07 AM, Tom Schrijvers wrote:

> I am in favor of 2 talks and a programming session. The programming
> session gives us more opportunity for informal discussion after the
> talks.
>
> Kenneths proposal is a good one: the talks first, then the programming
> which is more open ended.
>
>> Maybe problems from previous VPW programming contests are good for this?
>> Or Euler problems?
>> Do we want to focus on training for VPW-2012?
>
> We can have a mix of problems, ones that don't take too much time to
> solve, leaving room to compare alternative solutions.

Be aware that what one person deems easy may be quite hard for somebody else. We saw that frequently at the VPW ...

> Like inviting a speaker for a talk we can invite someone
> to select a problem and briefly present it.

I like that suggestion. Also, it is important that we do not require people to use one single language (read: Haskell), even though most of us will likely stick with that. It might be interesting to see how people tackle it in various languages, but perhaps we lack 'language specialists' to give feedback to e.g., students when they use something other than Haskell.

-- Andy

[Andy Georges]

19:00 may de difficult for people working and coming a long way

[Tom Schrijvers]

> 19:00 may de difficult for people working and coming a long way

How about polling the past (non-)participants about their starting
time and content preferences?

Tom

[Andy Georges]

All right, I'll draft a poll.

-- Andy

[Andy Georges]

On Apr 28, 2011, at 12:25 PM, Tom Schrijvers wrote:

You can fill in the poll here: https://spreadsheets.google.com/spreadsheet/formResponse?formkey=dHlnbjV6cEQxLUxjVDlaT084UTRJc2c6MQ&ptok=5785533074078047443&ifq

-- Andy