intfiction.org

I would like the idea of an "one update allowed" compromise. As in, midway during the voting period the creators can publish one update if they so wish, and they are all published at the same time. That would get rid of the "yeah, a tester found a game crasher one day after I submitted" issues with the games while not creating a huge chaos of game updates.

_________________
Janos
Vornasblogi: Gaming, diving & urban exploration

I can imagine a lot of people waiting for "one update day" before they begin playing, since then they'd know they'd be getting the final versions, and not bother with the earlier ones.

_________________
Games, Fonts, and Font-Toys For Gamers - http://www.cumberlandgames.com

This might be, although this is one of those uncomfortable things that get solved only by trying. Then again, I could imagine many people (including myself) "having a say" in the games and just generally making them better.

_________________
Janos
Vornasblogi: Gaming, diving & urban exploration

Interesting.. How many times do you let them change their answers and do you up their grades accordingly?

_________________
Work in progress: D-Day (Adrift V.5)

Isn't it pretty standard in writing/composition courses to hand in increasingly polished drafts, which (hopefully) take the teacher's feedback into consideration? Maybe not for final exams, but otherwise how would the students learn? I guess the question here is whether we want the Comp to be more like a final exam. (I personally am all in favor of intra-Comp updates, with the understanding that no one really wants to have to use them.)

Depends on the class. In some writing-intensive courses (which I haven't taught for a while) I've had students do one (mandatory) revision; sometimes I had the grade for the revision replace the grade for the original paper, and sometimes (mostly when I had got annoyed at the "I'm only going to change things where you wrote a note in the margin, and maybe not all of those" style of revision) I kept the original grade, gave relatively specific assignments for what they had to improve in the revision, and gave them a separate grade based on how much they had improved their paper. This is the Post-Comp Comp style of revision, I think.

Recently I taught a metalogic course where I let them revise their assignments as often as they liked to get it right. Part of this was that the material was difficult enough, or I was doing a bad enough job of teaching it, that it was a challenge to get some of the basic strategies down, and it showed a non-trivial amount of comprehension to be able to take a comment like "To finish the proof, you have to do this" and actually be able to do it.

In my intro logic courses, I don't let people revise their tests -- that would just be giving away the answer, and anyway part of the point is that they demonstrate mastery of technique rather than understanding of the specific proofs -- but they can use the final exam to make up for poor performance on the midterms, because in order to do well on the final they have to have learned the stuff they got wrong on the midterms.

The limiting factor on most of this is how much time it takes me to grade the stuff and give feedback. Giving people specific instructions for revision is pretty time-consuming; the metalogic stuff took them a lot more time to do than it did for me to grade so I didn't get swamped. But I think that isn't a problem in IFComp, because no one's obliged to give feedback or play the revised games. The other question is whether IFComp should test the authors' mastery or if it should be a means to getting them to produce good games by the end of the competition. I prefer the second, but the first might be more competitive.

(There's also the question of what the point of revising is, if people don't play the games at the end of the comp. I don't know about this, though I will say that if you do a revision of an Inform game don't post it to IF-Archive as a .zip -- that disables online play and will seriously cut down the number of people who try the revision.)

And I don't think that graded exams are a particularly good analogy for the Comp. It's not a matter of demonstrating retained knowledge, or showing off mastery of skills (which is what exams and graded papers are for); it's about the quality of the product. The comp isn't grading authors, it's grading games.

(I am planning to add zip support soon!)

This is excellent news!