Sticky

Post general comments here

by Alex on Apr 28, 2015 - 11:48:46 pm UTC
  • (3/0)
This is the place to post general comments, such as pressing bugs or broader questions.

Comments (Search)

I know it'd take a lot of time and code to implement these, so I know I won't be seeing anything soon, but I think the Judge'll get a lot better if these things happen:

1. Link PegJudge Accounts to Forum -> Will increase activity on forum. Necessary for 2.

2. Perhaps implement the 'Comments' section into a forum like topic? Might help with sorting the random walls of text and posts we have now. Also might separate some *decent pieces of information from the junk like the million '?' post.

3. A search feature for comments/site, not just problem names. It'll help with looking for the one hint comment which really helps with a different problem.

Again, I can't make any of these things happen by myself, so I have no right complain much, but I believe these will make the Judge much more effective.

I'm inclined to doubt that linking Judge accounts to the forum would increase forum activity. I could probably accomplish it if I set aside some time to learn about cookies, but I don't see the point right now.

I'll start on #3 though. EDIT: comment search is implemented

Thanks.
Comment search seems to be working fine.
I hope it didn't cause another security leak like the logged-off analyses thing.

Might I ask how comments would cause a 'security leak?'

I accidentally created a major security hole (allowing users to see other users' code when not logged on) when I was adding some features to the backend. This is a reference to that. Luckily, the code for comment search is mostly copied from Hanson's code for problem search, so I don't think any bugs could have been introduced.

I did recently create another bug though, which broke Java entirely. >.<

- First, I'm really impressed with the analysis of solved problems which many other OJs don't have.
- However, what keeps the members (or just me )most of the time is the ranking. The points allocated are quite random I supposed and this should be changed :)

We try to assign point value based on difficulty. Is there some other criterion you would like considered?

Maybe basing it on how long it takes to do the program as well?

Not sure whether you mean thinking time or coding time. Thinking time is highly variable; some people are "sprinters" and either see the algorithm quickly or don't at all; others are "long-distance runners" and spend a long time thinking about a problem, but usually get it eventually. All I can do is estimate how hard the problem is, based on how advanced the techniques required are. (e.g., Does a greedy algorithm work, or do you have to use dynamic programming?) We do also take into account how detailed a solution has to be to be correct, which is reflected in how long it takes to code; take for example the problem Landing from CCC '08 stage 2; the algorithm involved is a completely standard algorithm, but it is very tricky to implement, which is why that problem is worth 50 points instead of, say, 15 (which is a typical value for an "easy" advanced algorithm). You can always propose that a point value be changed if you feel it is inappropriate; I will always consider suggestions, and possibly ask for input from other community members.

oh yeah? wat about 'Hard Question'? It should be worth at least 5 points! (given that you take your algorithms away)

Hard Question is not hard at all; you just write down an equation and find a formula to solve it, and then you translate that to code. A strong math and science student should literally be able to finish it in 10 minutes. For this reason I cannot justify allocating more than 3 points.

To me, I believe that a hard question is one that few can solve . So, what I suggest is that let the points adjust themselves according to the number of people successfully solved it.

The link doesn't work XD. And yes, I would agree with basing a program's point value on length as well as difficulty. That would definitely encourage users to do longer questions. Not exactly a good thing but can you say that's bad?

I fixed the link. You should've posted this as a reply directly to that comment, though.

What!!! *Exasperated*
I did a big bunch of problems before, but they all disappeared. Like, now I just have that cross thing and it says that I got wrong for everything. And my answer just is a blank page.

How do you submit the acsl test questions?

No need to spam the question mark button.

If Joyce/Saravannan didn't tell you specifics, just email it to [email protected]

???
BETTER?

lolz still ???? spamming

It seems that the temporary 30pt questions (such as Waterpark) from last year have been moved down to 10 pts (on my own, and Daniel's accounts). Is that true, or is it just showing that, while actually being 30 pts on the total?

Yeah, they reverted to their original values. It makes sense; it's not really fair to disadvantage people who started later on the Judge, or something like that; this site isn't just for Woburn's CS classes anymore (although you guys still use it the most.)

Yeah, and most of us (except for Ammar) aren't in CS anymore.

Needs to be fixed to the 'Old' part of Peg judge.


Doesn't work on Chrome 9.0.517.16 Dev.
Says not all elements were secure so only goes to a red screen.

Why are you reporting a compatibility issue with a snapshot? I can't possibly be expected to know whether the problem lies with the browser or the site...

IE gives this:
Content was blocked because it was not signed by a valid security certificate.

For more information, see "Certificate Errors" in Internet Explorer Help.

and Firefox gives this:
This Connection is Untrusted

You have asked Firefox to connect
securely to 173.0.55.206, but we can't confirm that your connection is secure.

Normally, when you try to connect securely,
sites will present trusted identification to prove that you are
going to the right place. However, this site's identity can't be verified.

What Should I Do?

If you usually connect to
this site without problems, this error could mean that someone is
trying to impersonate the site, and you shouldn't continue.

i.e. (Pun intended) this isn't limited to one browser. I think it has something to do with the wcipeg.com domain not being attributed to the vps server. Anyhow, it's not working.

Yeah, the certificate is assigned to 173.0.55.207 right now. I'll issue one for wcipeg.com after we're finished the domain transfer.

For some reason, you can't log in to PEG Judge using IE. (Chrome works fine.)

IE 6 at least.
IE 7/8/Firefox works fine.

It works fine for me (IE 6 on Windows XP virtual machine). What are you talking about?

Okay, I discovered that I forgot to escape the URL string somewhere, which might have confused some browsers. Does it work now?

nope.

So what happens when you try to log in?

nothing. it works fine on safari. I have no problems whatsoever

I'm using IE 8 and it works perfectly. And I discovered that if I type in http://www.wcipeg.com/judge, it displays an error message.

use 173.0.55.207/main

you have ie8?
directed to sherwin

We're in the process of claiming the wcipeg.com domain; it won't work properly until the process is complete.

wat's our competition? and geerthan, you get to upgrade to ie 8 from ie 7.

At Woburn, it would simply return to the main page but would not identify as logged in. It affected all the computers in Room 211 so I don't think it was an isolated issue. (In hindsight, I didn't check if cookies were enabled.)

i just logged in on the server Saravannan and Joyce sent us and it works perfectly

I was doing a problem for a judge which had time limit exceeded.

When I switched my variable types from long long to int, it was faster and got accepted.

Why?

The int is 32 bits, which is the word size of the judge machine. Therefore, the machine can read, write, and calculate with 32-bit integers more quickly than with any other data size.