I think code-chef should follow code-forces. During the contest,the judge will test only on some test data(pre-tests) and after the completion of contest,system test will occur and all AC solutions will be rejudged with the whole test-data . This will reduce the load on server during the contest and reduces the submission queue. I think code-chef should follow this because for every cook-off the number of contestents will definitely increase and the submission load on server will increase. so judging by using only pre-tests is a very good idea.
Advantages
Reduces the submission-queue on sever because of fast testing for each submission.
Increases the competition among users because the user who submitted the perfect solution will only get accepted after the contest during the system-test.
I think it wont be a good idea, because codechef mainly aims for ACM ICPC. In the last contest, there was lot of participation due to which there were some problems in conducting it smoothly, but normally it runs fine
ofcourse !! but a code-chef contest is only to test the programming skills of a contestent but not for testing endurance and tolerance of a contestant.
I don’t think its a good idea. Getting a problem right lets me move on to the next one without hesitation. If I am filled with apprehension and doubt that my current score of 4 can change to 2 after the contest without me being able to guess what the error was, I will be depressed.
Not to mention that some problems fall within the user’s interest range and getting them wrong in the end when having moved on to a new one will frustrate people even more.
@trek: My suggestion will be–> please be patient. Seeing your record, you have participated only for the last few months on codechef, and unfortunately been acquainted only with the bad times. Keep calm and continue coding.
I compete in CodeForces only time to time, but half of those contests is unrated, because of some technical problems But it was the same with TopCoder - when I started, there was limit for 1800 participants in SRM (maybe less) and always when it was full, they increased the limit by 50. Today the the limit is something like 2200, but there were technical problems almost everytime when they increased the number of prticipants…
What I’d like to be implemented is the priority queue - judge the more difficult problems first.
In contests yesterday I had prepared 2 problems for submit in that 10 minuted break (when submitting was not allowed) and I was surprised, when I tried to submit second problems and there was a message, that I can submit only once in 180 seconds, that REALLY sucks…
I call bullshit on this. There was a time when many contests in a short time were unrated, but there hasn’t been an unrated round in months. To be precise: out of 67-5 (due to Black Day) rounds conducted during the past year, 1 was unrated (probably +1 that was lost due to Black Day). At best: 2 out of 67 contests. But sure, half is unrated, who cares about the real numbers just to prove my point!
You have some problem with me? You are always so offensive, at least read what I wrote - “I compete in CodeForces only time to time” (see my profile to understand) and probably I have a bad luck…
When you’re aware that you only compete from time to time, why do you think that you have enough information to claim anything about how frequent the system fails are that’d actually be statistically relevant?
It’s nothing personal, but you make CF sound like a totally dysfunctional site, while the only problem it’s had lately is occasional slow systests or rating updates.
But it was the same with TopCoder - when I started, there was limit for 1800 participants in SRM (maybe less) and always when it was full, they increased the limit by 50. Today the the limit is something like 2200, but there were technical problems almost everytime when they increased the number of prticipants…