| --- Log opened Tue Nov 22 00:00:48 2016 |
| 00:13 | | himi [sjjf@Nightstar-dm0.2ni.203.150.IP] has joined #code |
| 00:13 | | mode/#code [+o himi] by ChanServ |
| 00:21 | | Kindamoody is now known as Kindamoody[zZz] |
| 00:46 | | Derakon[AFK] is now known as Derakon |
| 02:20 | | Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has joined #code |
| 03:15 | <@celticminstrel> | My code is heavily bound to a specific map size. I don't think I can readily unbind it. Oh well. |
| 03:25 | <~Vornicus> | celmin: another pair of eyes may help |
| 03:26 | <@celticminstrel> | I suppose it could. |
| 03:26 | <@celticminstrel> | The thing is, the areas are mapped onto a rectangular prism, and I was unable to think of a general way of doing the ends. |
| 03:27 | <@celticminstrel> | Okay, so if I'm going to show it off I have to decide between opening a git or throwing it in a pastebin... |
| 03:27 | <@celticminstrel> | s/git/github/ |
| 03:27 | <@celticminstrel> | I probably would've eventually anyway, but not until I had something that could be described as playable. |
| 03:29 | <&ToxicFrog> | You should be using some sort of version control anyways, which makes putting it on github/bitbucket/etc easy |
| 03:29 | <@celticminstrel> | Yeah, I just made the initial commit in git a few minutes ago. |
| 03:30 | <@celticminstrel> | I forget if you can rename github repositories. |
| 03:30 | <&ToxicFrog> | I think s |
| 03:30 | <&ToxicFrog> | *so |
| 03:30 | <@celticminstrel> | Okay yeah, looks like it. |
| 03:31 | <@celticminstrel> | Because this project is currently nameless, so if I put it there I'd have to rename it once I came up with a name. |
| 03:31 | | * Vornicus is bad at version control partly because he tends to be in places where even his smoke tests are red for long periods of time |
| 03:32 | | * celticminstrel isn't super-worried about it being bound to a specific size, but that does make it harder if I decide to make the map bigger at some point. |
| 03:34 | <@celticminstrel> | So here you go: https://github.com/CelticMinstrel/wip-roguelike/blob/master/panels/Map.js |
| 03:34 | <@celticminstrel> | The border stuff, mainly in Map(). |
| 03:34 | <~Vornicus> | oh wild. |
| 03:34 | <@celticminstrel> | The number of regions is hard-coded and not easy to change. |
| 03:35 | <@celticminstrel> | Wild huh. |
| 03:36 | <~Vornicus> | I thought, when you said "rectangular prism", you meant the inside of such a thing, not the *surface* |
| 03:37 | <@celticminstrel> | Heh... |
| 03:38 | <~Vornicus> | Now *this* is my neighborhood. |
| 03:39 | <&Derakon> | A roguelike where you can walk on the walls and ceilings and the procgen is in all three dimensions. |
| 03:39 | <~Vornicus> | Typically the way I see this type of thing done is *three* coordinates - one tells you what zone you're in, and the other two coordinates tell you where in that zone you are. |
| 03:40 | <@celticminstrel> | Not quite like that, Derakon. |
| 03:40 | <~Vornicus> | Your global topology is then set by which zone edges align with which, in what directions |
| 03:41 | <&Derakon> | CM: no, that was random neuron firings |
| 03:41 | <@celticminstrel> | Vornicus: So basically you wouldn't make any attempt to make the zones' arrangement in memory reflective of the topology. |
| 03:41 | <@celticminstrel> | ? |
| 03:41 | <@celticminstrel> | Derakon: Fair enough. :p |
| 03:42 | <~Vornicus> | celmin: nope. Way easier to have six or twelve or twenty or what-have-you square arrays. |
| 03:46 | <@celticminstrel> | But is there a general way to set up the zone edges? |
| 03:46 | <@celticminstrel> | Or would it just have to be done manually for whatever size you choose? |
| 03:48 | <@celticminstrel> | I don't think my code is heavily dependent on the fact that zones are currently addressed by two coordinates. |
| 03:48 | <~Vornicus> | You can have a data table that handles it - basically you can say -- for side 2, if you go off the high y side, you end up on side 0, on the low x side, with reversed coordinate |
| 03:48 | <@celticminstrel> | But that data table would still need to be tailored for your chosen size? |
| 03:49 | <~Vornicus> | For your chosen topology but not for your chosen size |
| 03:49 | <@celticminstrel> | Hmm... |
| 03:49 | <~Vornicus> | So if you decide to have, for instance, a cube shape, you can hard code that, and then a cube of *any* size will work. |
| 03:50 | <@celticminstrel> | I have more of a pillar than a cube. |
| 03:51 | <~Vornicus> | Still possible. A little more annoying |
| 03:51 | <&Derakon> | Oh, today I accidentally wrote "temfile" instead of "tempfile". |
| 03:51 | <&Derakon> | And got distracted for a bit by thinking about Undertale. |
| 03:51 | <&Derakon> | Temmie should probably not code. |
| 03:52 | <~Vornicus> | But you can plop that in the setup & motion table too - what your sizing DOFs are. |
| 03:53 | <~Vornicus> | (I wonder what they are for a rhombic dodecahedron...) |
| 03:53 | <@celticminstrel> | DOF? |
| 03:53 | <~Vornicus> | Degrees of Freedom |
| 03:54 | <~Vornicus> | Different topologies will have different degrees of freedom because you're forced to have the opposite edges of each face have the same number of cells. |
| 03:58 | | * celticminstrel nod... |
| 03:58 | <~Vornicus> | So in a cube you have three degrees of freedom - there are three different tuneable dimensions. |
| 04:00 | <~Vornicus> | (yours are set to 2,2,8) |
| 04:00 | <@celticminstrel> | ...wait, what about the 10? |
| 04:00 | <@celticminstrel> | Shouldn't it be 2,2,10? |
| 04:00 | <~Vornicus> | sorry, 2,2,10 |
| 04:01 | <@celticminstrel> | Since the 8 is dependent on the 2,2. |
| 04:01 | <~Vornicus> | I wasn't looking at it when I spoke |
| 04:01 | <@celticminstrel> | Fair enough |
| 04:01 | <~Vornicus> | ah, rd has 3 as well. |
| 04:03 | <~Vornicus> | rt has 5, looks like |
| 04:04 | <@celticminstrel> | rt? |
| 04:04 | <~Vornicus> | rhombic triacontahedron |
| 04:05 | <~Vornicus> | oh, no, six. |
| 04:06 | <@celticminstrel> | So I guess one step in making it general would be figuring out how to calculate the 8 from the 2,2... |
| 04:06 | <@celticminstrel> | Though I might not need the 8 when building it up. |
| 04:06 | <@celticminstrel> | But it'll probably be needed somewhere for something. |
| 04:07 | <~Vornicus> | and I can't add either on the rd, it's 4 there |
| 04:07 | <~Vornicus> | (rhombic dodecahedron) |
| 04:08 | <@celticminstrel> | Yeah, I got that one since you mentioned it earlier. |
| 04:11 | <~Vornicus> | there's way more of these than I expected actually. I'd forgotten about the dyakis dodecahedron (which actually has 24 sides) for instance. |
| 04:13 | | macdjord|wurk is now known as macdjord |
| 04:14 | <~Vornicus> | ...which appears at first blush to have also 6 DOF in this sense, and even better there's pairs of such that don't intersect. |
| 04:15 | <~Vornicus> | ...and is actually topologically equivalent to a deltoidal icositetrahedron, which itself is best represented by a *cube* because of the way the 3- and 4-corner points |
| 04:17 | <~Vornicus> | (welcome to one of my long-time interests) |
| 04:18 | <@celticminstrel> | Heh. |
| 04:18 | | * celticminstrel is currently adding more named constants for readability and stuff. |
| 04:25 | <@celticminstrel> | I think I'm starting to see how to make all those assignments as a loop instead of a hard-coded block... not going to do it quite yet though. |
| 04:30 | <@celticminstrel> | Uh, I forget... x+h*y or x+w*y...? |
| 04:32 | <~Vornicus> | x + w*y |
| 04:42 | | * Derakon hacks at Lua for about half an hour before running out of brain. |
| 04:42 | <&Derakon> | Maybe 50/50 not being used to Lua and not having much in the way of brain reserves. |
| 04:56 | | Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [Connection closed] |
| 05:02 | | Derakon is now known as Derakon[AFK] |
| 05:52 | <@celticminstrel> | My "converge" coordinate transformation is not working. |
| 05:53 | <@celticminstrel> | Or to be more precise, it's working for the north but not for the south. |
| 05:53 | <@celticminstrel> | In the south the transformed coordinate is (SIZE-x-1,SIZE-y-1), where (x,y) is what it should be. |
| 05:58 | <@celticminstrel> | I guess I can just have different transform functions for this case... |
| 06:06 | <~Vornicus> | this is still the thing where you're laying it out in an array? |
| 06:07 | <@celticminstrel> | Well, it is laid out in an array, yes. I'm not sure that's relevant to this particuar question though. |
| 06:08 | <@celticminstrel> | This is the function that transforms an out-of-bounds coordinate in one zone to an in-bounds coordinate in the adjacent zone. |
| 06:08 | <@celticminstrel> | Different boundaries use different functions. |
| 06:09 | <@celticminstrel> | Specifically it was the two *_converge_bmap functions. |
| 06:09 | <@celticminstrel> | I added two additional functions to use for the south pole boundaries, and it seems to work. |
| 06:10 | <~Vornicus> | I was wondering if you were still on that or trying my method |
| 06:10 | <@celticminstrel> | I haven't done anything about that yet. Maybe tomorrow. Currently working on generating a faux-longitude value. |
| 06:10 | <@celticminstrel> | (Already did latitude) |
| 06:12 | <@celticminstrel> | Though I might just sleep and do both tomorrow. |
| 06:13 | <@celticminstrel> | I've figured out that I need to project an interior point onto the perimeter of a square. No idea how to do that yet. |
| 06:16 | | himi [sjjf@Nightstar-dm0.2ni.203.150.IP] has quit [Ping timeout: 121 seconds] |
| 06:24 | | celticminstrel [celticminst@Nightstar-h4m24u.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 08:51 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 08:51 | | mode/#code [+o himi] by ChanServ |
| 09:08 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 09:17 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 09:17 | | mode/#code [+o himi] by ChanServ |
| 09:28 | | Kindamoody[zZz] is now known as Kindamoody|afk |
| 11:30 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Ping timeout: 121 seconds] |
| 13:01 | | catadroid [catalyst@Nightstar-hl7rfp.dab.02.net] has joined #code |
| 13:05 | | Emmy-werk [NSkiwiirc@Nightstar-41pbej.static.chello.nl] has joined #code |
| 13:05 | | mode/#code [+o Emmy-werk] by ChanServ |
| 13:16 | | Emmy-werk [NSkiwiirc@Nightstar-41pbej.static.chello.nl] has quit [[NS] Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] |
| 13:17 | | Emmy-werk [NSkiwiirc@Nightstar-41pbej.static.chello.nl] has joined #code |
| 13:17 | | mode/#code [+o Emmy-werk] by ChanServ |
| 14:31 | | catadroid` [catalyst@Nightstar-qaemkg.dab.02.net] has joined #code |
| 14:32 | | catadroid [catalyst@Nightstar-hl7rfp.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 15:20 | | catadroid` is now known as catadroid |
| 15:35 | | macdjord is now known as macdjord|wurk |
| --- Log opened Tue Nov 22 17:29:31 2016 |
| 17:29 | | TheWatcher [chris@GlobalOperator.Nightstar.Net] has joined #code |
| 17:29 | | Irssi: #code: Total of 35 nicks [32 ops, 0 halfops, 0 voices, 3 normal] |
| 17:29 | | mode/#code [+o TheWatcher] by ChanServ |
| 17:29 | | Reiver [quassel@Nightstar-ksqup0.co.uk] has joined #code |
| 17:29 | | mode/#code [+ao Reiver Reiver] by ChanServ |
| 17:30 | | Irssi: Join to #code was synced in 55 secs |
| 18:34 | | catadroid [catalyst@Nightstar-qaemkg.dab.02.net] has quit [[NS] Quit: Bye] |
| 19:10 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has joined #code |
| 19:10 | | mode/#code [+o Emmy] by ChanServ |
| 20:27 | <&[R]> | <enderst> https://twitter.com/docsmooth/status/800639999364448256 |
| 20:27 | <&McMartin> | http://gizmodo.com/hedgehog-based-authentication-is-the-only-way-to-be-tru-17892 45671 |
| 20:31 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 20:31 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 20:36 | <@celticminstrel> | XD |
| 21:52 | <@celticminstrel> | Why does 'git add -N' require the file to exist? :/ |
| 21:55 | <&[R]> | Because git tracks files |
| 21:56 | <@celticminstrel> | But that means I can't 'edit xxx; git add -N xxx' |
| 21:56 | <@celticminstrel> | (Since edit doesn't create the file until saved.) |
| 21:57 | <&[R]> | Because edit deletes the file after it is done? |
| 21:57 | <&[R]> | Or did you mean `edit xxx & git add -N xxx`, which I'm not certain actually works. |
| 21:58 | <@celticminstrel> | It doesn't, because this edit just tells TextWrangler to open a document and then returns. |
| 21:58 | | macdjord|wurk [macdjord@Nightstar-cd2k4t.mc.videotron.ca] has quit [Connection reset by peer] |
| 21:58 | | macdjord|wurk [macdjord@Nightstar-cd2k4t.mc.videotron.ca] has joined #code |
| 21:58 | | mode/#code [+o macdjord|wurk] by ChanServ |
| 21:58 | <@celticminstrel> | Well, I suppose it would probably still work, but there's no point. |
| 22:00 | <&[R]> | Ah |
| 22:00 | <&[R]> | Eitherway, I don't think git likes to track things that don't exist yet, so you're best off making your edit shell script touch the file first. |
| 22:01 | <@celticminstrel> | It's not a shell-script, though I think it does have a flag to do that. Still annoying. |
| 22:01 | | Kindamoody|afk is now known as Kindamoody |
| 22:32 | <@celticminstrel> | For some reason ROT is inserting an extra space before the prime characters... except it's not really a space because it's simply not rendering anything there at all. |
| 22:33 | <&McMartin> | I have just been introduced to the phrase "server-farm-to-database-table" |
| 22:38 | <@celticminstrel> | Looks like the problem is ROT treating it as a full-width character. Why would it do that? |
| 22:39 | <@celticminstrel> | isFullWidth = (cc > 0xff && cc < 0xff61) || (cc > 0xffdc && cc < 0xffe8) && cc > 0xffee; |
| 22:39 | <@celticminstrel> | The prime is 0x2032, so it falls in that first range. |
| 22:40 | <@celticminstrel> | I bet a lot of other non-full-width characters fit in there too. |
| 22:41 | <&[R]> | Prime? |
| 22:41 | <@celticminstrel> | The prime symbol. |
| 22:42 | <@celticminstrel> | Used for minutes of arc in my particular case. |
| 22:42 | <~Vornicus> | celmin: uh, something seems wrong with the first range |
| 22:43 | <@celticminstrel> | Yes indeed. |
| 22:43 | <~Vornicus> | and also the &&, which is going to not catch the second range. |
| 22:43 | <@celticminstrel> | Very very wrong. |
| 22:43 | <@celticminstrel> | ...what? |
| 22:44 | <&McMartin> | Is cc signed by any chance |
| 22:44 | <@celticminstrel> | Oh, precedence? |
| 22:44 | <@celticminstrel> | McMartin: JS, so technically yes but probably doesn't matter. |
| 22:44 | <~Vornicus> | Precedence and also I don't think you want && in the first place |
| 22:44 | <&McMartin> | Oh wait no |
| 22:44 | <&McMartin> | Yeah, that last && looks like A Problem |
| 22:45 | <@celticminstrel> | Problem in that characters greater than FFEE aren't necessarily fullwidth? |
| 22:45 | <~Vornicus> | No |
| 22:45 | <~Vornicus> | Problem in that it makes you miss the entire second region of fullwidth characters |
| 22:45 | <&McMartin> | If any previous clause is true, that one is false and therefore the whole thing is false |
| 22:46 | <@celticminstrel> | Oh, hm... |
| 22:46 | <@celticminstrel> | isFullWidth definitely had a value of true for some cc, because commenting that out fixed my problem, though. |
| 22:46 | <~Vornicus> | for cc = 0xffe1, you get (true && false) || (true && true) && false = false, but it's supposed to come up true, being FULLWIDTH POUND SIGN |
| 22:48 | <~Vornicus> | You need to fix the start point of the first region, and use || instead of && to glom the third region on, and not everything beyond BMP is fullwidth; it starts, for one thing, with a bunch of ancient alphabets |
| 22:48 | <@celticminstrel> | Looks like && has higher precedence than || |
| 22:48 | <~Vornicus> | You're ignoring the problem |
| 22:48 | <~Vornicus> | It's *supposed to be || in the first place* |
| 22:48 | <@celticminstrel> | Okay. |
| 22:50 | <@celticminstrel> | I have no idea what is and is not fullwidth though... not sure if Character Viewer can tell me... do all fullwidth characters say so in the name? |
| 22:50 | <~Vornicus> | Because what it says now is "it has to be in fullwidth region 1, or it has to be in both fullwidth region 2 AND in the bottom-of-bmp control characters plus basic multilingual plane" |
| 22:50 | <~Vornicus> | There are 103 characters that say FULLWIDTH in their names. |
| 22:52 | <@celticminstrel> | Aren't all/most of the Chinese characters fullwidth too? |
| 22:52 | <~Vornicus> | They're all in the BMP and all in the two regions listed above, once you correct the start point for the first region. The third region is only a couple of control characters and then everything that's not in the BMP |
| 22:53 | <@celticminstrel> | Ah, I see, it should be FF00 not FF. |
| 22:55 | <@celticminstrel> | So, that plus changing the && to ||... |
| 22:55 | <~Vornicus> | beyond the BMP are quite a lot of decidedly not fullwidth characters |
| 22:55 | <@celticminstrel> | But from what you said it sounded like that last range includes some non-fullwidth. |
| 22:55 | <~Vornicus> | Lots of ancient languages, lots of mathematical symbols, a huge private use area... |
| 22:55 | <@celticminstrel> | So the CJK radicals are not fullwidth, I guess... |
| 22:56 | <&McMartin> | I would expect them to be combining |
| 22:57 | <@celticminstrel> | Not sure if it's worth fixing it for characters beyond BMP though... how many roguelikes will use things like Linear B anyway... |
| 22:57 | <~Vornicus> | What are you trying to "fix" here |
| 22:58 | <@celticminstrel> | The engine corrects for fullwidth characters by inserting a space before them. |
| 22:58 | <~Vornicus> | uh huh |
| 22:58 | <@celticminstrel> | Which looks bad when the character isn't fullwidth. |
| 22:59 | <@celticminstrel> | Though I wonder if it even looks good when it is fullwidth... |
| 22:59 | <@celticminstrel> | Given that it's skipping over that space rather than filling it with the background colour. |
| 23:01 | <~Vornicus> | Probably not. |
| 23:06 | <@celticminstrel> | So, just to be sure, those two are the only full-width ranges in BMP? |
| 23:07 | <@celticminstrel> | No Chinese characters or anything? |
| 23:07 | <@celticminstrel> | ff00..ff61 and ffdc..ffe8 |
| 23:08 | <~Vornicus> | right |
| 23:09 | <~Vornicus> | But I don't have any idea what it's really trying to do or why it thinks it must do this |
| 23:09 | <~Vornicus> | Unless-- wait, is it doing a translation step on the fullwidth characters? |
| 23:09 | <@celticminstrel> | No? |
| 23:09 | <@celticminstrel> | It's writing one character per cell. |
| 23:10 | <~Vornicus> | RIght, right -- if you put in a character like, say, one of the CJK ideographs, does it draw that as two cells wide? |
| 23:11 | <~Vornicus> | If so, and it's also pushing the fullwidth characters trhough a translation step that brings it down into, say, its basic latin equivalent, then cjk with embedded fullwidths will look funny unless you add spaces in |
| 23:12 | <~Vornicus> | Which would explain why only the fullwidth characters are getting this treatment: they're designed to be used with 2-cell-wide characters, but they're showing up as 1 cell wide and that makes it look weird. |
| 23:14 | <@celticminstrel> | It looks like it doesn't really work well with CJK ideographs. Before, they were treated as full-width, inserting an extra space; now they're squished together. That said, the font it's using is Courier New, so maybe with a more Chinese-oriented font it'd work better. |
| 23:15 | <@celticminstrel> | Here the ideographs simply don't fit in a cell. |
| 23:15 | <@celticminstrel> | The're not only wider but also taller. |
| 23:15 | <@celticminstrel> | ^They're |
| 23:16 | <~Vornicus> | Okay, this still makes a certain level of sense then: fullwidth characters -- you'd need to find this out from reading the stride widths I suspect, but you can safely assume all cjk ideographs -- need extra space or they push each other out |
| 23:22 | | catalyst [catalyst@Nightstar-bt5k4h.81.in-addr.arpa] has joined #code |
| 23:37 | | Derakon[AFK] is now known as Derakon |
| 23:59 | | Kindamoody is now known as Kindamoody[zZz] |
| --- Log closed Wed Nov 23 00:00:21 2016 |