I agree that it is useless for now but that is the counter of codingame Achievement Points (CP=codingame Points). So for now there isn’t any achievement related to those games, but it will probably change
Add more comprehensive tests, possibly randomized. As of now, there is no way to know whether the people I compete with have written legitimate code for the problem as stated, or shortcut code that only cares about the current test suite.
Also, I submitted a cheated solution to the Thor problem, then submitted the old (fair) one again, but the leaderboards still show the cheated score; that’s probably not very fair.
Also +1 to allow to submit several entries in different languages.
Regarding the possibility to send several entries in different languages, this feature would also be great for the multiplayer contests. I’ve been tempted multiple times to create another account to participate both in my first language of choice for the ranking and in Bash for the fun.
I second the idea of a randomized test suite.
At some point the possibility of hardcoding the different solutions becomes too tempting, which spoils the fun in size coding.
you had graphs for rank, achievements, solo conests and multi contests
now you should add graph for optimisations, because rank points are not equal to sum of all other points
I can see the total optimization points but I can’t see by puzzle. My suggestion is to add how many points I obtained by puzzle.
How is it calculated?
In my opinion, it’s strongly weighted compared to the normal puzzles.
For some reason, in the puzzle’s screen the optimization position and the code size are not updating. In the leaderboard, it is ok.
I also have the impression that the optimization points are not updating, can’t confirm.
I agree with meira, it seems there’s an updating bug. Even in the leaderboard section.
I’m on the Power of Thor optimization puzzle. I first submited my old PHP code from the easy challenge which is 1091 cars long. Now I optimized it (229 cars), but when I’m trying to resubmit it, it seems it doen’t do anything: The leaderboard appears, I can see the report (100%) and the score (229), and my new rank (500~) but when I go back to the puzzle selection page, it’s still written that my best is 1091. Same in the leaderboard section, I’m still at 1091 cars.
Then I did it in Python, and it seems it was correctly saved.
EDIT: It seems the bug concerns only the 1st language used to solve the puzzle. With others it updates correctly
Just to confirm the updating bug.
For Thor(codesize), in the games’ page I see the rank of my very first submission, not the best one.
The same goes for the Thor’s leaderboard header line.
Clicking on the “VIEW” boutton in the header brings the report and code of this first submission.
(My actual rankings in the scrollable section of the leaderboard are OK.)
One more question:
It seems that solving Thor’s code size optimisation puzzle didn’t add any points to my account (ranking 118/1761).
Are these points computed at midnight?
If so, why? (seems like it could be easily updated on the spot, like for any other puzzle).
It’s nice not being alone
Yes, points for optimisation puzzle and multiplayer games progression are added at midnight only.
I have a suspicion that the only way to solve the Thor puzzle using 50-60 symbols is by a very clever hardcoding of the solution, which to be honest is an art in itself
Maybe, but it’s still really buggy.
The leaderboard is showing (top bar) my old position 88 with 238 characters, but my actual score is 58 with 182 characters. Just scrolling down I can see the correct position. It’s been days that I don’t change the code and it still shows the non-updated result.
The inputs in the Mars Landing optimization puzzle seem inconsistently wrong. Some tests swap X Y with R P and others have X Y in HS VS position. As far as I have tested none of them are in the order mentioned in the puzzle description.
I didn’t find any bug,but I didn’t understand your problem neither.
@SaiksyApo you appears twice on the leaderboard of Bilbo’s Quest, I think puzzle not based on code size like this one and mars landers should only preserve best score and not best score of each language.
I appears more than twice
That’s not my point
the point is it’s odd because once you have a code that make like 3300 instructions, why keeps you for writing the exact same code in another language?
To win some language achievements ofc (even if there are not implemented yet :D).
There is a little inconsistency in the character count for code size optimization puzzles.
UTF-8 characters are counted as one, except 4-bytes characters which are counted as two.
A (1 byte, \x41) counts as one
é (2 bytes, \xc3\xa9) counts as one
€ (3 bytes, \xe2\x82\xac) counts as one
𝄞 (4 bytes, \xf0\x9d\x84\x9e) counts as two
I don’t know which is the fairest counting method between bytes and characters, but certainly not a mix of both.