Errors/Warnings/Bugs

[euler]

Thanks for letting me know. I'm currently trying to figure out why it hasn't renewed.
Hopefully I can get this sorted shortly.

[euler]

... it should be working now. It's not a nice experience arriving at a site and being told that the connection is not secure. Sorry for the disruption.

[radeye]

The statement for 814 is incorrect. The value for S(10) given is not the actual value (before taking the remainder modulo the prime) but the statement says it is. The value given for S(10) is *after* the modulo operation.

(This cost me way too much time.)

[bruce_love]

@Radeye: sorry for the inconvenience - it has now been changed to indicate that $S(10)$ is the value after taking the modulus.

[progheal]

In problem 816, the example answer for $d(10)$ is rounded wrong. The answer should be $\sqrt{85784932007125}\approx 9262015.54776955662$ (between $P_2(17939732,3034546)$ and $P_8(9561938,6983913)$), which rounds to $9262015.547769557$.

I may have a guess on where this inaccuracy comes from: The two nearest IEEE754 double precision floating point numbers to our number are 0x4161AA77F1875405 (9262015.54776955582201480865479) and 0x4161AA77F1875406 (9262015.54776955768465995788574). Our number is closer to the former than the latter, and rounded to that number; but it round to ...9556 when rounds to 9th decimal place. Basically this is a precision-not-enough error. This might not be a problem for 2 million points as points will be denser and the shortest distance will be smaller.

I found this error when using `long double` in the last square root step in my program.

If possible, I would suggest change it to $d(14)=\sqrt{298603741129}\approx 546446.4668464789333$ (between $P_{12}(18885138, 9652189)$ and $P_{13}(19047286, 9130354)$) which is one order of magnitude smaller, and gives room to the precision after decimal point: this value rounds to 0x4120AD1CEF06806F (546446.466846478986553847789764) and both rounds to 546446.466846479 for nine decimal places. This value retains until $d(121)$ so it is not necessary to use 14, any integer between 14 and 121 is ok.

[hk]

Thanks.
I went with your suggestion to use d(14)=546446.466846479.

[SirMrGuy]

I'm not sure if this is an issue elsewhere, but every so often my placement on the fastest solver table for 808 flips between third and fourth. I suspect this is because I submitted an answer within a second of someone else, but it's odd that it's inconsistent.

[euler]

Wow! I never expected that to happen. It turns out that you and neverforget solved it in exactly the same time, or at least as accurate as the Unix timestamp (to the nearest second) has recorded it. So each time the fastest solvers table is re-cached, I'm guessing that the DB engine just randomly shuffles matching times when it's asked to order by timestamp.

[pjt33]

It's probably not random: DB engines don't tend to randomise unless they're told it. I would guess that it's an artifact of the number of rows in the table. Wonder whether there's a PE question waiting to be written about unstable sorts and k-medians...

[SethTroisi]

I'm fairly sure there's a </tr> missing on the table header row on the fastest solver pages

Code: Select all

<table class="grid">
   <tr>
<th>&nbsp;</th>
<th class="username_column">User</th>
<th>&nbsp;</th>
<th class="language_column">Language</th>
<th class="time_column">Time To Solve</th>
[MISSING </tr> HERE]
<tr>
    <td><span class="smallest strong">1st</span></td>
     ....
</tr>

[euler]

That's really helpful. Thank you.

It should be fixed now. I've also taken the time to add some tabs and new line escape codes to the generating script to make it tidier and more readable.

[ETS1331]

My post to the most recent problem (822) does not seem to be permanent (according to the post tab in my statistics), despite it being the first post in the thread. I assume this is a bug? (not sure why it would affect just this one problem though.)

Edit: Checking the thread, it appears (at least to my account) that my post is the only non-permanent one on the thread.

[euler]

No idea how this happened. I'm not aware of it happening before either. Anyway I've marked it as permanent now. We'll have to keep an eye on things.

[Obergscheidle]

Dear Project Euler,

it looks as if the site is currently down.
I get a message

503 Service Unavailable
No server is available to handle this request.

Thanks a lot in advance for looking into it.

Best regards, Andreas

[KING-OLE]

Yes, it looks like the site has been down for 13.5 hours.

[euler]

I'm not sure why IsItDownRightNow is reporting 13 hours of outage. It's possible that some parts of the world have been experiencing issues, but I've been able to access the main (.net) site a couple of times in the past few hours and we have services running that monitor uptime and downtime. According to the server logs, the site has been inaccessible since 22:06:08 UTC. Our hosts are currently exploring the issue as it extends beyond projecteuler.net and is affecting part of their network. Hopefully they'll have it back up again soon.

[radeye]

Down for me.

[jaap]

It's working fine for me right now. I'm in the Netherlands, but it also still works if I switch my VPN on to the US.

[KING-OLE]

I can confirm that it is working from the US. Was able to post an answer around 11 PM Zulu Time.

[phatmo]

When i clicked on "Download solve history as CSV" at https://projecteuler.net/progress;show=history, the csv was not downloaded.
Instead, a page was opened with my history content, and it look like the content doubled.
Please check, thanks.