[QUOTE=Fourier;48610717]Berkin, for a second I thought you were a Nazi because 88 also means HH (H is 8 in a row), which stands for Heil Hitler. Then I googled piano 88 and saw that I thought wrong :v:.[/QUOTE]
Bit of a long stretch. Or did you think of that because I speak German? :v:
[editline]4th September 2015[/editline]
Something tells me I'm not doing this right...
Top is recorded voice, bottom is synthesized voice.
[img]http://i.imgur.com/BzAKWL4.png[/img]
Rant(TM) Quality Speech Synthesizer
I got huge Moon River vibes from that for some reason.
~moon river wider than a mile i'm crossing you in style some day~
[QUOTE=Map in a box;48611656]I got huge Moon River vibes from that for some reason.
~moon river wider than a mile i'm crossing you in style some day~[/QUOTE]
It sounds like that song from the original [I]Willy Wonka and the Chocolate Factory[/I] movie that plays during the edible plants scene.
[editline]4th September 2015[/editline]
[media]https://www.youtube.com/watch?v=r2pt2-F2j2g[/media]
Oh I know. It is. I just for some reason heard Moon River from it and got me thinking about Bayo 2 again.
At least fricatives are just noise patterns. Shouldn't be too hard to replicate with some band-pass filtering.
[img]http://i.imgur.com/XpJ28Bi.png[/img]
[editline]4th September 2015[/editline]
I managed to transform one fricative into another. This provides me with very useful data.
[vid]https://my.mixtape.moe/kiwxox.webm[/vid]
[QUOTE=DrDevil;48611012]He obviously must be a nazi![/QUOTE]
Close. I am not nazi but I have nazi friends and they tell all sort of (crazy) stuff.
[QUOTE=Fourier;48612053]Close. I am not nazi but I have nazi friends and they tell all sort of (crazy) stuff.[/QUOTE]
You associate with some interesting crowds.
[editline]4th September 2015[/editline]
[QUOTE=PelPix123;48611979]Ooh, speech stuff. I was working on a system that dynamically changes grunt line delivery:
[url]http://vocaroo.com/i/s1AsFWxUVWJ8[/url][/QUOTE]
Can I ask how you did this? That's pretty neat.
Well they are programmers and we were classmates, that is the only reason.
I also have friend that did meth and is math/programming genius, ... he helped me lots with higher math.
Anyway, back on topic: I am working on Dictionary attack.. because I forgot password... again. ... what a waste of time
[editline]4th September 2015[/editline]
So all memory fragments and combinations I tried right now are incorrect...
I think I am just gonna go to school and request reset... second time.. is this real life.
Really it's mesmerizing. I knew password for three days. Fourth day? Nope. Fuck you brains. Gonna write everything down now.
[QUOTE=Fourier;48603603]Bite a bullet and just do that damn PDF :v:. You must understand Unity doesn't want stuff with no documentation (who does?). If it is small script, everything should go to one A4 page.
You can do it with Microsoft Word (yes that Word) or maybe another open source stuff like OpenOffice.
With Word you can "Save As PDF", or if you cant, save it as DOCX and google "DOCX to PDF".. there are at least three free variants.
Or do HTML version and convert it to PDF. Never tried this route thou.
[url]http://pdfcrowd.com/[/url][/QUOTE]
Can't people on Windows just print to PDF? Is that not implemented in the basic drivers of the Windows printer system?
Finally got around to recreating my steam account batch creator.
It can make, verify email, and activate steamguard for about 10 accounts/minute right now. Around 90% of that time is in FINDING the damn email that Steam sends, so I'll probably do some tricksy stuff to cut down on that time.
[QUOTE=Fourier;48612174][...]
Anyway, back on topic: I am working on Dictionary attack.. because I forgot password... again. ... what a waste of time
[editline]4th September 2015[/editline]
So all memory fragments and combinations I tried right now are incorrect...
I think I am just gonna go to school and request reset... second time.. is this real life.
Really it's mesmerizing. I knew password for three days. Fourth day? Nope. Fuck you brains. Gonna write everything down now.[/QUOTE]
Since this happens to me somewhat regularly, I use [URL="http://keepass.info/download.html"]KeePass 2[/URL] to keep track of them. ([I]Professional Edition[/I] is free too.)
Now I only have to not-forget a few a bit more complex ones. It can also print the list in case you have a good way of storing that relatively securely.
[editline]4th September 2015[/editline]
[QUOTE=geel9;48612784]Finally got around to recreating my steam account batch creator.
It can make, verify email, and activate steamguard for about 10 accounts/minute right now. Around 90% of that time is in FINDING the damn email that Steam sends, so I'll probably do some tricksy stuff to cut down on that time.[/QUOTE]
Didn't they recently lock down abuse by players who never bought anything?
I suppose if you use it for scrap.tf you can just file that under business-expenses though.
[QUOTE=Tamschi;48612813]Since this happens to me somewhat regularly, I use [URL="http://keepass.info/download.html"]KeePass 2[/URL] to keep track of them. ([I]Professional Edition[/I] is free too.)
Now I only have to not-forget a few a bit more complex ones. It can also print the list in case you have a good way of storing that relatively securely.
[editline]4th September 2015[/editline]
Didn't they recently lock down abuse by players who never bought anything?
I suppose if you use it for scrap.tf you can just file that under business-expenses though.[/QUOTE]
For Scrap.TF's purposes we need an API key anyways so we just add a $20 steam wallet code to the accounts, buy 17 bp expanders (which we'd already need to do!) and other shit.
You could try connecting to the "new email!" notification that IMAP sends, that should arrive a little faster and only give you the new ones.
Though, you're probably already self-hosting those addresses and just get them delivered via SMTP.
The protocol is fairly straightforward so you likely could have them served directly into memory.
[QUOTE=mastersrp;48612711]Can't people on Windows just print to PDF? Is that not implemented in the basic drivers of the Windows printer system?[/QUOTE]
There are plenty of programs and whatnot you can install to print to a PDF.
'Printing' to a file format always seemed like a weird option to me, though. I guess it maximises compatibility with programs that don't have an option to save to PDF.
[QUOTE=BackwardSpy;48612886]There are plenty of programs and whatnot you can install to print to a PDF.
'Printing' to a file format always seemed like a weird option to me, though. I guess it maximises compatibility with programs that don't have an option to save to PDF.[/QUOTE]
That and PDF and PostScript aren't too dissimilar afaik, so the result is actually pretty good (for human consumption).
Back to GUI for cvision made some directX tabs and reskinned the cursor
[video=youtube;gz-HG-JPZxw]http://www.youtube.com/watch?v=gz-HG-JPZxw[/video]
[QUOTE=BackwardSpy;48612886]There are plenty of programs and whatnot you can install to print to a PDF.
'Printing' to a file format always seemed like a weird option to me, though. I guess it maximises compatibility with programs that don't have an option to save to PDF.[/QUOTE]
I thought Windows 7 had an XPS "printer", and 8+ an actual PDF "printer"?
adnzzzzZ wrote up his version of that [URL="http://gamasutra.com/blogs/AAdonaac/20150903/252889/Procedural_Dungeon_Generation_Algorithm.php"]dungeon generating tutorial on Gamasutra![/URL]
read a book at barnes and noble and now i get has tables. two things i apparently have to do:
make my table size p (prime number), and multiply all my pre-modulo hashes by some other (small) prime number. now i have to figure out how to expand the size of a table in C
[editline]4th September 2015[/editline]
Alright implemented a simple hash function and table
[code]#include <stdio.h>
#include <string.h>
#define TABLESIZE 137
int hash(char* key, int mod);
int main(int argc, char** argv){
int hashTable[TABLESIZE];
hashTable[hash(argv[1], TABLESIZE)] = 100;
printf("Key: \t%s\nHash:\t%d\nValue:\t%d\n", argv[1], hash(argv[1], TABLESIZE), hashTable[hash(argv[1], TABLESIZE)]);
}
int hash(char* key, int mod){
int hash = 0;
for(int i = 0; i < strlen(key); i++){
hash += key[i];
}
hash *= 31;
return hash % mod;
}
[/code]
That's just one method of making a hashmap. Use malloc/realloc or, less preferably, a linked list to resize your table at runtime. Of note, you can use the object pointer, too, to use as a hash key. Perhaps as a function overload?
[QUOTE=Map in a box;48614983]That's just one method of making a hashmap. Use malloc/realloc or, less preferably, a linked list to resize your table at runtime. Of note, you can use the object pointer, too, to use as a hash key. Perhaps as a function overload?[/QUOTE]
C doesn't do function overloading, and I reckon using the pointer value as the hash key would lead to a terrible distribution (haven't tested it though).
It could be interesting to use a function pointer for the hashcode implementation however, though you'd need to void* the things (if you want it to be a bit more general).
So you'd have something like:
[code]
hash_get(hashtable, some_key)
[/code]
Which would use a function pointer for the hashcode defined in the hashtable struct and pipe the key to that for the lookup.
Just an example :v:
Actually, nevermind my rambling, here instead have an article to a generic hashtable implementation which is now part of linux kernel since 3.7: [url]http://lwn.net/Articles/510202/[/url]
[QUOTE=Profanwolf;48615285]C doesn't do function overloading[/QUOTE]
You can get around this with void pointers and stuff.
I'll look into that for academic reasons but this very simple hash table (I will add buckets later) will work for what I'm using it for
[editline]4th September 2015[/editline]
Also I can't use the pointer as the key because I'm using this hash table to store values for variables in a language parser so they will be passed a raw string rather than the same object every time
[QUOTE=Profanwolf;48615285]C doesn't do function overloading, and I reckon using the pointer value as the hash key would lead to a terrible distribution (haven't tested it though).
It could be interesting to use a function pointer for the hashcode implementation however, though you'd need to void* the things (if you want it to be a bit more general).
So you'd have something like:
[code]
hash_get(hashtable, some_key)
[/code]
Which would use a function pointer for the hashcode defined in the hashtable struct and pipe the key to that for the lookup.
Just an example :v:
Actually, nevermind my rambling, here instead have an article to a generic hashtable implementation which is now part of linux kernel since 3.7: [url]http://lwn.net/Articles/510202/[/url][/QUOTE]
Oh didn't dawn on me that you were using C. The problem for using void* for everything is it wouldn't work for copyable things such as structs and strings.
[editline]4th September 2015[/editline]
However hash_gets and hash_getv would work
[QUOTE=Fourier;48603536]Firecode font, if anyone is interested.. cool stuff :)
[/QUOTE]
this is nice man
[QUOTE=proboardslol;48616414]I'll look into that for academic reasons but this very simple hash table (I will add buckets later) will work for what I'm using it for
[editline]4th September 2015[/editline]
Also I can't use the pointer as the key because I'm using this hash table to store values for variables in a language parser so they will be passed a raw string rather than the same object every time[/QUOTE]
Pure function!
Function which does not rely on MUTABLE state.
Rule: It returns ALWAYS same result for same input.
Which means:
You can generate pointer address from string. With help of pure function. Let's call this function int StrToPtr. The StrToPtr is hash function.
[code]
//Returns int or address
int StrToPtr(string str);
//You can also call this function HASH function
int Hash(string str);
[/code]
Each memory space is FIXED SIZE. This means we are limited to space.
Hash functions job is to project elements from INFINITE space to FINITE (countable). This is how we can map strings of INFINITE variants to fixed space (fixed memory).
[b]HashTable[/b]
So yeah either you use already working HashTable or make your own.
Just a hint, with % (modulo) operator you can limit infinite space to finite. Cryptography relies on this heavily.
[code]
int finite = infinite % spaceSize
[/code]
[editline]5th September 2015[/editline]
Now you can also guess why collisions occur.
[QUOTE=Fourier;48616667]Pure function!
Function which does not rely on MUTABLE state.
Rule: It returns ALWAYS same result for same input.
Which means:
You can generate pointer address from string. With help of pure function. Let's call this function int StrToPtr. The StrToPtr is hash function.
[code]
//Returns int or address
int StrToPtr(string str);
//You can also call this function HASH function
int Hash(string str);
[/code]
Each memory space is FIXED SIZE. This means we are limited to space.
Hash functions job is to project elements from INFINITE space to FINITE (countable). This is how we can map strings of INFINITE variants to fixed space (fixed memory).
[b]HashTable[/b]
So yeah either you use already working HashTable or make your own.
Just a hint, with % (modulo) operator you can limit infinite space to finite. Cryptography relies on this heavily.
[code]
int finite = infinite % spaceSize
[/code]
[editline]5th September 2015[/editline]
Now you can also guess why collisions occur.[/QUOTE]
So are you saying that using this strToPtr function, putting in "ab" will get me a different int than "ba"? and if I put in "ab", it will return the same int every single time?
So I took a half-summer hiatus from working on my project, primarily because it is extremely easy for me to give up when the going gets tough in the realm of programming.
However as of late I've been craving that feeling of making cool things, and felt guilty for just giving up on a past years worth of work.
So I gave myself a new set of goals and now I've got the motivation to continue working once again, and hopefully it can continue to fuel me for yet another year.
This time I'm starting to address problems with my rendering back-end. It occurred to me that it was largely inefficient, but I had no real idea on how to solve it until now. It turns out I misunderstood how deferred rendering should actually be implemented.
Step 1 of 3 is now complete. All this image shows is a simple scene but being rendered to a separate frame buffer, and split into several different render targets rendered as (clockwise) diffuse, normal, texture coordinates, and world.
[t]http://i.imgur.com/0VhM3cw.png[/t]
Next up I hope to have the lighting aspect re-introduced.
[QUOTE=Darkwater124;48613248]I thought Windows 7 had an XPS "printer", and 8+ an actual PDF "printer"?[/QUOTE]
You're probably right.
It just goes to show how long it's been since I've needed to print anything to a PDF, huh?
Sorry, you need to Log In to post a reply to this thread.