Usaco

[Sane]

Yeah. And I find IOI freaking difficult too. xD

Those problems are just at a whole nother level. I never know how to deal with numbers that large. Well, time to get back to USACO training. Rob fixed the Castle problem.

[klopyrev]

long long is your friend.

[Sane]

Ohshi. Damn Dev-C++. I didn't know long long could go up to 20 digits, because Dev-C++ doesn't support it.

I had been trying to do all those questions by manipulating char arrays. =.=

[Saad]

Sane @ Sun Apr 27, 2008 6:40 pm wrote:

Ohshi. Damn Dev-C++. I didn't know long long could go up to 20 digits, because Dev-C++ doesn't support it.

I had been trying to do all those questions by manipulating char arrays. =.=

Dev-C++ is not a compiler its an IDE. It uses the MinGW compiler which does support long long. Furthermore your forgetting literal constants. IE 1234LL is 1234 in long long format.

[thegoose]

klopyrev @ Sun Apr 27, 2008 5:06 pm wrote:

I don't think COI is anymore difficult than IOI. I think this was actually a particularly easy COCI contest. I didn't actually participate in it. I was too lazy.

Stop being lazy.....^_^

I'd say the difficulty of this COI is the geometric mean between CCC and the hardest of IOIs. I was a bit surprised that none of the COCI's trademark hard problems were on this COI.

[Sane]

Saad @ Sun Apr 27, 2008 6:51 pm wrote:

Sane @ Sun Apr 27, 2008 6:40 pm wrote:

Ohshi. Damn Dev-C++. I didn't know long long could go up to 20 digits, because Dev-C++ doesn't support it.

I had been trying to do all those questions by manipulating char arrays. =.=

Dev-C++ is not a compiler its an IDE. It uses the MinGW compiler which does support long long. Furthermore your forgetting literal constants. IE 1234LL is 1234 in long long format.

It does support it, but it's still bugged.

C:

#include <stdio.h>                                    typedef unsigned long long Long; int main(void)                                        {                                                         Long juul = 123456789012345678LL;       unsigned char *pjuul;                                 int i;                                                                                                       printf("sizeof(Long) = %d\n", sizeof (Long));    printf("juuld: %lld\n",juul);    printf("juulx: %016llX\n",juul);                         pjuul = (char *) &juul;                               printf("juulb: ");                                     for (i = 0;i < sizeof(Long); i++) printf("%02X",*(pjuul + i));    printf("\n");                                     return 0; }

I get:

C:

sizeof(Long) = 8 juuld: - 1506741426 juulx: 00000000A630F34E juulb: 4EF330A64B9BB601

Many people have found mingw's support for long long to be bugged.

The fix is to use:

C:

%I64d %016I64X

Which works. But the Dev-C++ documentation and C standard says to use %lld, so I'm not sure how portable that fix would be.

[Saad]

Sane @ Mon Apr 28, 2008 4:32 am wrote:

Saad @ Sun Apr 27, 2008 6:51 pm wrote:

Sane @ Sun Apr 27, 2008 6:40 pm wrote:

Ohshi. Damn Dev-C++. I didn't know long long could go up to 20 digits, because Dev-C++ doesn't support it.

I had been trying to do all those questions by manipulating char arrays. =.=

Dev-C++ is not a compiler its an IDE. It uses the MinGW compiler which does support long long. Furthermore your forgetting literal constants. IE 1234LL is 1234 in long long format.

It does support it, but it's still bugged.

C:

#include <stdio.h>                                    typedef unsigned long long Long; int main(void)                                        {                                                         Long juul = 123456789012345678LL;       unsigned char *pjuul;                                 int i;                                                                                                       printf("sizeof(Long) = %d\n", sizeof (Long));    printf("juuld: %lld\n",juul);    printf("juulx: %016llX\n",juul);                         pjuul = (char *) &juul;                               printf("juulb: ");                                     for (i = 0;i < sizeof(Long); i++) printf("%02X",*(pjuul + i));    printf("\n");                                     return 0; }

I get:

C:

sizeof(Long) = 8 juuld: - 1506741426 juulx: 00000000A630F34E juulb: 4EF330A64B9BB601

Many people have found mingw's support for long long to be bugged.

The fix is to use:

C:

%I64d %016I64X

Which works. But the Dev-C++ documentation and C standard says to use %lld, so I'm not sure how portable that fix would be.

Actually this bug would be windows related (specifically msvcrt.dll). However a simple #define and #ifdef could ensure portability.
More information here

On a smaller note, are you coding in C or C++?.

Sorry for going a little off topic

[Sane]

Usually C.

[Sane]

A.J, I found Camelot a lot easier than a lot of other chapter 3 problems. Camelot took me 2 tries to get working (forgot to handle the case where there are no knights). Some other problems like Humble Numbers took me a whole day to get working. Do you think Camelot was really the hardest one in chapter 3?

P.S. The most annoying part about that question was parsing the input in C. Since as far as I know, you can't use scanf if you want to stop at the EOF. I had to use getchar() and manually parse it.

[A.J]

k, I am almost done USACO training....

and does anyone know whats the deal with the 'new' training pages? what is it about and how does one get access to it?

[saltpro15]

Well, about 2 months ago i started hearing about how great the USACO training is for DWITE, so I thought I'd check it out. So after a month and a half learning C++, I have 1 question done -.- I understand how to get the answer, I just don't know enough C++ to code it

[phreadj]

A.J @ Sat Jan 31, 2009 11:15 pm wrote:

k, I am almost done USACO training....

and does anyone know whats the deal with the 'new' training pages? what is it about and how does one get access to it?

you can easily make an account (possibly even the same as your contest account, not sure)
and it is sort of like a contest which never ends. once you look at the question, the timer starts and it doesnt stop
until you submit the right solution. The questions get harder as you do more, it is good practice

[A.J]

nice

I am at the end of chapter 5 of the normal training pages

if I finish it, then I'll look into the new ones for sure. thanks

[SJ]

phreadj @ Sun Mar 29, 2009 6:23 pm wrote:

A.J @ Sat Jan 31, 2009 11:15 pm wrote:

k, I am almost done USACO training....

and does anyone know whats the deal with the 'new' training pages? what is it about and how does one get access to it?

you can easily make an account (possibly even the same as your contest account, not sure)
and it is sort of like a contest which never ends. once you look at the question, the timer starts and it doesnt stop
until you submit the right solution. The questions get harder as you do more, it is good practice

I thought the new training problems are "invitation only"? are you referring to ace.delos.com/traingate