I worked in Java, could I see the Haskell equivalent?

[Per Fredelius]

I worked with Java for the labs as I haven't worked in Haskell before.
But I would be rather interested in seeing the solutions made in
Haskell. It seems more powerful and fast doing future language
implementations in Haskell.

Could there perhaps be a code exchange now that the labs are over? Is
this a sensitive request since it might become available to next year
students?

Thanks for a great course anyway. Interesting content and well made
execution.

[klondike]

El 21/03/11 16:15, Per Fredelius escribió:

> I worked with Java for the labs as I haven't worked in Haskell before.
> But I would be rather interested in seeing the solutions made in
> Haskell. It seems more powerful and fast doing future language
> implementations in Haskell.
>
> Could there perhaps be a code exchange now that the labs are over? Is
> this a sensitive request since it might become available to next year
> students?

If teachers agree I and my partner have no problems in publishing our
code under a free license.

[Arash Rouhani]

Hi Per!

I would be interested in seeing a java solution actually, and am willing
to publish my haskell solution, but I want an ok from someone so I don't
get accused for cheating or something ...

I believe Chalmers is just being evil+dumb when not granting this in all
other courses. Discussing solutions should be encouraged, and technical
aid for that should ideally be provided through fire (now this is can be
a question of not having resources, unfortunately :-/)! Certainly the
possibility to see your friends solutions (or the teachers solution)
would improve learning. Actually, I've heard that some friends of mine
formally proposed this idea, but I don't really know what Chalmers (or
whoever was in charged) responded to this.

Arash Rouhani

[csvan...@spray.se]

The problem is that people might just as well use it to cheat and use
it to get their labs done in a cheap way. I believe it is much better
as it is now - the key is finding a lab partner you are comfortable
with and with whom you can discuss your solutions.

[klondike]

El 21/03/11 22:34, gussv...@student.gu.se escribió:

The problem is that people might just as well use it to cheat and use
it to get their labs done in a cheap way. I believe it is much better
as it is now - the key is finding a lab partner you are comfortable
with and with whom you can discuss your solutions.

I come from Spain, where this things are a real issue, and hey, just keeping track of the solutions and reports sent in the last 2, 3 years suffices to catch most of the copy attempts.

I'm going to say something, I got access to one of the secret study documentation sharing networks in there and the material I got was only helpful for:

1. Getting out of a rush when you couldn't meet the deadline at all. But usually...
2. Understanding better what I was supposed to do seeing how others did it.

Again that's just my opinion.

[Arnar Birgisson]

Hi all,

We do encourage discussion in this course, and indeed we think that
this year has been particularly good in terms of how you have used the
mailing list and each other's assistance. So kudos for that.

The reason we ask you not to share code with each other too much is
mainly to avoid controversy when grading. Sadly, there are often cases
where groups submit the same or very similar solution, even sometimes
without one group knowing about it. Letting this happen is very bad
for the university's reputation, and thus it is also bad for the value
of the degrees you will be receiving. If full sharing is allowed, it
becomes so much harder for us to uphold such a standard, as we could
not distinguish between fair sharing and cheating. This is why we
encourage you to discuss solutions without sharing too much detail or
code.

As for publishing solutions, I would kindly ask you not to make your
solutions available publicly. The reason for this is that this would
mean we will have to abandon the current labs and make new ones every
year, since otherwise the course would lose some credibility. If it is
easy to pass a course like this by just taking solutions found online,
regardless if students do it or not, then e.g. industry would not have
much faith in the degrees that Chalmers gives. I'm sure you can see
that this would not be in your best interest either.

It is also quite hard to make good, instructive programming
assignments, and the ones we have in this course have evolved over a
long time. They are improved every year, and there are already a few
things that you discovered this year that will be fixed next year.
Having to write new labs every year would severely hurt the quality of
the course. In addition, the more time we have to spend on looking
for copied submissions, the less time we have to help you on the labs
(and as you know we already have quite limited time).

So on both points the issue is not about cheating, or to make our
lives easier, but rather to maintain the quality of the education.
After all, that's why most of you go to Chalmers in the first place.

That said, we are certainly not interested in preventing you from
learning as much as you can. Sending your solution privately to a
fellow student that has already completed the lab shouldn't be too bad
(but keep in mind that I'm just an assistant, so I don't make the
official policy). Just keep the above points in mind when you do, and
try not to give your solutions to someone that you wouldn't trust to
keep them private.

cheers,
Arnar

[klondike]

El 22/03/11 11:16, Arnar Birgisson escribió:

Hi all,

We do encourage discussion in this course, and indeed we think that
this year has been particularly good in terms of how you have used the
mailing list and each other's assistance. So kudos for that.

The reason we ask you not to share code with each other too much is
mainly to avoid controversy when grading. Sadly, there are often cases
where groups submit the same or very similar solution, even sometimes
without one group knowing about it. Letting this happen is very bad
for the university's reputation, and thus it is also bad for the value
of the degrees you will be receiving. If full sharing is allowed, it
becomes so much harder for us to uphold such a standard, as we could
not distinguish between fair sharing and cheating. This is why we
encourage you to discuss solutions without sharing too much detail or
code.

In what's that different to Spain? How can you be sure they didn't copy their solutions from each other or get it from a third source?

It's not the first time cheating has been discovered on a Spanish university and turns out that discovering cheating increases the courses credibility as it shows that cheating there is harder.

As for publishing solutions, I would kindly ask you not to make your
solutions available publicly. The reason for this is that this would
mean we will have to abandon the current labs and make new ones every
year, since otherwise the course would lose some credibility. If it is
easy to pass a course like this by just taking solutions found online,
regardless if students do it or not, then e.g. industry would not have
much faith in the degrees that Chalmers gives. I'm sure you can see
that this would not be in your best interest either.

Again this is usually your responsability, nowadays:

1. It is easy to pass the course just by leeching from your partner since there are few to no protection in that, and I can tell because this has happened to me in other courses.
2. It is easy to just pay somebody for the solution, funny that teachers have been able to find cases like those on internet.
3. Nothing ensures that anybody isn't catching and sharing the practices code in private networks, this happens in Spain so why it should be different here?

So I think that putting it from this side it is in my best interest to share the code to enforce you to strengthen your anticheating measures.

It is also quite hard to make good, instructive programming
assignments, and the ones we have in this course have evolved over a
long time. They are improved every year, and there are already a few
things that you discovered this year that will be fixed next year.
Having to write new labs every year would severely hurt the quality of
the course. In addition, the more time we have to spend on looking
for copied submissions, the less time we have to help you on the labs
(and as you know we already have quite limited time).

You know what they do on the Compilers course in Spain?

1. Even though the base lab (ie the tools and objectives) are the same, the target language implementation changes, for example one year you have integers and boolean logic, other you have for loops instead of while ones etc.
2. There is a mandatory lab examination, this just consists on you taking with you as many help (books, source code in pendrives etc) as you want and make a small modification to the target language, for example add the ? ternary operator.

And hey, that and comparing the submissions to those of two or three years prior seems to work very well.

So on both points the issue is not about cheating, or to make our
lives easier, but rather to maintain the quality of the education.
After all, that's why most of you go to Chalmers in the first place.

To maintain quality you must maintain quality on all aspects, security through obscurity is not going to work and this has been proved, last case being the PS3.

Keep my words in mind, after all spanish students are sadly famous outside for making other universities reinforce their cheating policies so you can bet we've got some knowledge on this ;)

klondike

[Per Fredelius]

It would be nice to get an 'official' statement on this.

[Arnar Birgisson]

Hej Per,

Official statement on what? As I said, we can't stop you from sharing solutions (nor do we want to), but just ask that you don't publish them for the reasons listed in my message.

cheers,
Arnar

[Aarne Ranta]

Dear All,

Slightly more officially perhaps, as the course responsible I would very much like you not to distribute your lab solutions. Arnar has given all the good arguments, which I don't need to repeat.

Sharing lab solutions *during* a course is, as we all know, strictly forbidden and may lead to disciplinary measures. Maybe this can even be extended to students who continue in our programmes and help people in other courses to cheat.

But after the course and at any case after your graduation - well, we cannot prevent you, we just hope you are cooperative. The web is full of stuff that should better not be there. But fortunately this is the work of very few people, and most people are nice and responsible.

With best regards

  Aarne.

[Arash Rouhani]

Hello,

While I think Arnars arguments are mostly correct, I think you seem to for no reason ignore the advantages of code sharing.

Here below, I'll advocate that it'll be inspirational to show Accepted students the solutions of other Accepted students, and why it isn't harmful:

Given that a student got an 'Accepted' on fire, that means he has a correct solution available and has the possibility to give his code to next years student, that is highly immoral, cheating, and wrong. Ok, so such a student has that possibility. Now we therefore conclude that he don't get any more "cheating power" after getting his hands on another solution of the same lab. ==> Conclusion: It's a not dangerous for a student that has got Accepted on lab X to be able to see his friends solutions for lab X. Rather on the contrary, the student will benefit from it. The student can learn by seeing other students solutions, and nor is it cheating, and nor does it enhance his influence to let students of the next year to cheat.

What do you people think? Is this nonsense or are there any strong benefits in controlled solutions sharing? I for one am very interested in seeing how fellow students implement labs and to what extent they have code reuse, testing, etc.


Cheers,

Arash Rouhani

[Adam Bengtsson]

Hi All,

I have followed this discussion with great interest, and I think Arash has made a good point.

Why not in a future release of Fire, add the possibility for students which are accepted at all their labs, and have agreed to share their solutions, the possibility to see other students which in return share their own solutions? This way the code would not become "more public" than it already is, and we still get the benefit from code sharing. 

[klondike]

Seeing as things stand all I can say is that I agree with Arash and
Adam and maybe that will be the best solution for all of us.