| --- Log opened Sat May 02 00:00:55 2015 |
| 00:02 | | Turaiel[Offline] is now known as Turaiel |
| 00:19 | | Turaiel is now known as Turaiel[Offline] |
| 01:21 | | Turaiel[Offline] [Brandon@Nightstar-olc20d.resnet.mtu.edu] has quit [[NS] Quit: Bouncer terminated] |
| 01:56 | | * Vornicus fails miserably at getting graphviz to do what he wants wit gentle prodding |
| 01:58 | | Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [Connection reset by peer] |
| 02:25 | | * McMartin sets about trying to install the full HaxeFlixel stack. |
| 02:34 | | Ogredude [quassel@Nightstar-dm1jvh.projectzenonline.com] has joined #code |
| 02:34 | | mode/#code [+o Ogredude] by ChanServ |
| 02:44 | <~Vornicus> | Flixel as in the thing that, uh, Manufactoria was built in? |
| 02:51 | <&McMartin> | Probably. It's a common Flash game library. Haxe is a language and system for cross-platform development that is what Papers Please was written in. |
| 02:51 | <&McMartin> | My understanding is that HaxeFlixel bridges the two. |
| 02:57 | | VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has quit [[NS] Quit: Leaving] |
| 02:57 | | VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has joined #code |
| 03:05 | <&McMartin> | I'm also back to poking at Monocle to make it good enough to implement, well, anything I've previously released |
| 03:08 | <~Vornicus> | including Sable? |
| 03:16 | <&McMartin> | Anything means "any one of" here, not "all of" |
| 03:17 | <&McMartin> | So Sable, Target Acquired, or Hex Inverter |
| 03:17 | <&McMartin> | I can't implement Hex Inverter because mouse-sprite collisions aren't really there |
| 03:17 | <&McMartin> | I can't implement Target Acquired because I don't yet have pixel-perfect collisions |
| 03:17 | <&McMartin> | Sable is Out Of Scope |
| 03:19 | <&McMartin> | One of the sample programs along the way to getting that stack installed is a bejeweled clone named Pirate Pig (part of OpenFL) |
| 03:19 | <&McMartin> | It's almost cute |
| 03:19 | <&McMartin> | But not quite, and so it is just half-assed and twee. =( |
| 03:25 | < VirusJTG> | need to reboot brb |
| 03:25 | | VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has quit [Connection reset by peer] |
| 03:26 | | VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has joined #code |
| 03:27 | < VirusJTG> | well... getting there |
| 03:41 | <&McMartin> | OK, it looks like I've got the stack up and running |
| 04:02 | | Vornicus [vorn@ServerAdministrator.Nightstar.Net] has quit [[NS] Quit: Leaving] |
| 04:09 | <&McMartin> | Now to make it all go on Linux as well. |
| 04:29 | | TheWilfox [awilfox@Nightstar-b8i.hbu.99.192.IP] has joined #code |
| 04:46 | <&McMartin> | http://the-toast.net/2015/05/01/every-question-in-every-qa-session-ever/ |
| 05:08 | <&McMartin> | And now it's working on Linux |
| 05:41 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 06:02 | | kourbou|zzz is now known as kourbou |
| 06:04 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [[NS] Quit: Don't think much of what others think of you. Think of what you think of others.] |
| 06:11 | | Checkmate [Z@Nightstar-pdi1tp.customer.tdc.net] has quit [Ping timeout: 121 seconds] |
| 06:11 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has quit [[NS] Quit: ] |
| 06:17 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 06:33 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 06:48 | | TheWilfox [awilfox@Nightstar-b8i.hbu.99.192.IP] has quit [[NS] Quit: Don't hang on, nothing lasts forever but the earth and sky.] |
| 07:21 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has quit [[NS] Quit: ] |
| 07:40 | | Kindamoody[zZz] is now known as Kindamoody |
| 08:08 | | kourbou [kourbou@Nightstar-pjt.geh.161.37.IP] has joined #code |
| 08:10 | | kourbou [kourbou@Nightstar-pjt.geh.161.37.IP] has quit [[NS] Quit: Don't think much of what others think of you. Think of what you think of others.] |
| 09:37 | | Kindamoody is now known as Kindamoody|afk |
| 10:00 | | mac is now known as macdjord|slep |
| 11:01 | | Meatyhandbag [sebastianfe@Nightstar-dk0.5eb.224.136.IP] has quit [Client exited] |
| 11:06 | | GreenGuy [GreenGuy@Nightstar-fik25i.gnomino.eu] has joined #code |
| 11:10 | | kourbou [kourbou@Nightstar-tk5.i3r.165.37.IP] has joined #code |
| 11:16 | | GreenGuy [GreenGuy@Nightstar-fik25i.gnomino.eu] has quit [[NS] Quit: Leaving] |
| 11:31 | | kourbou [kourbou@Nightstar-tk5.i3r.165.37.IP] has quit [Ping timeout: 121 seconds] |
| 11:32 | | kourbou [kourbou@Nightstar-tk5.i3r.165.37.IP] has joined #code |
| 11:38 | | kourbou [kourbou@Nightstar-tk5.i3r.165.37.IP] has quit [Ping timeout: 121 seconds] |
| 11:42 | | kourbou [kourbou@Nightstar-tk5.i3r.165.37.IP] has joined #code |
| 12:00 | | kourbou|phone [kourbou@Nightstar-mqk.m1t.161.37.IP] has joined #code |
| 12:01 | | kourbou [kourbou@Nightstar-tk5.i3r.165.37.IP] has quit [NickServ (RECOVER command used by kourbou|phone)] |
| 12:01 | | kourbou|phone is now known as kourbou |
| 12:16 | | Meatyhandbag [sebastianfe@Nightstar-dk0.5eb.224.136.IP] has joined #code |
| 12:17 | | Kindamoody|afk is now known as Kindamoody |
| 12:25 | | kourbou|phone [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 12:26 | | kourbou [kourbou@Nightstar-mqk.m1t.161.37.IP] has quit [NickServ (RECOVER command used by kourbou|phone)] |
| 12:26 | | kourbou|phone is now known as kourbou |
| 12:38 | | catadroid` [catalyst@Nightstar-ndh970.dab.02.net] has joined #code |
| 12:40 | | catadroid [catalyst@Nightstar-p9cb1l.range86-174.btcentralplus.com] has quit [Ping timeout: 121 seconds] |
| 12:41 | | catadroid` [catalyst@Nightstar-ndh970.dab.02.net] has quit [A TLS packet with unexpected length was received.] |
| 12:42 | | kourbou is now known as kourbou|foodz |
| 12:42 | | catadroid [catalyst@Nightstar-ndh970.dab.02.net] has joined #code |
| 12:42 | | catadroid [catalyst@Nightstar-ndh970.dab.02.net] has quit [A TLS packet with unexpected length was received.] |
| 12:49 | | GreenGuy [GreenGuy@Nightstar-fik25i.gnomino.eu] has joined #code |
| 12:51 | | kourbou|foodz [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [Ping timeout: 121 seconds] |
| 13:19 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 13:30 | | GreenGuy [GreenGuy@Nightstar-fik25i.gnomino.eu] has quit [Connection closed] |
| 13:38 | | GreenGuy [GreenGuy@Nightstar-5sh.fa6.179.88.IP] has joined #code |
| 13:39 | | GreenGuy is now known as NSGuest58407 |
| 13:40 | | NSGuest58407 is now known as GreenGuy |
| 14:06 | | GreenGuy [GreenGuy@Nightstar-5sh.fa6.179.88.IP] has quit [Connection closed] |
| 14:07 | | GreenGuy [NSkiwiirc@Nightstar-fik25i.gnomino.eu] has joined #code |
| 14:07 | | Kindamoody is now known as Kindamoody|afk |
| 14:22 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 14:22 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 14:27 | | GreenGuy [NSkiwiirc@Nightstar-fik25i.gnomino.eu] has quit [[NS] Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] |
| 14:56 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has quit [[NS] Quit: ] |
| 15:15 | | GreenGuy [GreenGuy@Nightstar-5sh.fa6.179.88.IP] has joined #code |
| 15:16 | | GreenGuy is now known as NSGuest2800 |
| 15:25 | | Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 15:25 | | mode/#code [+o Vash] by ChanServ |
| 16:13 | | Meatyhandbag [sebastianfe@Nightstar-dk0.5eb.224.136.IP] has quit [Client exited] |
| 18:06 | | Meatyhandbag [sebastianfe@Nightstar-dk0.5eb.224.136.IP] has joined #code |
| 18:21 | | NSGuest2800 is now known as GreenGuy |
| 18:51 | | Orthia [orthianz@Nightstar-tap.qer.184.203.IP] has quit [Ping timeout: 121 seconds] |
| 18:55 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [[NS] Quit: What does the 'x' do?] |
| 18:56 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 18:56 | | Orthia [orthianz@Nightstar-629.6n6.184.203.IP] has joined #code |
| 18:56 | | mode/#code [+o Orthia] by ChanServ |
| 19:02 | | GreenGuy [GreenGuy@Nightstar-5sh.fa6.179.88.IP] has quit [Connection closed] |
| 19:49 | < [R]> | Is there something like HTML-tidy that doesn't shove linebreaks everywhere? |
| 19:52 | <&McMartin> | I'm kind of surprised that's not a setting |
| 19:54 | | kourbou is now known as kourbou|foodz |
| 19:55 | | * McMartin finds himself once again in a position where he needs to implement something like inheritance or closures in C. |
| 19:56 | <&McMartin> | Step one is probably to do a couple takes of it with copy-paste to see which bits can be kept the same, and then turn the rest into method-like functions. |
| 19:57 | <&McMartin> | (This is for collision detection in Monocle. I'll also need to stockpile a bunch of efficient collision algorithms for various shapes~) |
| 20:09 | <&McMartin> | GENTLEMEN. LADIES. KITTENS THAT HAPPEN TO HAVE FOUND THE BUTTON THAT MAKES THE WARM LIGHT BOX TURN ON. |
| 20:10 | <&McMartin> | I am pondering collision detection between axis-aligned bounding boxes and circles. |
| 20:10 | <&McMartin> | I believe, but am not certain, that I merely have to check to see if any corner of the AABB is inside the circle, or if any "axis extremum" of the circle is inside the rectangle. |
| 20:10 | <&McMartin> | But so far, my proof of this is "I can't seem to draw a counterexample" |
| 20:10 | | kourbou|foodz [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [Ping timeout: 121 seconds] |
| 20:11 | | macdjord|slep is now known as macdjord |
| 20:11 | <&jerith> | Hrm. |
| 20:12 | <&McMartin> | Also it is the stroke of noon. Perhaps I will also think about this over lunch. |
| 20:12 | <&McMartin> | (By "axis extremum" I mean the point on the circle that is the most positive or negative on the X or Y axis, so a circle at the origin of radius 1 has axis extrema at (0, +/- 1) and (+/- 1, 0)) |
| 20:15 | <&McMartin> | Aha, I *have* found a counterexample |
| 20:15 | <&jerith> | Circle with diameter slightly bigger than a square? |
| 20:15 | <&McMartin> | ... no I haven't |
| 20:16 | <&McMartin> | ... yes I have |
| 20:16 | <&McMartin> | The circumscription cases all work, though |
| 20:16 | <&McMartin> | Draw a narrow bar through the top quarter of the circle. |
| 20:16 | <&jerith> | Something like the Vitruvian Man's boxes? |
| 20:17 | < [R]> | Isn't circle-based colision the simplest though? |
| 20:17 | <&McMartin> | R: I'd often say that AABB collision is the simplest |
| 20:17 | <&McMartin> | However, sometimes you want circles |
| 20:17 | <&McMartin> | And in this case I'm ultimately also going to need pixel-perfect |
| 20:17 | <&McMartin> | So a sleazy way to do this is to turn "circle" into "pixel mask" and go from there if it's not circle-to-circle, but I'd prefer not to because there should be easier ways. |
| 20:17 | < [R]> | It's just "is the circle's radius lesser than the point closest to the circle?"? |
| 20:17 | <&jerith> | AABB/AABB and circle/circle are both pretty easy. |
| 20:18 | <&McMartin> | Yeah |
| 20:18 | <&McMartin> | I suspect the question here is "how to most efficiently compute "the point on the AABB with the point closest to the center of the circle" |
| 20:18 | < [R]> | Gets hard when you have weird shapes though. |
| 20:18 | <&McMartin> | ... but it's an AABB, which means it's either directly axis-aligned with the center of the circle (that is, if projected against an edge, the center is *on* the edge) or it's a corner, yes? |
| 20:20 | <@macdjord> | McMartin: Isn't there are easy way to calculate the intersection of a line and a circle? Do that for each of the 4 lines, then check for one object completely within the other. |
| 20:21 | <&McMartin> | macdjord: I believe that is the algorithm for the OBB case. |
| 20:21 | <&McMartin> | I think I can simplify it for the axis-aligned case. |
| 20:21 | <&jerith> | McMartin: I think your corners and axis extrema work if you also add a check for the centre of each shape being within the other. |
| 20:21 | <@macdjord> | jerith: Nope. |
| 20:21 | <&McMartin> | jerith: I think that fails the same way |
| 20:21 | <&McMartin> | If I extend my "horizontal bar" arbitrarily far to the right, I can move the center of the rectangle out. |
| 20:22 | <@macdjord> | Assume the long box extending through - then make it longer in one direction until- yeah. |
| 20:22 | <&jerith> | Oh, right. |
| 20:23 | <&McMartin> | macdjord: I think the "intersection of line and circle" case sort of ends up becoming "read this value" and I can simplify further because I'm not trying to find *where* it intersects, merely *that* it does. |
| 20:23 | <&McMartin> | The geometric construction is, I think, "find the point on a line closest to some other point" |
| 20:24 | <&McMartin> | We can restrict ourselves to the two lines of the rectangle nearest the center, which makes it more like a pair of rays, or an angle ABC |
| 20:25 | <&McMartin> | So we find the closest point to the circle's center on AB and BC - and if it's not on the segment, we check point B for that |
| 20:25 | <&McMartin> | If the distance between those points is less than or equal to r, there is a collision. |
| 20:26 | <&McMartin> | And since they are axis aligned, "the point closest to the circle's center", if the circle is at (x, y), are basically (0, y) and (x, 0) |
| 20:26 | <&McMartin> | (Well, really it's more shuffling constants around) |
| 20:27 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 20:27 | <&McMartin> | It's probably actually cheaper to just do all four comparisons, looking at it, rather than figure out which one or two comparisons are truly necessary; that's even more branches. |
| 20:29 | <@macdjord> | McMartin: Actully, I would expect it would be faster to check for which zone the circle's center is in: below, below-right, right, up-right, etc. |
| 20:29 | <&McMartin> | Only if it saves more than two branches. |
| 20:29 | <&McMartin> | Because that's two branches each time~ |
| 20:29 | <@macdjord> | Then you either just check the distance tot he nearest corner, or find the nearest point on the correct line. |
| 20:29 | <@macdjord> | Oh, right. |
| 20:29 | <&McMartin> | That's "I'll have to write it and see" |
| 20:31 | <&McMartin> | Yeah, this will work. |
| 20:31 | <&McMartin> | And I'm starting to see what the architecture will look like |
| 20:31 | <&McMartin> | ... it's going to look like multiple dispatch -_- |
| 20:31 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has quit [[NS] Quit: Kernel upgrade] |
| 20:35 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has joined #code |
| 20:35 | | mode/#code [+ao McMartin McMartin] by ChanServ |
| 20:36 | <&McMartin> | Correction: Previous algorithm assumes circle's center is outside of the rectangle. |
| 20:36 | <&McMartin> | If it's inside the rectangle, you're done~ |
| 20:38 | <&McMartin> | So that means it is basically macdjord's suggestion but being axis-aligned lets you find "closest point" with basically no math. |
| 20:38 | <&McMartin> | It's basically a min or max operation, depending |
| 20:57 | | kourbou_ [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 20:57 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [NickServ (RECOVER command used by kourbou_)] |
| 20:57 | | kourbou_ is now known as kourbou |
| 20:58 | | kourbou|phone [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 21:06 | | catadroid [catalyst@Nightstar-sbtq4p.dab.02.net] has joined #code |
| 21:11 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 21:11 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 21:17 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [NickServ (RECOVER command used by kourbou|phone)] |
| 21:17 | | kourbou|phone is now known as kourbou |
| 21:28 | | Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [Connection closed] |
| 21:54 | | catadroid [catalyst@Nightstar-sbtq4p.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 22:00 | <&McMartin> | Vorn! We were performing feats of geometry earlier. It would be awesome if you could sanity-check it. |
| 22:13 | | Kindamoody|afk is now known as Kindamoody |
| 22:14 | <~Vornicus> | Tallyho |
| 22:15 | <~Vornicus> | Okay, first claim |
| 22:16 | | * abudhabi presses it. |
| 22:17 | <~Vornicus> | "I believe, but am not certain, that I merely have to check to see if any corner of the AABB is inside the circle, or if any "axis extremum" of the circle is inside the rectangle." - not true, consider a long thin rectangle passing through a circle but *not* through the center |
| 22:18 | <~Vornicus> | ah, you found it |
| 22:23 | <~Vornicus> | okay, aabb vs circle collision. 0. xlo - r <= xcen <= xhi + r and ylo - r <= ycen <= chi + r; outside of this is definitely false |
| 22:23 | <~Vornicus> | er, yhi, not chi |
| 22:24 | <~Vornicus> | 1. if 0, xlo <= xcen <= xhi or ylo <= ycen <= yhi; this true means collision is yes |
| 22:25 | <&McMartin> | That's "0. If the aabb that circumscribes the circle doesn't intersect the aabb, no collision; 1. If the center of the circle is inside the aabb, there is a collision", yes? |
| 22:26 | <~Vornicus> | 1 is a little more loose |
| 22:26 | <~Vornicus> | note the *or* |
| 22:27 | <~Vornicus> | It covers the cross shape made by pushing the edges of the aabb out by r |
| 22:27 | <&McMartin> | Oh, I see. The idea is that we know the center is at least r from some edge at this point, so if it projects onto any line segment, it's in. Right. |
| 22:28 | <&McMartin> | That leaves the "webbing" |
| 22:28 | <&McMartin> | between the branches of the cross |
| 22:29 | <~Vornicus> | if xcen < xlo: xoffset = xlo - xcen; else: xoffset = xcen - xhi, and the y equivalent |
| 22:29 | <~Vornicus> | then if xoffset**2 + yoffset**2 <= r**2, you've hit a corner |
| 22:32 | <~Vornicus> | mmmm, with images: http://www.metanetsoftware.com/technique/tutorialA.html#section3 |
| 22:32 | <&McMartin> | That was pretty clear, really. |
| 22:33 | <&McMartin> | 2 is "if you're outside of the cross, check the closest corner with the point-in-circle method" |
| 22:33 | <~Vornicus> | Yep. |
| 22:33 | <&McMartin> | (I need point-collision anyway to do mouseclicks right) |
| 22:34 | <&McMartin> | Which turns into six short-circuiting tests followed by the actual comparison. |
| 22:41 | <~Vornicus> | Using SAT the way metanet does also gives you the shortest path out |
| 22:45 | <&McMartin> | This is for Monocle's object collision. I want to get to a boolean value as rapidly as possible |
| 22:48 | | Checkmate [Z@Nightstar-pdi1tp.customer.tdc.net] has joined #code |
| 22:48 | | mode/#code [+o Checkmate] by ChanServ |
| 23:05 | | Kindamoody is now known as Kindamoody[zZz] |
| 23:17 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has quit [Ping timeout: 121 seconds] |
| 23:20 | | kourbou [kourbou@Nightstar-deqg8j.fbx.proxad.net] has joined #code |
| 23:26 | | Vornotron [vorn@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 23:29 | | Vornicus [vorn@Nightstar-uhn82m.ct.comcast.net] has quit [Ping timeout: 121 seconds] |
| 23:48 | | catadroid [catalyst@Nightstar-p9cb1l.range86-174.btcentralplus.com] has joined #code |
| --- Log closed Sun May 03 00:00:11 2015 |