Clash of Code : Bugs & Improvement

Feedback & Bugs
#81

Has anyone noticed that sometimes the 5 minute clashes will actually take 10 minutes?

#82

Another improvement: If someone tries to start a new clash and there is already a waiting room for that clash, they should be presented with the opportunity to join the existing waiting room before creating another waiting room.

#83

The ā€œ5 minutesā€ is an estimation, you might need less or more time. Thatā€™s why we have added more time.

#84

Regarding the clashes that never ends, we have started to investigate and we have found a minor issue that has always existed. However, we do not know why the number of errors have increased and it seems to be back to normal nowā€¦ We will patch our server but this might take some time.

For clashes that finally terminate after a while, I am interested to know how long it takes. If itā€™s less that 30 seconds after the end of the clash, it is not surprising.

#85

When someone starts a new clash, the timer is set to 5 minutesā€¦
I have noticed that many people, start a clash and do not want to wait so long, they abort their clashā€¦
Perhaps it could be a good idea to reduce that timeā€¦
2 minutes seems to me a good timing.
What do you think about it?

#86

In case of the browser has been closed for some reason while a user was clashing, :worried:
I suggest to add a button ā€˜Join the clashā€™ in the clashes displayed at the right side of ā€˜Clash Of Codeā€™ section, that is in case the user is present among participants. :wink:

This happened to me, and i couldnā€™t join the clash again, all what i can see is to view the report which shows that iā€™m still clashing which is not the case.
This feature is available in worldcup & multiplayer contests.

Thanks :blush:

1 Like
#87

I have the feeling that it takes sometimes several hours.
Iā€™m not sure at 100%, but I believe that these results took more than 12 hours to appear.
Maybe you have logs to double-check this.

#88

Thanks, Iā€™ll have a look to it.

#89

Il like Clash of Code. Itā€™s funny. So after several games, the fatest drag and drop wins. Iā€™m not Lucky Luck just Rantanplan :smile:

#90

Here is another one, where the results are still not complete after more than 10 hours: it says ā€œEn cours de clashā€¦ā€ for the last player.

#91

Thank you, I will work on this issue asap. We were a bit busy with the last release and the preparation of the next contest but I hope this will be addressed for the next release.

Feel free to ping me on IRC.

#92

Iā€™m sure itā€™s been mentioned before, but Iā€™d love to have some more feedback on my code. Iā€™m on the website to improve my coding skills, I want to know why my code got 80% even though 100% of the test cases back. Iā€™m completely new to the golfing challenges, so I have no idea what is done to get the code so short by other people. Maybe some tutorials or a cheat sheet type of thing?

To stop advantages in certain languages for some programs, I think having users fill in what programming languages they can do would be good. In CoC, if all users have a common programming language, the challenge will be in that programming language. Otherwise, itā€™ll be any of them as normal. Probably wouldnā€™t result in many specific language questions, but hey.

Also, a possible idea for the stopping of the copy/paste. If someone completes a challenge in a certain language, they cannot use the same language if they get the same challenge. Will probably have some problems with people not being able to program in any of the languages left, or actually running out of languages, but a system like that. In any case, the copy/pasting needs to stop.

1 Like
#93

Sadly, this solution has the same flaw as imposing that people can only compete once for each puzzle statement and challenge type (but can have the same puzzle twice for two different challenge types): people can perfectly, after the clash, spend time to develop a good golfing solution or a solution in a different language, and to copy/paste them the next time they see that puzzle.

#94

There is a chance that it may be mentioned before but i think that a sound to notify the start of a clash may be usefull

3 Likes
#95

And even without having seen a particular CoC before you may have seen something similar on the solo puzzles for instance, there is a CGX Formatter CoC and a Size of Lake CoC for example. I canā€™t imagine a flawless Clash Of Code, the only thing that seems to make sense is to reduce the incredible amount of points it gives to reduce the moral hazard.

#96

I have not seen the CGX formatter one, but I have met the Size of lake one. I had solved Surface before, but I found it faster to just quickly rewrite a solution than dig back into the specification of the larger puzzle.

After some time, more or less all problems (be it games or real-world problems) that one faces look like something weā€™ve done earlier, and one can find some similar code to copy, paste and adapt, or even recode it faster than if it were the first time.

Itā€™s just like an experienced chess player storing countless games in his memory to analyse the current one, somehow. Thatā€™s part of the game in my opinion, and not really anti-competitive.

The problem resides when people donā€™t need to recode the thing or adapt existing code. Then it just becomes a game of who finds the relevant clash or pastes it fastest.

#97

I am loving this feature, thank you!

How can I view the code that others wrote after the clash has finished? If this is not possible then I would like to add to requested features.

By the way, CoC stands for Clash of Clans - popular iOS game. Make sure you donā€™t get sued for using their nameā€¦

1 Like
#98

Iā€™d love to play those mods

#99

I have many suggestions:

  • Records kept for each challenge, which are viewable after the challenge is over. For example, for shortest code I would like to see the shortest length of code ever submitted (not the actual code). For others I would like to see fastest submission times.

  • Ability to chose clash type.

  • Each challenge should be novel. If youā€™ve seen it before then you can simply copy+paste from a local library of solutions.

  • Clash report should say what the clash type is.

  • You shouldnā€™t be able to invite people who are already clashing

  • Chat between clash participants during the clash

  • Global clash ranking filtered by language of choice

  • When I hover over a person I should see their clash ranking, not their overall ranking.

  • Shortest code contests should not count whitespace characters and comments. Deleting whitespace is a real time sink for me. Also you should be ranked within your own language.

  • Execution speed would be an interesting contest too. It doesnā€™t matter that some languages are faster - we have the same problem with some languages being more terse in shortest code contests. If people can choose their language to use then no problem.

  • Finally you can copy HackerRank and have contests for bug finding.

1 Like
#100

[quote=ā€œdimkadimon, post:99, topic:833ā€]

  • When I hover over a person I should see their clash ranking, not their overall ranking.[/quote]
    Yes, please!

In early optimization games, people were able to abuse this by encoding their code as whitespaces. So not really doable.

If you are the only one with your language, how would you compare to the other clashers?

Already suggested too, iirc the answer is that it is technically difficult to get consistent measures of execution time at the moment.

2 Likes