Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[MagmaMcFry]

MagmaMcFry: Try doing the 3D thing again with a smoother heightmap! By smooth, I mean really smooth. Like the initial displacement being less than 1.

Er, which one? The green thing or the moon thing? And what do you mean by really smooth? I can tweak resolution, change the exponent, blur the heightmap, or simply decrease the altitude multiplier. Which of those do you want to see?

[Soadreqm]

If you're a new programmer, it's not a force of habit. If I see you do it one more time, I'm going to post in 72 point font not to do it again.

Hey, at least they're not strings.

Well, booleans are like this:

false = false

true = true

Apparently, in at least some versions of Python, you can set True = False.

[MadocComadrin]

I'm surprised at how common it is to make values that represent true/false (I.E a boolean) integers for new programmers mostly because I started with that...

I wonder if his professor was majorly a C programmer?

[Skyrunner]

MagmaMcFry: Try doing the 3D thing again with a smoother heightmap! By smooth, I mean really smooth. Like the initial displacement being less than 1.

Er, which one? The green thing or the moon thing? And what do you mean by really smooth? I can tweak resolution, change the exponent, blur the heightmap, or simply decrease the altitude multiplier. Which of those do you want to see?

I thought diamond square worked with a displacement value each step. :S
Just make the max heights and the min heights have a smaller difference, see if it looks less like an asteroid, re the moon thingy.

I'm surprised at how common it is to make values that represent true/false (I.E a boolean) integers for new programmers mostly because I started with that...

I wonder if his professor was majorly a C programmer?

Doesn't C have true and false too?

[MagmaMcFry]

MagmaMcFry: Try doing the 3D thing again with a smoother heightmap! By smooth, I mean really smooth. Like the initial displacement being less than 1.

Er, which one? The green thing or the moon thing? And what do you mean by really smooth? I can tweak resolution, change the exponent, blur the heightmap, or simply decrease the altitude multiplier. Which of those do you want to see?

I thought diamond square worked with a displacement value each step. :S
Just make the max heights and the min heights have a smaller difference, see if it looks less like an asteroid, re the moon thingy.

Spoiler: SCREENSHOTS (click to show/hide)

Displacement factor fifth'd.


No displacement at all.


Cranked down resolution.

I'm surprised at how common it is to make values that represent true/false (I.E a boolean) integers for new programmers mostly because I started with that...

I wonder if his professor was majorly a C programmer?

Doesn't C have true and false too?

Nope, true and false are C++ keywords.

[Skyrunner]

Question:

Which one would be faster for large (>=1 GB) data?

1. Reading and writing in one or more threads, and encrypting the data in another thread (at the same time), in small blocks of memory.
:: reasoning: page faults prevent one from using large amounts of RAM. Multithreading lets a user to read, encrypt, and write at the same time.
2. Reading large blocks of memory into RAM (0.3 GB ~ 1 GB), using multiple threads to encrypt it, then outputting it again.
:: reasoning: RAM I/O is orders of magnitude faster than hard drive I/O. Multithreading and disk I/O also are more efficient in large batches.

I think the operation being done is block encryption.

Someone asked a question and me and a different person have different answers XD

[Siquo]

Question:

Which one would be faster for large (>=1 GB) data?

1. Reading and writing in one or more threads, and encrypting the data in another thread (at the same time), in small blocks of memory.
:: reasoning: page faults prevent one from using large amounts of RAM. Multithreading lets a user to read, encrypt, and write at the same time.
2. Reading large blocks of memory into RAM (0.3 GB ~ 1 GB), using multiple threads to encrypt it, then outputting it again.
:: reasoning: RAM I/O is orders of magnitude faster than hard drive I/O. Multithreading and disk I/O also are more efficient in large batches.

I think the operation being done is block encryption.

Someone asked a question and me and a different person have different answers XD

Probably #2 for speed, disk IO is serial anyway, having multiple processes access it won't speed up the reading. If you have small amounts of RAM (on a phone or something) you might want to go for #1 as it seems more robust.

[GlyphGryph]

I kind of miss programming like this:

Code: [Select]

go :- hypothesize(Animal),
      write('I guess that the animal is: '),
      write(Animal),
      nl,
      undo.

hypothesize(cheetah)   :- cheetah, !.
hypothesize(tiger)     :- tiger, !.
hypothesize(giraffe)   :- giraffe, !.
hypothesize(zebra)     :- zebra, !.
hypothesize(ostrich)   :- ostrich, !.
hypothesize(penguin)   :- penguin, !.
hypothesize(albatross) :- albatross, !.
hypothesize(unknown).

cheetah :- mammal,
           carnivore,
           verify(has_tawny_color),
           verify(has_dark_spots).
tiger :- mammal, 
         carnivore,
         verify(has_tawny_color),
         verify(has_black_stripes).
giraffe :- ungulate,
           verify(has_long_neck),
           verify(has_long_legs).
zebra :- ungulate, 
         verify(has_black_stripes).

ostrich :- bird, 
           verify(does_not_fly),
           verify(has_long_neck).
penguin :- bird,
           verify(does_not_fly),
           verify(swims),
           verify(is_black_and_white).
albatross :- bird,
             verify(appears_in_story_Ancient_Mariner),
             verify(flys_well).

mammal    :- verify(has_hair), !.
mammal    :- verify(gives_milk).
bird      :- verify(has_feathers), !.
bird      :- verify(flys),
             verify(lays_eggs).
carnivore :- verify(eats_meat), !.
carnivore :- verify(has_pointed_teeth),
             verify(has_claws),
             verify(has_forward_eyes).
ungulate :- mammal,
            verify(has_hooves), !.
ungulate :- mammal,
            verify(chews_cud).

ask(Question) :-
    write('Does the animal have the following attribute: '),
    write(Question),
    write('? '),
    read(Response),
    nl,
    ( (Response == yes ; Response == y)
      ->
       assert(yes(Question)) ;
       assert(no(Question)), fail).

:- dynamic yes/1,no/1.

verify(S) :-
   (yes(S)
    ->
    true ;
    (no(S)
     ->
     fail ;
     ask(S))).

undo :- retract(yes(_)),fail.
undo :- retract(no(_)),fail.
undo.


[Nadaka]

Question:

Which one would be faster for large (>=1 GB) data?

1. Reading and writing in one or more threads, and encrypting the data in another thread (at the same time), in small blocks of memory.
:: reasoning: page faults prevent one from using large amounts of RAM. Multithreading lets a user to read, encrypt, and write at the same time.
2. Reading large blocks of memory into RAM (0.3 GB ~ 1 GB), using multiple threads to encrypt it, then outputting it again.
:: reasoning: RAM I/O is orders of magnitude faster than hard drive I/O. Multithreading and disk I/O also are more efficient in large batches.

I think the operation being done is block encryption.

Someone asked a question and me and a different person have different answers XD

Probably #2 for speed, disk IO is serial anyway, having multiple processes access it won't speed up the reading. If you have small amounts of RAM (on a phone or something) you might want to go for #1 as it seems more robust.

Disk IO for most hardware actually does benefit from multi-threading if you are dealing with many files. Reading/writing a single file is unlikely to use your disks/controllers full bandwidth. You can see a x2 to x4 read/write performance increase with 8 threads. But I wouldn't go over 8 threads.

[MadocComadrin]

I kind of miss programming like this:
..snip...

What I like about Prolog is that the more you try to force it to be more procedural, the more the word "fail" appears. XD

[Valid_Dark]

So, I started a new project, well not really a *new* project,  I was helping a guy work on a game 2-3 years ago, but back then I was doing 3D models for it,  got in contact with him and he wanted me to make a few more models for it, but saw that he really needed a programmer, so I offered to do the programming instead.
So I'm putting my current/old project on hold until I help finish this, because not only should it make me some money (hopefully) but when I finish it'll be a good little something in my portfolio.

It's on the source engine; which I know little about, and I have to use Visual Studio which I've never used before.

Finally got it all set up, and wow, the Source engine is really a clusterfuck, there's just so much to it.  I'm sure I can figure this all out though, doesn't seem to hard, just looks a bit daunting at first because of its size. 

right now I'm working on the xp/level system, then after that is the inventory system and saving the character to a file, all which should be rather easy-ish.  The real hard part is I might have to redesign the combat, or have someone to help me with that, because the combat in game at the moment is all placeholder.

Spoiler (click to show/hide)

The game looks pretty good, and if I can get gameplay on par with how good it looks, it should do pretty decent on steam.

[Skyrunner]

Code: [Select]

1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(2911): error C2065: '__CREATE_IMAGE2D_ERR' : undeclared identifier
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(2965): error C2065: '__CREATE_GL_TEXTURE_2D_ERR' : undeclared identifier
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(3107): error C2065: '__CREATE_IMAGE3D_ERR' : undeclared identifier
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(3160): error C2065: '__CREATE_GL_TEXTURE_3D_ERR' : undeclared identifier
1>gpgpu_virtualPipes.cpp(166): warning C4018: '<' : signed/unsigned mismatch
1>  main.cpp
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(2911): error C2065: '__CREATE_IMAGE2D_ERR' : undeclared identifier
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(2965): error C2065: '__CREATE_GL_TEXTURE_2D_ERR' : undeclared identifier
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(3107): error C2065: '__CREATE_IMAGE3D_ERR' : undeclared identifier
1>C:\Program Files (x86)\AMD_APP\include\CL/cl.hpp(3160): error C2065: '__CREATE_GL_TEXTURE_3D_ERR' : undeclared identifier

A set of errors from a simple

#define CL_VERSION_1_0                              1
#undef CL_VERSION_1_1
#undef CL_VERSION_1_2

I am continuing my efforts to make my program portable, but they are foiled yet again!

[olemars]

That sucks 
I guess that leaves you with two options then:
1. Find and install an older version of the APP SDK, or an alternative SDK like Intel's.
2. Forgo the C++ wrapper and use the C api directly.

Technically there's a third option: Telling compatibility to sod off and lock your application to 1.2 and the hardware that supports it, but that's not very helpful if you actually need this to run on other people's computers.

[Skyrunner]

My own PC can only run 1.1
The weird thing is it runs OpenCL without a hiccup. Also, no mention of clRetainDevice in the error messages. Maybe I should go ask stackoverflow? :S

[Mego]

I'm surprised at how common it is to make values that represent true/false (I.E a boolean) integers for new programmers mostly because I started with that...

I wonder if his professor was majorly a C programmer?

Doesn't C have true and false too?

Nope, true and false are C++ keywords.

Well... It does, actually. C99 introduced the stdbool.h header, which defines bool, true, and false.