JOIN
Get Time
forums   
Search | Watch Thread  |  My Post History  |  My Watches  |  User Settings
View: Flat (newest first)  | Threaded  | Tree
Previous Thread  |  Next Thread
[ 1 2 3 ]    NEXT >
ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? | Reply
Manila Regional Contest 2009 has ended and I heard that 6 teams got full score (solved all 10/10 problems). I'm wondering that the contest must be like coding speed contest rather than algorithmic. What do you guys think? I only knew there was a Guidelines for Producing
a Programming-Contest Problem Set
. Should this guidelines enforced from now on? For example, by having ICPC "verification teams"?

Talking about "problemset verifications", Several months before ACM ICPC Regional Jakarta 2009, a person claimed as Rujia Liu (rujialiu.srbga@gmail.com) contacted BINUS to "verify" Jakarta 2009 problemsets. AFAIK, Rujia Liu's email is rujia.liu@gmail.com. I was thinking that this was a scam email and ignored it. The email seemed convincing though by cc-ing Prof C.J. Hwang's email (cjhwang@acm.org). Did anybody who host regional contest received similar emails?
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
http://raffysaldana.blogspot.com/

From the result above, there are 2 teams with 10 problems.
The 1st rank local team solved all in 119 (one shot one AC).

Edit: The blog's updated. 6 teams got 10.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by dungcs) | Reply
Is that really the final rank?
Because I cannot see any team solve anything after 240 mins

and the final rank should not be released yet.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
Try the Greater New York region problems. 8 teams (possibly 9, I didn't stay for the awards ceremony) finished the 9 problems (including my team, Cooper 2). Standings/solutions/data are not up yet, but see how easy those problem statements are (a next permutation problem (E) when almost everyone uses c++...yeah that'll stump them...)! I had three problems correct by myself in 12 minutes. The first one has always been easy because organizers wanted everyone to be able to solve at least one problem, but this was truly ridiculous! And all teams had 2 computers to boot. I think the top 2 teams finished under 2 hrs, mostly because G and H can take a bit of time to implement.

The test data will probably be a lot easier than it could (should?) have been, especially for the Running Median problem (in the contest, the naive n*2 * log(n) sort_every_iteration solution passed instantly though n = 9999 for enough cases would have been enough to force that to time out).

Too bad I got fixated on using bad convex hull code early instead of doing other problems...
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by dungcs) | Reply
Are the problems available?
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by dungcs) | Reply
The blog about the result has been removed?
I didn't see any result there...
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
Well, I see it...

I think this is a very bad problemset btw. Even if 1 team solves all problems, it's probably too easy, but more than 1 team solving all problems is very bad in my opinion.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by narri) | Reply
Well... from my experience, all US semi-finals are like that. We used a couple of their problemsets to practice before a final, and the main problem was usually to parse good-looking input and print good-looking output. Or, sometimes, simulate some tedious process with hundred of rules.
Although, I was always puzzled why.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
The mail of rujia liu was a scam but I was almost fooled.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by broken_arrow) | Reply
Ic... So it was indeed a scam... It's fortunate that I took a security module and since then I've been very sensitive to security issues :D. I'd like to share the email that was sent to BINUS (Lomba TI). Hopefully, nobody fall/fell to this scam:

From: Rujia Liu [rujialiu.srbga@gmail.com]
Sent: Monday, September 14, 2009 12:41 AM
To: Lomba TI
Cc: cjhwang@acm.org
Subject: Problems and data verification request (Jakarta site)

To whom it may concern,

Since the date of the Jakarta regional contest is nearby, the asia regional contest director, Prof. C. J. Hwang, told me that I should contact you, the chief judge of the contest about whether you would like us to help you verify the problem descriptions and data. The problems and data will be released to ICPC Official website (http://icpc.baylor.edu<http://icpc.baylor.edu/>) and CII Online Judge Archive (http://cii-judge.baylor.edu<http://cii-judge.baylor.edu/>) if you accept our request. However, due to the fact that the problems and data sets may have some minor problems, we would like to help you verify the test data.
If you accept our request, please send the problems and data into a zip or rar archive and send it back to me.

PS: (To Prof C. J. Hwang) This letter is also forwarded to you for your information.

Thanks and regards,

Rujia Liu

The email looked convincing, right? My basic intuition was to check whether this mail was a spoof? Like the one happened to Bartholomew Furrow in GCJ09 milist (this incident made me worry trusting anything from Gmail or Google Groups). Is there any Googlers in this thread? Can you check whether the spoofing issue has been fixed?

Knowing the spoofing issue, first I need to verify the sender address so I googled the sender email address and Google couldn't find anything. This raises my security alarm! Then I googled "Rujia Liu gmail" and found several results that says his email is "rujia.liu@gmail.com" and later I found his paper about The UVa Online Judge Experience (now the email address in such paper is a lot more trustworthy). So the lousy scammer didn't try to spoof Rujia's email after all :P Googlers out there, can you help identifying who is behind rujialiu.srbga@gmail.com ?
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
But there is a cc to Asia Contest Director. I wonder if any attacker will give a cc like that. Did the Asia Contest Director say anything about it?

May be Rujia Liu had an alternative email address just for all these testing purpose, so that his usual account doesn't get flooded.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
Do you think it is normal for a team to get all problems correct, each in one submission, in about half of the contest time?

I heard that there was a problem with wrong examples, and there were some clarifications after that. But the first ranked team has solved that problem before the clarification.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by rajeshsr) | Reply
It was definitely a scam, and after it was identified as such, all contest directors got an email warning as far as I know. It seems that the email was not send by cc to the real email address of Prof. Hwang.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
The penalty of the top team is 606 minutes after solving 10 problems, which does not seem natural when the others have twice penalty time and those other universities are well reputed contest strengths. I hope this has nothing to do with the contest director being fooled by Rujia Liu scam type email and handing over the data to a participating team :-P. I would love to be proved wrong but everything sounds fishy to me now.
Re: ICPC Regional Manila 2009 problems were too Easy? Regional problemset verifications? (response to post by felix_halim) | Reply
I am Rujia Liu. Just like Felix has found (by googling), I only use rujia.liu@gmail.com, not rujialiu.srbga@gmail.com. That fake email was sent long ago. I first realized this trick when a problemsetter of regional in China asked me when is the "deadline" for sending testdata for verification. Then I phoned the professor who's in charge of the problemset immediately, and he contacted Prof. Hwang and other RCDs in China.

A few days later, Prof. Hwang sent emails to various related people (e.g. RCDs outside China) clarifying this, as Adrian pointed out. But until now, I've no idea about who wrote that email (though I'm very angry).

About Manila: I was very busy with my company and did not notice the strange result of Manila until someone told me. I still haven't read the problemset but the result looks ridiculous... Though I myself do not have time verifying problemsets (due to the 'terrible' company I have), enforcing a good guideline seems a good idea.
[ 1 2 3 ]    NEXT >

RSS