EPT"PNGwriter:: plot_text - ERROR **: FreeType: Could not find or load font file."
Not sure if im just being bad.
[Tool] SC2 Map Analyzer - Page 3
| Forum Index > SC2 Maps & Custom Games |
|
Wihl
Sweden472 Posts
"PNGwriter:: plot_text - ERROR **: FreeType: Could not find or load font file." Not sure if im just being bad. | ||
|
Logo
United States7542 Posts
| ||
|
Wihl
Sweden472 Posts
| ||
|
Wihl
Sweden472 Posts
"Opening .\space2.SC2Map... Warning: Could not open enUS.SC2Data\LocalizedData\GameStrings.txt for reading. Warning: Could not extract map name from archive, using archive filename instead. Warning: Map name sc2map begin (amagads typo) reused, map-specific output for previous instance will be clobbered." It used "sc2map" for all the 8 maps instead of using, lets say, space2 and so on. | ||
|
chuky500
France473 Posts
| ||
|
kraemahz
United States4 Posts
you'll need a manual You already have the -h parameter including this information, I don't understand why you wouldn't mention that. It would also help make it feel more like a "standard" command line exec if you included -? as help also. Another suggestion: you could extend the functionality of help by adding detailed help to the param features that better explains what each function does. E.g: sc2mapanalyzer.exe -? -w | ||
|
Logo
United States7542 Posts
| ||
|
Pharow
United States19 Posts
| ||
|
Wihl
Sweden472 Posts
| ||
|
Shanedon
United States147 Posts
| ||
|
dimfish
United States663 Posts
On June 26 2010 02:32 MoreFasho wrote: I've noticed a bigger imbalance on scrap station, distance from main to ramp which matters quite a bit for zerg. In ZvZ when I spawn top on scrap station I will wall -in a little bit with a roach warren/ evo chamber to protect against speedlings/banelings, but this is impossible on the bottom as the creep doesn't extend to the ramp! This is an interesting point--the map analyzer doesn't compare all of the terrain of the whole map on one side versus another. That's pretty hard to do because its hard for a computer to know which area should correspond to another unless there is perfect symmetry. That's why what it does now is compare the terrain surrounding bases only (a radius of 12 cells by default). So it'll look at whether my main's mineral line is butting up against a cliff and if your mineral is hanging out in a wide open space and report a difference. One of my plans is to sample the openness of every point along the shortest path from one start location to another. You should get a roller coaster of values as the chokes tighten and open along the way, and this is like a signature that should roughly match for every start location. It's in the works, anyway, and would detect something like you observed, that one base's main choke is closer to the hatchery than the other. | ||
|
dimfish
United States663 Posts
On June 26 2010 05:38 monkh wrote: I don't think it recognises areas blocked by doodads (i think) i had area blocked by trees on one of my maps and it didn't seem to account for them in travelling distances Yes, you are right. The main project page over at SC2Mapster explains the problem in detail. Essentially the map data says "put a tree here, put a mineral patch there, put a destructible rock over there" but doesn't tell you the footprint of any of them! I am currently hard-coding the shape of stuff like the various destructible rocks but it would take eons to write down the footprint of everything in the editor that can block pathing. Two immediate solutions: 1) ask me for a small list of really important doodads that SC2 Map Analyzer should look for OR 2) temporarily add something the analyzer already recognizes, like destructible rocks, over your important line of trees and then take them out after getting the info you needed. | ||
|
Archerofaiur
United States4101 Posts
| ||
|
dimfish
United States663 Posts
On June 26 2010 06:15 chuky500 wrote: I have a suggestion, you could generate a picture that shows how safe spots are on the map, depending on the distance to several things. For example on Steppes of War, on the big ramp below the natural I remember Day9 saying the safe spot to place tanks was on the cliff between the ramp and the natural, just far enough to be able to shoot enemies. You could assign points to judge how safe a spot is, like on a cliff or under a cliff, or if a spot can be shot but not reached by melee units, if a kind of unit can shoot there but not another kind of unit. That would be useful to know what's the safest route to attack for example or if a large way to attack is safer on the left or the right. Or where to place your units in defense. But maybe it would require too much computing time. This is neat and fits in with something I'm toying with. Stay tuned for future updates! On June 26 2010 06:15 chuky500 wrote: It would be nice to have informations on the distance if you take an alternate route too, but I don't know how hard it would be to find the alternate routes. Alternate routes is a hard concept for a computer. There is an exact shortest path (easy for computer) and many ways to count alternate paths. At some level us humans just gotta get in there and augment what the analyzer can do. Still, if you have a proposal for an alternate route algorithm I am willing to entertain it. On June 26 2010 06:15 chuky500 wrote: Another thing is to run your program I had to launch cmd.exe but I had to place myself in your program's directory then type sc2mapanalyzer.exe and then drag my map in the console, but it would have been easier if I could just drag your exe in the console, then drag my map because this doesn't work, it says it can't find the free font (I guess because it doesn't run the exe from the right directory). Or even running your program without opening the console by drag and dropping the map file onto the exe would be fine. I'm using Windows XP btw. Roger that, verified that running program from command line but in directory other than where it is sitting it cannot find the font file. Will work on it and post back in this thread when a better version is available. On June 26 2010 06:15 chuky500 wrote: Also, I tested 2 of my maps and since you display the full path in the picture one of them had the informations cropped in the picture that compares the 6 o' clock vs the 12 o' clock. I don't know if the size of the font is related to the size of the map because my huge map had a smaller font size and the picture didn't crop the 6 and 12 o' clock infos in the file name but for the regular sized map it was cropped. And all my maps are in the My Documents directory anyways so it's not really usefull to display the c:\documents and settings\... and all The size of the font is fixed and the images are actually different sizes in pixels. If you view them with Windows Photo Gallery, for instance, it shrinks all the images to a fixed size and the bigger images will appear to have smaller font. Also, have you given your map a name in the editor? Go to map properties and fill it out and see if your images display better. Basically, if the analyzer can't find a name for the map in the data it makes up a name for you based on the filename, which yeah can be a long mess. On June 26 2010 06:15 chuky500 wrote: One last thing, on my huge map the cliffs were tinted a bit grey, I don't know if it's a feature but I didn't see my cliffs were a bit grey in my normal sized map. All of the images use black and grey to show you the 4 possible cliff levels. The lowest elevation which is always unplayable shows up as black. The other three levels are grey, dark for the lowest and getter lighter for higher levels. Is your normal sized map all one cliff level? | ||
|
Jeffadore
United States5 Posts
and nice work man, keep it up | ||
|
dimfish
United States663 Posts
On June 26 2010 08:56 Logo wrote: Does your tool respect no-pathing fills? Also I'm getting up to 8% positional imbalances on my map that's almost entirely symmetrical. Possibly due to no-pathing fills? The edge of my map, separated by a gap has uneven terrain since it's no-pathing). Other than that there's a small 1 tile imbalance (bases along the line of symmetry can't be positionally balanced) and 1 extra tile right on the line of symmetry (useless tile, just makes it prettier), but is that enough for the map to show something like a 10% difference? Correct, it does not respect no-pathing fills at the moment. It's on the todo list. And now that you mention it, it would be helpful if you could go to the the project page at SC2Mapster and put bugs or feature requests in as tickets. That way I can organize what's wrong and all. The imbalance sounds odd for sure. There are two components to the balance, one of which is the distance to the offending base(s). The other is an attribute of the base. So let's say the naturals are exactly in the same place but one natural only has 7 mineral patches. There will be an imbalance by total resources even though the bases are in symmetrical positions. | ||
|
dimfish
United States663 Posts
On June 26 2010 09:47 Wihl wrote: Analyzed all my own maps, really liking this. However.. I put like 8 maps in the same directory as the exe and ran the program. "Opening .\space2.SC2Map... Warning: Could not open enUS.SC2Data\LocalizedData\GameStrings.txt for reading. Warning: Could not extract map name from archive, using archive filename instead. Warning: Map name sc2map begin (amagads typo) reused, map-specific output for previous instance will be clobbered." It used "sc2map" for all the 8 maps instead of using, lets say, space2 and so on. Gotcha--I'm hard-coding this beast to grab the English-localized game strings because I was having a problem with getting documentation for StormLib with regard to localization. It's on my todo list now to get this resolved. Maybe you can analyze each map in separate directory until I get a fix for you? | ||
|
dimfish
United States663 Posts
On June 26 2010 09:56 chuky500 wrote: Another small suggestion : since your program can identify bases it would be nice to know how much time it takes to move from the main to the 3rd base, or from the natural to the 3rd base. So you can have an idea if it's easy or hard to defend. The philosophy of the map analyzer is that if it is the same for both players then it is balanced. Its sort of up to you to decide if you want your map to have hard-to-defend third bases, etc. Still, I got this and your earlier suggestion recorded because I have an idea that would notice whether one base is "cliffable" and another isn't and stuff like that. | ||
|
dimfish
United States663 Posts
On June 26 2010 10:01 kraemahz wrote: You already have the -h parameter including this information, I don't understand why you wouldn't mention that. It would also help make it feel more like a "standard" command line exec if you included -? as help also. Another suggestion: you could extend the functionality of help by adding detailed help to the param features that better explains what each function does. E.g: sc2mapanalyzer.exe -? -w I wanted to release soon (trying to beat phase 2!) so I intentionally put the master explanation at the manual site. I've got your suggestions recorded and on the todo pile! | ||
|
PrinceXizor
United States17713 Posts
Alternate routes is a hard concept for a computer. There is an exact shortest path (easy for computer) and many ways to count alternate paths. At some level us humans just gotta get in there and augment what the analyzer can do. Still, if you have a proposal for an alternate route algorithm I am willing to entertain it. How about the shortest possible path coming closest to as many expansions as possible? i don't know the exact math, but would something like that be possible? (it's also more likely to be an alternate path, from guarding your expos. | ||
| ||
