On Euler you're asked for the answer, not an implementation solving the problem. This means that you can solve the problem with an implementation that takes days to run or require some manual steps. On Kattis you submit source code to the system, so the runtime limit will be on the order of seconds (definitely not days).
Topcoder has "more interface". With Kattis you would typically code in whatever editor/IDE you have and prefer. Also, Topcoder asks you to implement a function, Kattis asks you to write an entire program. The most important (?) difference being that the latter includes I/O.
Im pretty sure that this is a public version of the platform that has been running at KTH (royal institute of technology) for years. It was used as a test suite as well as performance test for assignments there for the CS school.
There are many problem archives online, each with its own characteristics. Project Euler has math problems where answers are submitted via plaintext, TopCoder has its own flavor of problems with an ancient Java applet interface (although it has a new, still very primitive, web version in beta).
Compared to other problem archives, the Kattis Problem Archive has a very high ratio of quality/quantity. Also, the average problem statement on Kattis is more interesting than the average problem statement on TopCoder/Codeforces (e.g. "Fox Ciel just got this sequence of integers as a birthday present from his grandma, ...".)
This is a good time to bring up the fact that Advent of Code[1] will be returning this December. I had a blast working through it last year and can't wait for more.
At the time, I was a little surprised to see it not get upvoted more. OTOH it was kind of a slow period (Sunday evening IIRC). The original submission stayed for a few hours, then dropped off the frontpage with only a vote or two.
Which does make one curious, in what way could this occur? Had I bungled a planned HN launch? (and then maybe a mod helped out to resubmit). Or is there some other mechanism at work.
This is kind of creepy, I registered this morning and started practicing with the trivial exercises, I didn't fill anything special in the registration form besides the username, and now — around six hours later — my profile has been automatically updated with an association to my university and country (which I also didn't select when I signed up). I am not surprised by the country selection, they can easily get that information with a good GeoIP database, but I wonder who did they get the name of my university, maybe from LinkedIn since it is the only place that I am aware of contains that information. Kattis seems like a fun project nonetheless, but these two things scared me out a little bit.
17 comments
[ 2.9 ms ] story [ 47.8 ms ] threadTopcoder has "more interface". With Kattis you would typically code in whatever editor/IDE you have and prefer. Also, Topcoder asks you to implement a function, Kattis asks you to write an entire program. The most important (?) difference being that the latter includes I/O.
Compared to other problem archives, the Kattis Problem Archive has a very high ratio of quality/quantity. Also, the average problem statement on Kattis is more interesting than the average problem statement on TopCoder/Codeforces (e.g. "Fox Ciel just got this sequence of integers as a birthday present from his grandma, ...".)
1. http://adventofcode.com/
Which language did you want to use?
So here's something rather unusual (or not?): While HN says I submitted this story 6 hours ago, actually I submitted it 2 days ago.
A HN Algolia search confirms
https://hn.algolia.com/?query=kattis or http://imgur.com/a/O0pws
At the time, I was a little surprised to see it not get upvoted more. OTOH it was kind of a slow period (Sunday evening IIRC). The original submission stayed for a few hours, then dropped off the frontpage with only a vote or two.
Which does make one curious, in what way could this occur? Had I bungled a planned HN launch? (and then maybe a mod helped out to resubmit). Or is there some other mechanism at work.