| 00:04.49 | Noshei | lol |
| 00:06.35 | onsight666 | time to try and get 52 - 60 Q.Q |
| 00:07.07 | Noshei | Great Hunts are pretty decent xp |
| 00:08.01 | onsight666 | yeah i did quite a few on my main but thats back when i was a patron with the xp pots xD |
| 00:24.53 | *** join/#riftuidev Portalboat (62a0e1c6@gateway/web/freenode/ip.98.160.225.198) |
| 00:24.58 | Portalboat | quick question |
| 00:25.12 | Portalboat | can someone explain to me what pairs() does? |
| 00:26.38 | seebs | It iterates over a table in an unspecified order. |
| 00:27.06 | seebs | for key, value in pairs(t) do … end |
| 00:27.45 | seebs | if you had a table t["a"] = 1, t["b"] = 2, that would iterate through the loop once with key = "a", value = 1, once with key = "b", value = 2, but there is no guarantee as to which order. |
| 00:28.22 | Portalboat | huh, alright |
| 00:29.44 | Noshei | now ipairs() only works on sequentially indexed tables and it will always iterate 1 then 2 then 3, etc |
| 00:30.37 | Portalboat | If the order isn't constant, then how come with the QuestFinder code it always prints out the quests in the same order? |
| 00:30.47 | Portalboat | It uses a pairs() loop |
| 00:31.33 | Noshei | it isn't guarenteed to always be the same order, doesn't mean it can't be |
| 00:31.58 | Portalboat | hm |
| 00:32.08 | Portalboat | I'll have to run it a couple times, then |
| 00:32.37 | Noshei | ipairs() is guarenteed to always be the same order, though you will get strange resules if you don't have a proper table |
| 00:32.51 | Portalboat | yeah |
| 00:37.01 | Ivnedar | It'll come out in hash order most likely |
| 00:37.05 | Ivnedar | an internal value you don't see |
| 00:37.47 | Portalboat | no, it's in the same order |
| 00:38.54 | seebs | Well, yes. It's quite likely that, for a given Lua implementation, putting the same things in a table will produce the same order. |
| 00:38.57 | Portalboat | strange |
| 00:39.19 | Portalboat | But you said that it would return the pairs randomly, right? |
| 00:40.12 | seebs | Not randomly. |
| 00:40.16 | seebs | In an order which is not defined. |
| 00:40.48 | seebs | There's no particular *guarantee* that you get the same results one time after another. You might. You might not. |
| 00:41.07 | seebs | This is by contrast with, say, Ruby, in which if you populate a hash, and then iterate over it, you get things in insertion order. |
| 00:41.19 | Noshei | here is what the lua doc's say. |
| 00:41.19 | Noshei | pairs() - The pairs() function will allow iteration over key-value pairs. Note that the order that items are returned is not defined, not even for indexed tables. |
| 00:41.19 | Noshei | ipairs() - The order in which elements are returned is guaranteed to be in the numeric order of the indices, and non-integer keys are simply skipped. |
| 00:41.45 | Portalboat | oh, okay |
| 00:41.55 | Portalboat | the order isn't DEFINED by the user |
| 00:42.00 | Portalboat | but it isn't random either |
| 00:42.01 | Portalboat | i get it |
| 00:42.35 | Noshei | well it isn't defined by anything really |
| 00:43.08 | Noshei | it is simply undefined |
| 00:44.13 | Noshei | http://lua-users.org/wiki/ForTutorial < that has some good informations |
| 00:44.42 | seebs | What "not defined" here means is not just not defined by the user, but that the language explicitly does not specify it. |
| 00:44.58 | Noshei | that site in general is very helpful |
| 00:45.05 | onsight666 | Servers going down in 15 Est downtime 3 Hrs |
| 00:45.08 | Ivnedar | each implementation of lua can do it differently, it could change between versions of lua, etc |
| 00:45.50 | seebs | So basically, there is no guarantee at all that you'll get the same order. You might, and for most implementations, it's probably common that you will. Sometimes. But next week's hotfix, maybe Zorba's checked in a bug fix to the Lua implementation and suddenly the order changes. |
| 00:45.50 | Noshei | i take it your EU? |
| 00:45.54 | onsight666 | indeed |
| 00:46.04 | onsight666 | < Scottish.. like a baus |
| 00:47.17 | Noshei | nice, I was in Scotland the first time I ever got shit face wasted |
| 00:47.40 | onsight666 | We do know how to drink ;) |
| 00:47.41 | onsight666 | XD |
| 00:48.00 | Noshei | I wasn't actually drinking with any scottish people though |
| 00:48.15 | onsight666 | aww, then you were doing it wrong xD |
| 00:48.31 | onsight666 | where in scotland did you visit? |
| 00:48.42 | Noshei | well I was drinking with some random people i met in the hostel i was staying in. |
| 00:50.06 | Noshei | Started in Aberdeen, went to a couple other places south on my way to Edinburgh |
| 00:50.50 | onsight666 | ah nice, im from glasgow myself but i used to live in edinburgh :) |
| 00:53.15 | Noshei | thats where I was when I got wasted, was at a pub maybe 3 blocks north of the castle (downhill, I think that was north) and then I had to walk back to the hostle, further down hill from the pub |
| 00:55.53 | Noshei | oh looks like that is actually more east |
| 00:56.47 | onsight666 | i think i know the hostel your on about |
| 00:58.21 | onsight666 | the castle rock hostel if memory serves |
| 00:58.57 | Noshei | might have been, but there were like 5 or 6 that I saw in the area |
| 01:01.53 | onsight666 | aaand the servers down Q.Q |
| 01:02.18 | Noshei | could switch to the us servers and play there |
| 01:02.41 | onsight666 | wouldn't i need to DL the US client? |
| 01:02.53 | Noshei | nope, same client |
| 01:03.01 | Noshei | in the launcher you can change it |
| 01:03.03 | onsight666 | nice.. |
| 01:03.15 | onsight666 | didnt think you could change the launcher settings xD |
| 01:03.18 | Noshei | just hit settings and change region |
| 01:03.35 | onsight666 | now i have something to do for 3 hours xD |
| 01:24.34 | Ivnedar | woah |
| 01:24.44 | Ivnedar | cqp decay starts at 48 hours now instead of 23! |
| 01:24.49 | Noshei | yeah |
| 01:24.55 | Noshei | which is fucking amazing |
| 01:27.31 | Ivnedar | a few nice QoL fixes in that patch |
| 01:27.40 | Ivnedar | dendrome quest givers not being dead 99% of the time is nice |
| 01:29.11 | Noshei | gah I have 50 mins left on my 80% flasks and I don't want to do any more SH's |
| 01:34.44 | Noshei | only 12k rep to go with torvan |
| 01:34.51 | Ivnedar | nice! :D |
| 01:35.03 | Ivnedar | Wish I had time to finish my factions |
| 01:35.11 | Ivnedar | EA/Achyati to go |
| 01:35.37 | Noshei | it is really fast for torvan with vials and SH's |
| 01:36.13 | Ivnedar | yeah |
| 01:38.34 | *** join/#riftuidev Portalboat (62a0e1c6@gateway/web/freenode/ip.98.160.225.198) |
| 01:38.53 | Portalboat | Is there any way to make text look like the text does on the quest log? |
| 01:39.07 | Portalboat | As in, having a drop shadow or whatever it has and not just pure text? |
| 01:39.32 | Noshei | :SetGlowEffect() |
| 01:40.09 | Noshei | http://www.riftui.com/downloads/info374-GlowTest.html < if you want to play with how :SetGlowEffect() looks easily |
| 01:41.20 | Portalboat | Can you call that with a Text element? |
| 01:41.31 | Portalboat | I don't see the function on the API index. |
| 01:42.52 | Noshei | test is the only frame type that can use it. |
| 01:43.52 | Portalboat | So it would just be Text:SetGlowEffect? (sorry for the confusion, it's not showing up here http://portal.wildtide.net/RiftAPI) |
| 01:44.16 | Noshei | http://pastebin.com/GLAHrbr1 |
| 01:44.43 | Portalboat | alright, thanks |
| 01:44.50 | Ivnedar | http://www.seebs.net/rift/live/Text_index.html#Text:SetEffectGlow |
| 01:44.59 | Portalboat | that must be a bit out of date then |
| 02:05.44 | Portalboat | You would pass a table for that function, right? I made a table for the values, but it's returning an error. "Unknown tags found in table." |
| 02:06.51 | Noshei | so the keys in the table have to be the exact same as what is in the documentation, including case |
| 02:07.11 | Noshei | Text:SetEffectGlow({colorR = 0.91, colorG = 0.424, colorB = 0.098, strength = 5, blurX = 1.5, blurY = 1.5}) |
| 02:07.20 | Noshei | as an example |
| 02:07.24 | Portalboat | right, those are keys |
| 02:07.30 | Portalboat | Bleh. |
| 02:07.44 | Portalboat | I thought it was indexed. |
| 02:08.03 | Portalboat | For some reason. |
| 02:08.10 | Portalboat | wheee, c++ |
| 02:09.03 | Ivnedar | you don't have to specify all of them either |
| 02:09.26 | Portalboat | Yeah, I got that. |
| 02:23.18 | Portalboat | There wouldn't happen to be a list of text colors, would there? (as in, the exact colors for quest log text, item types, etc.) |
| 02:24.28 | Noshei | I don't believe so, but you can always take a screenshot and use something like paint to sample the color and get a code for it |
| 02:25.41 | Portalboat | That's what I've been doing, but it doesn't seem quite right. Maybe it's the font size difference. |
| 02:27.00 | Ivnedar | ZorbaGama made a post recently |
| 02:27.08 | Ivnedar | the art group gave him a lot of them |
| 02:27.33 | Portalboat | Where is it? |
| 02:27.40 | Portalboat | The API forum? |
| 02:27.48 | Ivnedar | yeah, looking |
| 02:27.57 | Noshei | http://forums.riftgame.com/technical-discussions/addon-api-development/333518-official-addon-information-station.html#post4175070 |
| 02:28.03 | Ivnedar | there ya go :) |
| 02:28.32 | Portalboat | yep, just found it |
| 02:32.07 | Portalboat | hm |
| 02:32.11 | Portalboat | can't seem to find it |
| 02:32.15 | Portalboat | unless I'm blind. |
| 02:33.58 | Noshei | <PROTECTED> |
| 02:34.02 | Noshei | one of those |
| 02:34.29 | Portalboat | Tried both of those. #FF9730 was orange, and #FCF8CB is too pale. |
| 02:34.30 | Portalboat | Oh well |
| 02:34.45 | *** join/#riftuidev Baanano (~chatzilla@156.203.221.87.dynamic.jazztel.es) |
| 02:39.01 | *** join/#riftuidev onsight666_ (~onsight66@cpc15-uddi22-2-0-cust106.20-3.cable.virginmedia.com) |
| 03:01.28 | Portalboat | Huh, that's strange. |
| 03:01.43 | Portalboat | Apprently the max background color for this frame is yellow? |
| 03:01.57 | Portalboat | (SetBackgroundColor(255, 255, 255, 0.25)) |
| 03:03.23 | Noshei | SetBackgroundColor(r,g,b,a) is only meant to use values from 0 to 1. |
| 03:03.50 | Portalboat | Oh, okay. |
| 03:49.10 | *** join/#riftuidev Jydan (marcel@D57DD135.static.ziggozakelijk.nl) |
| 03:56.09 | *** join/#riftuidev seebs (~seebs@97-92-17-178.dhcp.stcd.mn.charter.com) |
| 03:56.10 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 03:59.56 | *** join/#riftuidev Jydan (marcel@D57DD135.static.ziggozakelijk.nl) |
| 04:49.57 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 04:49.57 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 08:27.42 | *** join/#riftuidev Kheldar (~Kheldar@217.108.40.234) |
| 09:19.00 | ZorbaGama | http://www.twitch.tv/zorbathut back to streaming again, finally |
| 09:19.48 | Ivnedar | welcome back! :) |
| 10:55.54 | *** join/#riftuidev Adelea (~martin@109.224.129.157) |
| 11:10.17 | *** join/#riftuidev Kheldar_ (~Kheldar@ANantes-557-1-195-193.w2-8.abo.wanadoo.fr) |
| 17:08.42 | seebs | I have this sort of half-baked notion that it might be possible to optimize some code paths a lot with a way for Inspect.*.Detail() to specify subsets of information they actually want to look at, meaning they wouldn't have to compute (or allocate storage for) other items. |
| 17:09.29 | seebs | e.g., if all I want is the names and locations of a bunch of units, something like Inspect.Unit.Detail({ group01 = true, group02 = true, …}, { name = true, coordX = true, coordY = true, coordZ = true }) |
| 17:27.56 | *** join/#riftuidev Kheldar (~Kheldar@ANantes-557-1-195-193.w2-8.abo.wanadoo.fr) |
| 18:00.23 | *** join/#riftuidev Portalboat (62a0e1c6@gateway/web/freenode/ip.98.160.225.198) |
| 18:01.12 | Portalboat | So I've got a really strange bug. |
| 18:01.57 | Portalboat | "_qf.Background[int]:EventAttach(Event.UI.Input.Mouse.Cursor.In, function() _qf.Background[int]:SetBackgroundColor(0.2, 0.2, 0.2, 0.25) end, "")" |
| 18:02.25 | Portalboat | That's what's throwing the error, with 'attempted to index a nil value' |
| 18:02.47 | Portalboat | The strange thing is, throwing in a _qf.print(string.format(int))" |
| 18:02.58 | Portalboat | doesn't throw the error, and it works fine. |
| 18:05.29 | seebs | Well, the question is which of several things in there is the index of a nill value. |
| 18:05.36 | seebs | nil. |
| 18:05.45 | Portalboat | Yeah. |
| 18:05.48 | seebs | Every . and every [] and every : is an index of something. |
| 18:05.55 | Portalboat | Yep |
| 18:06.21 | seebs | So, is _qf.Background[int] a non-nil value? I'm assuming that the actual contents of the function aren't executing yet, so we can ignore that. |
| 18:06.23 | Portalboat | The only thing I changed is the "_qf.Background[int]:SetBackgroundColor(0.2, 0.2, 0.2, 0.25)" between throwing the error and not throwing the error, so it has to be something in there. |
| 18:06.46 | seebs | So you have _qf.Background[int]:EventAttach(Event.UI.Input.Mouse.Cursor.In, function() do_something end, "") |
| 18:06.56 | Portalboat | yep |
| 18:07.05 | seebs | Because if the contents of the function mattered, you wouldn't be seeing an exception here, but later on mouse-in. |
| 18:07.35 | Portalboat | I know it isn't int, because that's what the print function was for. |
| 18:07.40 | seebs | So, there's two possibilities, one's that _qf.Background[int]:EventAttach is indexing a nil value, the other is that Event.UI.Input.Mouse.Cursor.In is. |
| 18:07.57 | seebs | The print() isn't doing something that makes much sense to me. |
| 18:07.58 | Portalboat | But I don't think it's either. |
| 18:08.06 | seebs | string.format(x) is almost always wrong. |
| 18:08.12 | Portalboat | Well |
| 18:08.23 | seebs | The only purpose of string.format is to expand %-sequences in a string, which implies corresponding arguments. |
| 18:08.25 | Portalboat | It's working fine and doing what I expected it to do. |
| 18:08.39 | Portalboat | <PROTECTED> |
| 18:08.57 | seebs | In that case, you can get away with _qf.print(int) most likely. |
| 18:09.16 | seebs | My guess would be that _qf.Background[int] is a nil value, or that _qf.Background is a nil value. |
| 18:09.38 | seebs | You might find the dump() call useful. dump(_qf.Background) should show you a textual representation. |
| 18:09.39 | Portalboat | But why would it change all of a sudden if it was working fine with the print function? |
| 18:10.05 | seebs | int can be a non-nil value, but _qf.Background[int]:Foo() indexing a nil value. |
| 18:10.12 | Baanano | you were printing a number, it doesn't imply the number is actually used as an index |
| 18:10.24 | seebs | t = { x = { y = 1 } } |
| 18:10.33 | seebs | print(t.x.y) => 1 |
| 18:10.41 | seebs | z = "x" |
| 18:10.52 | seebs | print(t[x].y) => 1 |
| 18:10.56 | seebs | Er t[z].y |
| 18:11.00 | seebs | z = "foo" |
| 18:11.04 | seebs | print(z) => "foo" |
| 18:11.12 | seebs | print(t[z].y) => attempt to index a nil value |
| 18:11.17 | seebs | Because t[z] is nil. |
| 18:11.31 | seebs | So my guess is that qf.Background[int] is nil, even though int isn't nil. |
| 18:11.51 | seebs | And then qf.Background[int]:EventAttach is attempting to index a nil value. |
| 18:11.52 | Portalboat | But it's working fine for the EventAtttach. |
| 18:11.57 | seebs | Is it? |
| 18:12.01 | Portalboat | Yeah |
| 18:12.08 | seebs | I thought this was an error thrown at the point where the EventAttach is called. |
| 18:12.09 | Portalboat | If I remove the "_qf.Background[int]:SetBackgroundColor(0.2, 0.2, 0.2, 0.25)" |
| 18:12.24 | seebs | Ahh. |
| 18:12.26 | Portalboat | and substitute the print function, it works fine |
| 18:12.33 | seebs | In that case, I have a suspicion. |
| 18:12.38 | seebs | Which is that int is a global. |
| 18:12.44 | seebs | And thus, that its value is changing. |
| 18:12.51 | seebs | Or that _qf.Background[int] is changing at some later point. |
| 18:13.16 | Portalboat | That's....quite possible. |
| 18:13.28 | seebs | Keep in mind, unless int is a local (and thus becomes a part of the closure of the inline function), it's just a variable whose value can change whenever. |
| 18:13.32 | Portalboat | Yeha |
| 18:13.34 | Portalboat | Yeah* |
| 18:13.47 | Portalboat | (and apprently all addons get the same data scope) |
| 18:13.53 | seebs | The idiom of doing stuff like b = 3; function(a) do_something(b, a) end |
| 18:13.58 | seebs | only really works if b is local. |
| 18:14.08 | Portalboat | Are arguments local? |
| 18:14.13 | seebs | Well, it's Lua. Variables are global unless declared local, which I dislike, but it's the way they work. |
| 18:14.14 | Portalboat | Because that's what it's passed as. |
| 18:14.17 | seebs | Usually, yes. |
| 18:14.25 | seebs | Can you paste the whole chunk to pastebin or something? |
| 18:14.52 | Portalboat | yeah, hang on |
| 18:15.56 | Portalboat | http://pastebin.com/ts8uhGjA |
| 18:15.58 | Portalboat | bleh, formatting |
| 18:16.13 | Portalboat | this is with the print function in there, and it works fine |
| 18:16.37 | Portalboat | The line is supposed to be _qf.Background[int]:EventAttach(Event.UI.Input.Mouse.Cursor.In, function() _qf.Background[int]:SetBackgroundColor(0.2, 0.2, 0.2, 0.25) end, "") |
| 18:17.23 | seebs | ah-hah |
| 18:17.36 | seebs | <PROTECTED> |
| 18:17.37 | Portalboat | yes? |
| 18:17.48 | Portalboat | ? |
| 18:17.52 | seebs | Okay, presumably "int" can have more than one value, like, this can get called more than once, right? |
| 18:17.59 | Portalboat | yep |
| 18:18.01 | seebs | Say it gets called twice, once with int == "a" and once with int = "b" |
| 18:18.06 | Portalboat | yep |
| 18:18.13 | Portalboat | (that's what it's supposed to do) |
| 18:18.14 | seebs | First time through: You set _qf.Background to an empty table, then populate _qf.Background["a"]. |
| 18:18.22 | Portalboat | yep |
| 18:18.29 | seebs | Second time through: You set _qf.Background to an empty table, then populate _qf.Background["b"] |
| 18:18.35 | Baanano | "a" isn't there anymore |
| 18:18.38 | Portalboat | oh |
| 18:18.40 | seebs | Now what happens when you try to refer to _qf.Background["a"]? It's a nil value. |
| 18:18.45 | Portalboat | huh |
| 18:18.45 | seebs | The Lua idiom is: |
| 18:18.48 | seebs | x = x or {} |
| 18:18.59 | Portalboat | damn, that's obvious in hindsight |
| 18:19.00 | Portalboat | wow |
| 18:19.02 | seebs | which is "if x doesn't exist yet, make x into an empty table". |
| 18:19.19 | seebs | A note (not that most of the people here need it): |
| 18:19.26 | Portalboat | And to think I spent like an hour looking at this last night. |
| 18:20.00 | seebs | I spend a lot of time answering questions in chat. Generally what determines whether I keep answering the same person's questions is mostly their attitude. People who get insulting and defensive don't get more answers. People who say "hah, wow, I can't believe I did that", I usually keep talking to. |
| 18:20.27 | seebs | I have almost never met a competent developer who doesn't have moments like that. I've met a lot of incompetent software developers who never have that problem, though. |
| 18:20.42 | seebs | Because it's *never* their fault, it's *always* a design flaw in the language, bad documentation, or something else. |
| 18:20.53 | Portalboat | Oh, wow |
| 18:20.53 | Portalboat | yeah |
| 18:20.57 | seebs | … I actually do know one person I've never seen make a mistake like that, but he's… magic. |
| 18:21.03 | Baanano | yeah documentation is horrible |
| 18:21.10 | Portalboat | Just 'magic', huh? |
| 18:21.12 | seebs | Give him an arbitrarily hard problem, wait a while. |
| 18:21.22 | seebs | You will get the answer or an exhaustively-researched explanation of why it can't be done. |
| 18:22.04 | seebs | Like, I've been working with him to varying degrees for over ten years, and not only have I not found bugs in his code, I don't think anyone else has either. |
| 18:22.17 | Portalboat | it's like the what the fuck comment |
| 18:22.30 | seebs | He's on the very far opposite end of the programming-style spectrum from me, really. :) |
| 18:22.56 | Portalboat | have you heard of that? |
| 18:23.17 | Baanano | i haven't |
| 18:23.24 | seebs | I don't think so. |
| 18:23.46 | seebs | I collect fascinating comments from source code. Favorites include Bill Paul's realtek driver for FreeBSD. |
| 18:24.00 | Portalboat | It was something for Quake Arena, I think... |
| 18:24.52 | Portalboat | http://pastebin.com/29jMxRJT |
| 18:25.07 | seebs | heh |
| 18:25.13 | Portalboat | This guy needed to find a inverse square root for raytracing. |
| 18:25.25 | seebs | That's a reasonably famous value. I think it may even be in HAKMEM. |
| 18:26.07 | seebs | Guess not, it's slightly more recent. |
| 18:27.22 | Portalboat | awsome, it's working now |
| 18:27.42 | Portalboat | Thanks for the help. |
| 18:39.17 | Portalboat | There's no way to get frames to have borders or anything fancy like that, right? |
| 18:39.28 | seebs | RiftWindw |
| 18:39.32 | seebs | RiftWindow, I mean. |
| 18:39.51 | seebs | You can also do slightly less elaborate frames by making a frame in one color with another frame inside it in a different color. |
| 18:40.04 | Portalboat | Yeah, but I'm talking about a basic block of color. |
| 18:40.18 | Portalboat | Something that's defined as Frame. |
| 18:40.46 | Portalboat | hm |
| 18:40.47 | seebs | Yeah, there isn't really much for that. There's ways to add texturey things, but nothing super cool. |
| 18:40.50 | Portalboat | doesn't look like it. |
| 18:40.54 | Portalboat | Darn. |
| 18:41.11 | seebs | There's an idiom I've seen elsewhere where you have 8 textures, four corners and four sides, that are stretched and modified to fit a frame, but I don't think Rift has anything like that natively. |
| 18:41.22 | seebs | I'm not even sure whether we get UV coordinates, although probably we do, I just never looked. |
| 18:42.30 | Portalboat | lua doesn't care about whitespace, right? |
| 18:42.31 | Baanano | no we don't have that |
| 18:43.00 | seebs | So we have to do multiple textures, not texture maps? |
| 18:43.12 | seebs | Portalboat: Mostly true, although as always there's probably occasional exceptions. |
| 18:43.32 | Portalboat | Alright. |
| 18:43.37 | seebs | C is whitespace-insensitive except when it isn't. |
| 18:43.59 | seebs | x = - - y; is not the same as x=--y; |
| 18:46.22 | Baanano | well you can use texture maps combining a texture and a mask, i think wildtide does that in gadgets |
| 18:47.22 | Baanano | but you can't use coordinates to stretch different parts of a texture with different scales |
| 18:48.26 | seebs | Oh, I wasn't thinking of that. |
| 18:48.32 | seebs | Oh, wait, yes I was. |
| 18:49.00 | seebs | Well, what I'd been thinking of was basically treating a texture as 8 subtextures, and setting each of those subtextures as the texture for a different frame, and scaling those frames. |
| 18:49.26 | seebs | So I wasn't trying to stretch different parts of a single object at different scales, but trying to stretch different objects at different scales, where they all have the same texture as their backing. |
| 18:50.03 | Baanano | yeah but, at that point it's better to use 8 textures and drop the masks |
| 18:50.44 | seebs | The main reason it exists, I think, is that in the underlying system a single texture being treated as a grid of textures is often more efficient. |
| 18:51.56 | Baanano | but you'd need 8 texture frames + 8 mask frames, and we've been told repeteadly that masks are bad :P |
| 18:54.12 | Portalboat | godamnit i almost had that exact same problem again |
| 18:54.30 | Portalboat | defining things in a looped function is bad, portalboat |
| 18:55.48 | Baanano | don't worry, once you get used to it, lua is a cool language |
| 19:02.05 | Portalboat | and now it's throwing the same error again |
| 19:02.06 | Portalboat | bleh |
| 19:08.08 | Portalboat | Huh, that's strange. |
| 19:08.29 | Portalboat | When I do anything with my increment inside of the function that gets called, it doesn't work. |
| 19:09.05 | Portalboat | But as soon as I "outsource" it, (as in, define it in the previous function and increment it there as well) it works fine. |
| 19:16.17 | Baanano | pastebin it, please |
| 19:16.55 | *** join/#riftuidev Veritin (~martin@109.224.129.157) |
| 19:21.14 | Portalboat | It's not a problem, I'm just curious |
| 20:47.58 | *** join/#riftuidev Imhothar (miranda@ppp-188-174-82-122.dynamic.mnet-online.de) |
| 21:04.07 | *** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net) |
| 21:49.29 | ZorbaGama | seebs, for what it's worth, the upcoming Canvas lets you sort-of have access to UV coordinates |
| 21:57.17 | Imhothar | "sort-of"? |
| 22:00.12 | Noshei | any eta on canvas ZorbaGama? |
| 22:00.30 | ZorbaGama | you don't actually have access to UV coordinates, but what you *can* do is linearly transform the texture to wherever you want it |
| 22:01.07 | Imhothar | so, one matric per texture? not per vertex |
| 22:01.18 | ZorbaGama | noshei, sadly there was a branch SNAFU that means it'll be a bit delayed - I don't have a good estimate right now. Sorry :/ |
| 22:01.42 | ZorbaGama | imhothar, yep. it means that if you want multiple textures that don't stitch together perfectly, you'll need one canvas for each |
| 22:01.56 | Imhothar | what's the drawback? |
| 22:01.58 | ZorbaGama | that said, note that canvas is pretty dang fast |
| 22:02.09 | Imhothar | (except for not specifyong UV per vertex) |
| 22:02.18 | ZorbaGama | implementation time, I guess - if you're not familiar with matrix math, now is a good time to brush up on it ;) |
| 22:02.35 | Imhothar | I gues 2x2 matrices are sufficient? |
| 22:02.42 | Imhothar | unless I want perspective.. hmm |
| 22:02.48 | ZorbaGama | they're actually 3x2 for translation also |
| 22:02.54 | ZorbaGama | but I don't believe perspective is possible |
| 22:03.42 | Imhothar | sure it is, just take homogenous math and remove the z row |
| 22:03.47 | Imhothar | and column |
| 22:04.12 | Imhothar | then it's "sort-of" perspective |
| 22:04.23 | Imhothar | i.e. not affine |
| 22:05.25 | Imhothar | or we just do regular 4x4 matrices ;) |
| 22:05.47 | ZorbaGama | no z coordinate for positions, so 4x4 matrices would be kinda silly :) |
| 22:07.56 | seebs | I am a sort of mix of good at and bad at matrix math. |
| 22:08.14 | seebs | I could never find anything which explained the significance of the values in a 3x2 affine transform matrix. |
| 22:08.29 | Imhothar | well, why not? ;) |
| 22:08.38 | seebs | Finally I gave up. I wrote down the three obvious cases (translate, rotate, scale), multiplied them, and *poof* instantly it makes sense. |
| 22:08.55 | Imhothar | topleft 2x2: rotation&scale, third column: tralnslation |
| 22:09.10 | seebs | I was at the time working with Java, and the Java docs are really good at saying "a matrix with the following 6 values", and really bad at saying what those values mean or how you generate them. |
| 22:09.40 | Noshei | I dont even remember matrix math, that was like 10 years ago for me (when I learned it in school) |
| 22:12.55 | seebs | I learned matrix math in the general case. I didn't learn graphics terminology or applications. |
| 22:13.12 | Noshei | yeah same |
| 22:13.52 | Imhothar | hmm adding a z kooridnate with 4x4 math wouldn't really make things worde, would it? |
| 22:14.02 | seebs | Well, it probably wouldn't *do* anything. |
| 22:20.16 | Imhothar | you could do 3d UI?! |
| 22:21.13 | seebs | Well, you'd have to have other rendering code aware of it too. |
| 22:30.45 | Imhothar | not if it's just a matrix transformation |
| 22:31.01 | Imhothar | in 2d space, inside a closed canvas |
| 22:39.32 | Imhothar | the math is easy, but it would have tpo work without a depth test |
| 22:45.41 | Imhothar | wtf since why do I get kicked out due to inactivity today? |
| 22:46.34 | Imhothar | this is strange. I sometimes spend an hour or more in Notepad and don't get disconnected |
| 22:47.55 | Baanano | now it's 15 mins |
| 22:48.23 | Imhothar | it certainly wasn't 15 minutes yesterday |
| 22:48.44 | seebs | What's interesting to me is, the time from the AFK notification to the disconnect is WAY shorter than it used to be. |
| 22:49.10 | seebs | I note, BTW, that it used to be at one point that you would get disconnected if, and only if, the server was basically full. Otherwise it just didn't care. |
| 22:49.58 | Baanano | yeah, i spent hours in notepad without getting disconnected, but shortly after the patch they changed the afk timer to 15 mins |
| 22:50.05 | Noshei | they turned it back on with 2.3 or shortly after to help with server congestion |
| 22:50.54 | Imhothar | that's just stupid |
| 22:51.06 | Imhothar | how many resources does an idle player consume? |
| 22:51.27 | Noshei | a lot |
| 22:51.43 | seebs | Not-zero. Think of all the chat channels a user might be on. And they load all sorts of stuff into RAM when a player is logged in. |
| 22:51.44 | Imhothar | proof |
| 22:51.51 | Imhothar | sure it's nonzeot |
| 22:51.55 | Imhothar | nonzero |
| 22:52.02 | seebs | Also, what happens if they let a ton of players sit around idle, way more than they could support if they were active, and then the players become active? |
| 22:52.52 | Imhothar | but how much compared to player who is acticely doing things |
| 22:53.10 | seebs | Probably not THAT much less, especially if they're, say, inactive in a busy city. |
| 22:53.26 | seebs | I would guess that the notifications to your client of 50 other players moving are a much bigger issue than your movements. |
| 22:53.28 | Noshei | I dont think there will be much of a difference on the server end between an idle player and an active player |
| 22:54.32 | Imhothar | thinking about it... it makes sense |
| 22:54.49 | Imhothar | it's still stupid |
| 22:55.09 | Noshei | yeha I know |
| 22:55.29 | Imhothar | when do we get a "addon dev" flag on our accounts? |
| 22:56.02 | Imhothar | if the game would at least throw one back to character selection, but noooooo it has to quit the entire game |
| 22:56.05 | Imhothar | WTF? |
| 22:56.53 | Baanano | i still do my development on icewatch (trial shard now), so i don't think i'm consuming that much resources to get kicked twice an hour :P |
| 22:59.41 | seebs | Wait, what? |
| 22:59.43 | seebs | That's… odd. |
| 22:59.52 | seebs | I don't get dropped to character select from this, I don't think. |
| 23:00.00 | Imhothar | no you don't |
| 23:00.03 | Imhothar | it quites the game |
| 23:00.07 | seebs | Unless what's really happening is that, shortly after I get the "you are now AFK", I get disconnected for other reasons. |
| 23:00.13 | seebs | Er. I meant I don't get kicked out entirely. |
| 23:00.31 | seebs | What I've been seeing happening has given me a "disconnected from server" message and then left me back at character select. |
| 23:00.37 | Imhothar | I just get a dialog telling me I'm being kicked from inactivity. Hit OK, game quits |
| 23:06.23 | Noshei | could create a log parser that loogs for the chat text telling you will be kicked soon and have it notify you oudside of the game |
| 23:16.03 | seebs | Requires a working way to ensure that the log is enabled. |
| 23:16.37 | Imhothar | it's just stupid |