|
Id4ever posted:Fixed now. Excellent work. The font looks pretty drat good. It's a bit misaligned in places, but that's my own drat fault. I may write out another one more carefully and try again soon.
|
# ? Aug 25, 2009 22:15 |
|
|
# ? May 11, 2024 10:10 |
|
Brilliant service, we'll probably end up using it for a game i'm working on if it makes scalable fonts easy.
|
# ? Aug 25, 2009 23:10 |
|
brian posted:Brilliant service, we'll probably end up using it for a game i'm working on if it makes scalable fonts easy. It generates TrueType fonts, which are vector based and should be usable at (almost) any size.
|
# ? Aug 25, 2009 23:25 |
|
Id4ever posted:
Definitively going to try this out as soon as I install my printer/scanner. Great idea.
|
# ? Aug 26, 2009 02:30 |
|
https://www.fontcapture.com rocks, and I pity whoever graded my AP Comp Sci test. font: http://www.2shared.com/file/7398277/72deee93/myfont.html Trabisnikof fucked around with this message at 18:50 on Aug 27, 2009 |
# ? Aug 26, 2009 22:23 |
|
(Scheme and Erlang in MofontQc Bold!) It looks good enough that I might use it for parts of my book. Only thing missing is the underscore.
|
# ? Aug 27, 2009 01:04 |
|
Id4ever posted:www.fontcapture.com That's pretty cool. You should put up some sample font to download. And people posting in this thread should post links to the fonts they created.
|
# ? Aug 27, 2009 17:21 |
|
MononcQc posted:Only thing missing is the underscore. I'll add underscore to the template either today or tomorrow. It'll be in the next available box, to the right of the '±' character. If you still have your completed template, just write the underscore in that box and re-scan and upload it. FigBug posted:That's pretty cool. You should put up some sample font to download. And people posting in this thread should post links to the fonts they created. Yes, I'm planning to add a 'gallery' section where you can preview and download fonts created by other users (with their permission of course). In the meantime, here are some interesting / pretty / weird fonts that people have created so far:
|
# ? Aug 27, 2009 20:39 |
|
This is pretty excellent. A minor issue: the fonts don't play nice with the "smart quotes" feature in Word since they don't include the upside down " and ' characters. Edit: on the same note, en and em dashes would be nice too, even if they are just duplicates of minus. Nippashish fucked around with this message at 23:09 on Aug 27, 2009 |
# ? Aug 27, 2009 23:05 |
|
It's nowhere near as useful as that font maker, but I'm implementing maze algorithms that have already been done a million times. Hue represents distance from the start (top-left corner).
|
# ? Aug 28, 2009 10:27 |
|
The hues are cool.
|
# ? Aug 28, 2009 14:43 |
|
Are you planning on sharing the code when you are done Scaevolus? I'd really love to see how you implemented that.
|
# ? Aug 28, 2009 15:00 |
|
What's the hue if there are multiple routes to the same location? Or does it only work for mazes with one path to each location?
|
# ? Aug 29, 2009 15:11 |
Screeb posted:What's the hue if there are multiple routes to the same location? Or does it only work for mazes with one path to each location? I'm guessing it uses the lowest distance to the location, unless he's using something other than A*.
|
|
# ? Aug 29, 2009 17:54 |
|
I think that errors sometimes look more interesting than the expected results. Click here for the full 1296x758 image.
|
# ? Aug 30, 2009 04:56 |
|
Screeb posted:What's the hue if there are multiple routes to the same location? Or does it only work for mazes with one path to each location? Right now it generates perfect mazes (only one path between any two locations), but it would display the hue for the shortest distance if there were loops. Jo posted:I'm guessing it uses the lowest distance to the location, unless he's using something other than A*. I'm using Dijkstra's algorithm because it lets you find the shortest path between a vertex and every other vertex on the graph, while A* just uses heuristics to find a path between two vertices on a graph.
|
# ? Aug 30, 2009 09:43 |
Scaevolus posted:I'm using Dijkstra's algorithm because it lets you find the shortest path between a vertex and every other vertex on the graph, while A* just uses heuristics to find a path between two vertices on a graph. A* is shortest path, too. It's also faster than Dijkstra, if I remember right. Jo fucked around with this message at 15:21 on Aug 30, 2009 |
|
# ? Aug 30, 2009 15:19 |
|
Jo posted:
A* is shortest-path only if the estimation is good enough.
|
# ? Aug 30, 2009 15:24 |
|
Id4ever posted:www.fontcapture.com Hey you're on lifehacker now. Congratulations!
|
# ? Aug 30, 2009 16:51 |
|
MonDubious posted:A* is shortest-path only if the estimation is good enough. The way I remember it: If the heuristic you use is admissable (no estimation is greater than the actual cost), then A* is guaranteed to eventually find the shortest path. If the heuristic is monotonic (the decrease in the estimated cost from node X to node Y is smaller than the distance between X and Y) then A* will never visit the same node twice before finding the shortest path.
|
# ? Aug 31, 2009 04:43 |
|
akadajet posted:I think that errors sometimes look more interesting than the expected results.
|
# ? Aug 31, 2009 14:32 |
|
Jo posted:
I think you missed the point; Dijkstra's finds the least costs from one node to every other node. A* finds the least cost from one node to one of a set of goal nodes.
|
# ? Sep 2, 2009 04:59 |
|
Ugg boots posted:A* finds the least cost from one node to one of a set of goal nodes. And even then, that's not guaranteed unless your heuristic is admissable (as AgentF said), which it rarely is. I'll draw an example if people still don't get that.
|
# ? Sep 3, 2009 08:34 |
|
Scaevolus posted:And even then, that's not guaranteed unless your heuristic is admissable (as AgentF said), which it rarely is. I always thought that the most frequently used A* heuristic would be euclidian distance, either in a grid or in a graph. And that is admissable. I'd love to see realistic examples of the contrary!
|
# ? Sep 3, 2009 16:10 |
|
Heuristics tend to only be as good as their extremely constrained "guessing function" in practice, and while it's quite acceptable if you don't always get the best route, Dijkstra is what people tend to use in practice. For example, see almost all the routing protocols in existence - they're based around gathering information from neighbors (local optimizations) to compose them up to globally optimal values. A* is not really great for that, but it's pretty sweet if you need a route and you need it fast. If you are privy to information that will yield an admissable heuristic that beats other straightforward algorithms, then you have yourself (amortized) a better algorithm.
|
# ? Sep 3, 2009 16:40 |
|
wlievens posted:I always thought that the most frequently used A* heuristic would be euclidian distance, either in a grid or in a graph. And that is admissable. I'd love to see realistic examples of the contrary! This is the case I was thinking of (applet here): So yes, A* would work fine for finding the shortest solution for the maze, but I wanted the distance for every node, making Dijkstra's Algorithm more efficient.
|
# ? Sep 4, 2009 00:09 |
|
Fontcapture is fantastic, I'm going to love writing with my own handwriting! This game is nearing completion, you can probably guess the premise:
|
# ? Sep 4, 2009 16:50 |
|
blueberrypudding posted:Fontcapture is fantastic, I'm going to love writing with my own handwriting! You're going to have to change the name. The company that owns the rights to Tetris has constantly sued people who use the "tris" suffix. There was a big deal about this several years ago with "Netris", and has been going on for decades. The company has proven in court that use of the "tris" suffix is enough to infringe on their rights, and is essentially "bandwagoning" for any "falling block game". They sued an iPhone developer once before: http://www.savvywallet.com/2008/08/27/iphone-application-tris-gets-the-boot/
|
# ? Sep 4, 2009 18:28 |
|
Inverse Icarus posted:You're going to have to change the name
|
# ? Sep 4, 2009 19:25 |
|
As far as I know, it was specifically the "tris" suffix. Textro should be fine. But I am not a lawyer, so do whatever you think would be best to avoid being sued.
|
# ? Sep 4, 2009 19:55 |
|
Canvas based pixel art wiki maze I have been working on. Still has a lot of rough edges but it works. http://maze.heroku.com
|
# ? Sep 6, 2009 16:28 |
|
After the first maze page, if you hold down the move key whilst the page is loading you can miss whole pages out. Looks pretty cool though
|
# ? Sep 7, 2009 07:32 |
|
I had some fun with this, but I found the edit feature to be really laggy.
|
# ? Sep 7, 2009 16:29 |
|
blueberrypudding posted:I had some fun with this, but I found the edit feature to be really laggy. The editor is laggy because its loading up like 100 color pickers. I am sure I can be smarter about it I just wanted to get it out the door. So far here is what I've learned people like drawing way better then adding walls. . Most people that draw do relatively simple designs probably because the editor sucks but whoever drew the Mario is awesome. The walls that do get added are mostly closing off exits to the screen or spelling things. I may just randomly generate the maze walls when the first person enters an area.
|
# ? Sep 7, 2009 20:15 |
|
openGL is hard
|
# ? Sep 7, 2009 20:26 |
|
tripwire posted:openGL is hard Glad to see you got it working-- what was the bug?
|
# ? Sep 7, 2009 20:54 |
|
Scaevolus posted:Glad to see you got it working-- what was the bug? Well, I still haven't found the bug, and I'm not so sure it's working. That picture is from many thousands of generations into the simulation where it should be converging to look like the source picture but the fitness seems to stay the same at roughly 50% similar to the source, forever. This is what you'd expect to happen if there was zero evolutionary pressure, or if there was no way at all for the drawings to pass on fit genes to future drawings (or put another way, it seems like either the drawing or some aspect of the evaluation of the drawing is failing completely). All it spits out is random canvas after random canvas with no improvements. I know the genetic algorithm itself works because when I switch the rendering engine back to cairo it works like normal again.
|
# ? Sep 7, 2009 20:59 |
|
tripwire posted:I know the genetic algorithm itself works because when I switch the rendering engine back to cairo it works like normal again.
|
# ? Sep 7, 2009 21:10 |
|
I'm a naughty boy. http://maze.heroku.com/maze?x=15&y=20
|
# ? Sep 7, 2009 22:03 |
|
|
# ? May 11, 2024 10:10 |
|
RMS is hacking the maze! http://maze.heroku.com/maze?x=30&y=200
|
# ? Sep 7, 2009 23:08 |