Makha's blog

System tests on virtual participation?
By Makha, history, 3 years ago, translation, In English

Mostly, people take a virtual participation to practice. In my opinion, it would not be bad for the solutions to be checked on pretests during the virtual round. And after the end of the virtual round, the system tests would start checking. I believe that in this way, people will benefit more from virtual participation.

What do you think about it?

MikeMirzayanov

  • Vote: I like it
  • +207
  • Vote: I do not like it

Comments Comments (21)
| Write comment?
»
3 years ago, # |
  Vote: I like it +19 Vote: I do not like it

The thing is that some people might not like that, so maybe an option, like "Use pretests/systests" that would be on by default (because I think many people, including me, will use it)

→ Reply
  • »
    »
    3 years ago, # ^ |
      Vote: I like it +11 Vote: I do not like it

    I think that this option should not be, because virtual round must be as similar as possible to real round.

    → Reply
»
3 years ago, # |
← Rev. 2   Vote: I like it +16 Vote: I do not like it

You are not alone who thinks the same.

Solution to this

→ Reply
»
3 years ago, # |
  Vote: I like it -67 Vote: I do not like it

Different candidates can take virtual test at different time, unlike the actual rounds. So, if Codeforces will try to do system test for vitual participations as well, it will put immense load on its server, which is somewhat unnecessary as of now.

→ Reply
  • »
    »
    3 years ago, # ^ |
      Vote: I like it -32 Vote: I do not like it

    If you were a red coder your comment would be upvoted.

    → Reply
    • »
      »
      »
      3 years ago, # ^ |
        Vote: I like it -19 Vote: I do not like it

      why did you downvote me?

      → Reply
    • »
      »
      »
      3 years ago, # ^ |
        Vote: I like it +12 Vote: I do not like it

      This comment isn’t being downvoted because it was posted by a gray—it’s being downvoted because the argument isn’t compelling. There’s no reason to believe that running some tests immediately and some tests later will significantly increase the load on the CF servers, which are not especially strained outside of official contest times anyway. There are a number of legitimate reasons not to pursue this update (most notably, that it trades off with other improvements the developers could make to the platform), but I don’t think this is one of them.

      → Reply
  • »
    »
    3 years ago, # ^ |
      Vote: I like it -37 Vote: I do not like it

    Here, no values are given to comments of Newbies and Pupils, as if everyone started as Redcoder only. Shame on coders of such mentality.

    → Reply
  • »
    »
    3 years ago, # ^ |
      Vote: I like it +20 Vote: I do not like it

    Doesn't codeforces already judge submissions on all tests during virtual round?

    → Reply
    • »
      »
      »
      3 years ago, # ^ |
        Vote: I like it 0 Vote: I do not like it

      I think sayand thought that every single submission to every single problem would be retested on every single test case for every single virtual that someone does.

      Also, it looks weird to see you without the rainbow name (if you know you know)

      → Reply
      • »
        »
        »
        »
        3 years ago, # ^ |
          Vote: I like it 0 Vote: I do not like it

        Well, every problem has X pretests and Y system tests. Where X is a subset of Y. But does the codeforces database even retain the historical information about how many of each problem's testcases used to be a part of its pretests? If not, then maybe it's a good idea to start tracking this information from now on?

        If such information is already available, then simulating the effects of passing pretests vs. passing system tests for each submission of each participant shouldn't have any extra overhead during virtual contests. If somebody's historical submission for some problem passed N testcases and N>=X, then display such submission as "passed pretests" in the scoreboard and award preliminary points. After the virtual contest is over, finalize the scoreboard to award points only for the solutions which have N==Y.

        But I guess, I'm stating the obvious.

        → Reply
        • »
          »
          »
          »
          »
          3 years ago, # ^ |
            Vote: I like it 0 Vote: I do not like it

          Good point. Maybe if they haven't, the first several systests could be chosen to be used as pretests, maybe the first $$$20\%$$$ or $$$25\%$$$ of tests.

          → Reply
»
3 years ago, # |
  Vote: I like it +2 Vote: I do not like it

Yes, something like in topcoder.

→ Reply
»
3 years ago, # |
  Vote: I like it 0 Vote: I do not like it

Lets just do a thing that would be helpful for the whole community. Just have sample tests == system tests. :clown:

→ Reply
»
3 years ago, # |
  Vote: I like it 0 Vote: I do not like it

Sad that MikeMirzayanov did not answer under this blog.

→ Reply
»
3 years ago, # |
← Rev. 2   Vote: I like it +11 Vote: I do not like it

While we're at it, I'd like being able not to see virtual participants in the standings

→ Reply