| 00:35.15 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 00:35.15 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 01:45.15 | Ivnedar | lol, amused that people are calling dream orbs "rng runes" now |
| 01:46.39 | Ivnedar | so if you're lucky/persistent enough to end up with bis runes on everything, ~15% of your core stats will come from the orbs |
| 01:46.46 | Ivnedar | that's a pretty big stat boost |
| 03:51.31 | seebs | Heh, one of our trolls (Madball) has hopped to Laethys, and is now organizing groups of people from Laethys to join Faeblight 1-29 and spam chat with religious trolling. |
| 03:51.47 | seebs | That's a little better effort than he's been doing recently, but he's still sorta phoning it in. |
| 03:52.59 | Ivnedar | Advocating for the Vigil? |
| 03:54.36 | Ivnedar | organising a group of trolls to act as Vigil zealots could be an amusing use of time |
| 04:00.03 | seebs | No, "arguing" about Christianity. Badly. |
| 04:00.29 | seebs | Basically, pretending to be a really obnoxious Christian, and having someone else be a hostile and ill-educated atheist, lets you provoke all sorts of stuff. |
| 04:06.34 | Noshei | this is odd, rift just stopped making any sound |
| 04:08.02 | Ivnedar | did you hit alt-s or whatever the global sound toggle key is? |
| 04:11.08 | Noshei | I did not, I dont have a key bind for it in fact |
| 04:13.41 | Noshei | restart fixed it |
| 04:16.55 | seebs | I have had it do that once, I think. |
| 04:50.49 | Ivnedar | lol, guildie is trying to work out what to tell his account this year at tax time |
| 04:50.58 | Ivnedar | when the account asks what all those TRION payments were |
| 04:55.03 | seebs | "hobby" |
| 04:55.53 | Ivnedar | he thinks his accountant will class it as gambling |
| 04:55.56 | Ivnedar | which is probably fair |
| 04:58.15 | seebs | I don't think anything I spend credits on is gambling. |
| 04:58.52 | Ivnedar | oh, for him it's 99% lockboxes |
| 05:17.13 | *** join/#riftuidev Imhothar_ (~Imhothar@ppp-93-104-190-26.dynamic.mnet-online.de) |
| 05:28.48 | *** join/#riftuidev Imhothar__ (~Imhothar@ppp-188-174-134-155.dynamic.mnet-online.de) |
| 05:43.10 | Amadeus- | Ivnedar, seebs, or ZorbaBeta ...can one of you give me a simple example of how you would avoid "performance warnings" on an event callback function? I mean, I understand the general principal ...but, I don't want to make a mistake. |
| 05:44.29 | *** join/#riftuidev Imhothar___ (~Imhothar@ppp-188-174-22-144.dynamic.mnet-online.de) |
| 05:44.47 | Amadeus- | in other words, some users are getting performance warnings for, as far as I can tell, the event callback function being called too many times at once in a frame.... The function itself (and the functions that are called) don't seem to be too intensive on their own...and I don't see any loops. |
| 05:45.14 | Ivnedar | do you have any control over the event callback function? |
| 05:45.44 | Ivnedar | and if they're called too many times, what did you want to happen instead? they get skipped or put off to later frames? |
| 05:47.59 | seebs | What I mostly do is if an event callback can happen a lot, often some of them are inherently redundant. So I have the event callback just record the data in some trivial way, and have a once-per-frame update that, if there's anything recorded from events, processes that. |
| 05:48.09 | *** join/#riftuidev Imhothar____ (~Imhothar@ppp-188-174-121-27.dynamic.mnet-online.de) |
| 05:48.44 | Ivnedar | if you're trying to smooth out bursts, you could have an event queue that the callback pushes something into and have an update event that pulls items from the queue until a watchdog threshold is met |
| 05:49.05 | Ivnedar | doesn't work so well if you're always getting more events than you can process though |
| 05:54.53 | *** join/#riftuidev Imhothar_____ (~Imhothar@ppp-93-104-180-152.dynamic.mnet-online.de) |
| 06:00.17 | *** join/#riftuidev Imhothar (~Imhothar@ppp-93-104-172-69.dynamic.mnet-online.de) |
| 06:05.26 | *** join/#riftuidev Imhothar______ (~Imhothar@ppp-188-174-25-237.dynamic.mnet-online.de) |
| 06:22.36 | Amadeus- | naw, I don't think that would happen ..it'd be nice to see a really simple example of a once-per-frame-update queue type thing like you're both describing |
| 06:23.45 | Amadeus- | I mean, I understand the logic ..just would like to see it once in LUA and/or the Rift Addon API sincE I'm not overly familiar with either just yet |
| 06:24.32 | Imhothar______ | Lua not LUA. And yes, I *will* keep on insisting on it ;) |
| 06:48.33 | Amadeus- | I suppose saying LUA is better than saying "that piece of shit language" |
| 06:50.20 | Amadeus- | but, regardless of what anyone tells you, in English a word that is all capitalized does not necessarily mean that it's an acronym or anything special. It could denote something of great importance when the tools used for writing do not allow for underline or bold, etc. |
| 06:50.44 | Amadeus- | it's not uncommon for nouns to be fully capitalized in documents that are written in unformatted text (like IRC) :) |
| 06:51.44 | Amadeus- | heck, there's even a font option in MS Word to make whole lines of text "small caps" |
| 07:36.43 | seebs | Yes, but name case is a part of the name. |
| 07:36.55 | seebs | Anyway, it's not a piece-of-shit language, it's a language that has a different idiom than you're used to. |
| 07:37.12 | seebs | Okay, simplest form of the queue thing. Assume you have a table foo that you stash your addon stuff in. |
| 07:37.15 | seebs | foo.updates = {} |
| 07:37.38 | seebs | function foo.event_hook(event, data) foo.updates[#foo.updates + 1] = data end |
| 07:38.05 | seebs | function foo.frame_hook(event) if #foo.updates > 0 then do_stuff_with(foo.updates); foo.updates = {} end end |
| 07:38.33 | seebs | Hook the event (say, Event.Inventory.Slot or whatever) to event_hook, and Event.System.Update.Begin to frame_hook. |
| 08:28.57 | Amadeus- | seebs, thanks :) |
| 08:30.08 | Amadeus- | so basically you can do pretty much anything you want on the Event.System.Update.Begin and not worry about watchdog? ...also, why "Begin" and not "End" (or does End not exist?) |
| 08:30.23 | Amadeus- | guess it really doesn't matter since frames happen pretty fast |
| 08:31.48 | Amadeus- | seebs, also ..what's the best practice ...to have a table (foo.updates in your example) for each "queue" ...or, have a global table contain a table (for each entry) that also indicates the function to be called, the data, etc.? |
| 09:15.17 | Amadeus- | seebs, if you get a chance can you look at this? For some reason my changes don't work ...tried to make it as simple as possible just to test it. No errors and it looks like it SHOULD work exactly the same ....but, it doesn't. |
| 09:15.19 | Amadeus- | https://dl.dropboxusercontent.com/u/19522021/DisplayChanges.html |
| 09:17.38 | Amadeus- | I'm not fancy enough to figure out how to make the queue store the function with the parameters all at once :) |
| 09:28.59 | Thorarin | Hmmz, that -1/12 proof is just weird :P |
| 09:29.43 | Thorarin | If you accept that S1 = 1 - 1 + 1 - 1 ... = 1/2 then why isn't S2 = 1 - 2 + 3 - 4 ... = 0 by that same reasoning |
| 09:30.38 | Thorarin | I can't think of a good reason :P |
| 09:53.23 | Ivnedar | where's the -1/12 proof? |
| 10:02.54 | Ivnedar | http://www.robg3d.com/?p=1175 |
| 10:03.09 | Ivnedar | response from a ccp dev about why they're still using python 2.6 and not 3.x |
| 10:03.37 | Ivnedar | I can't stand python 3, hate that that's the direction the python devs are going |
| 10:04.38 | Amadeus- | Yea, I saw that earlier today |
| 10:05.15 | Amadeus- | CCP really uses their own version of python anyway for EVE ...they have a wild version of stackless they use ...so, his opinion is a little skewed |
| 10:05.42 | Ivnedar | arguments are still pretty valid |
| 10:06.03 | Ivnedar | and probably are the same for a lot of the "why don't they just use/upgrade X" |
| 10:11.12 | Ivnedar | didn't realise eve was python on windows |
| 10:49.55 | Amadeus- | it has a C++ wrapper ....and the game engine is C++ ...but, the guts of the client is python |
| 10:49.59 | Amadeus- | stackless python |
| 10:50.03 | Amadeus- | on the server adn the client |
| 10:59.36 | Thorarin | https://www.youtube.com/watch?v=w-I6XTVZXww |
| 10:59.44 | Thorarin | ^-- Ivnedar |
| 11:00.29 | Ivnedar | ah, thanks |
| 11:05.19 | Ivnedar | when you're working with infinite sets, you can pretty much end up with any result surely |
| 11:09.49 | Ivnedar | It really lowers the belief in anything else quantum physics comes up with if it's all based on things like that :P |
| 11:12.58 | Ivnedar | I can see how S2 would be 1/4 |
| 12:20.25 | Ivnedar | wow, rift crashed that time with no crash handler popup |
| 13:10.09 | *** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net) |
| 13:32.43 | Thorarin | I can see how it would be 1/4 if the other one is 1/2 |
| 13:32.47 | Thorarin | But it could also be 0 |
| 13:33.06 | Thorarin | Following the same reasoning, which leads me to believe the whole reasoning is flawed tbh :P |
| 13:33.21 | Noshei | you talking about 1+2+3+4+5+6+7+... = -1/12 Thorarin? |
| 13:33.26 | Thorarin | Yes |
| 13:33.45 | Thorarin | <Thorarin> If you accept that S1 = 1 - 1 + 1 - 1 ... = 1/2 then why isn't S2 = 1 - 2 + 3 - 4 ... = 0 by that same reasoning |
| 13:34.34 | Noshei | the part that you have to keep in mind, is that that is the sum of the infinite series, if you take the sum at any given point it will be either 0 or 1 |
| 13:34.49 | Thorarin | I know... |
| 13:34.57 | Thorarin | But that doesn't explain what I just said |
| 13:35.12 | Noshei | that one is explained in further detail in another one of their videos and is proven a couple of different ways |
| 13:35.19 | Thorarin | I think the 1/2 is is dodgy, but I suck at maths so I accept it as fact |
| 13:35.41 | Thorarin | I follow the same reasoning on S2, which then says the outcome is 0 |
| 13:35.57 | Thorarin | Being the average of 1 and -1 |
| 13:36.18 | Thorarin | How can it be 0 and 1/4 at the same time? :P |
| 13:37.01 | ZorbaBeta | I don't think you can follow the same reasoning on S2 - it isn't the average of 1 and -1. the series as we evaluate it is 1, -1, 2, -2, 3, -3, 4, -4, 5, -5 |
| 13:37.16 | Thorarin | And Euler Riemann thing includes axiomas I'm not familiar with and applies more dodgy reasoning as well :P |
| 13:37.20 | ZorbaBeta | the "it's the average of the two terms" logic works when our series is alternating between two terms, but in the case of S2, it isn't alternating between two terms |
| 13:38.31 | Thorarin | Hmmz Zorba, still we're talking about a discrete set, the result of any number minus the next one is -1 |
| 13:38.51 | Thorarin | So I don't see how it would be different |
| 13:38.59 | Noshei | but that is when you think of it as a finite series |
| 13:39.02 | ZorbaBeta | I guess my point is that I don't see an obvious reason that the logic given should extend to a nonperiodic series, which that is |
| 13:39.10 | ZorbaBeta | if it was periodic I'd agree with you, note |
| 13:39.20 | Noshei | when you deal with infinite series, things get really strange |
| 13:40.18 | Thorarin | Infinitity is strange yes, but the way they explain it in the movie doesn't suggest any reason why I couldn't apply that same logic to S2 |
| 13:40.37 | Thorarin | So in that sense I find the explanation lacking |
| 13:41.15 | Noshei | they explain 1-1+1-1+1-1+1-1... = 1/2 thing in further detail in another video and provide multiple ways to get that result |
| 13:41.27 | Noshei | so they aren't simply relying on that one method |
| 13:41.41 | Thorarin | That isn't the issue tho |
| 13:41.52 | Thorarin | I already stated that I accept it as fact |
| 13:42.35 | Thorarin | But I don't see why it applies only to x-x+x-x... |
| 13:42.51 | Noshei | https://www.youtube.com/watch?v=PCu_BNNI5x4 - watch that |
| 13:43.24 | Thorarin | Unfortunately I need to do some coding as well today :< |
| 13:43.30 | ZorbaBeta | thorarin, I guess my argument would be that if you average a specific two consecutive items, you get 0 |
| 13:43.31 | Thorarin | I blame highlights :P |
| 13:43.36 | ZorbaBeta | but if you average a different two consecutive items you get 1/2 |
| 13:43.48 | ZorbaBeta | in fact, no matter how large a space you decide to average, you get a different average depending on where you start |
| 13:44.27 | ZorbaBeta | that's not a property that 1-1+1-1+1 has, as long as you average a multiple of the period length. S2 doesn't have a period length, though. To me, that's a pretty huge difference. :) |
| 13:44.49 | Thorarin | Where is the average 1/2? |
| 13:45.16 | ZorbaBeta | the series is in S1 ends up evaluating to 1, 0, 1, 0, 1, 0, 1, 0, ... |
| 13:45.37 | ZorbaBeta | any two (or four, or six, or (n*2) for positive integer n) adjacent numbers averaged come out to 1/2 |
| 13:45.48 | Thorarin | And the other one to 1, -1, 1, -1, 1, -1 ... |
| 13:46.05 | ZorbaBeta | no, it doesn't - you're evaluating it wrong |
| 13:46.13 | ZorbaBeta | it's 1, -1, 2, -2, 3, -3, 4, -4, ... |
| 13:46.44 | Noshei | you apply the operations to the first number |
| 13:46.56 | Thorarin | I guess I just need coffee :P |
| 13:53.13 | Ivnedar | I need to spend more time on string theory basics I think |
| 13:53.27 | Ivnedar | failing to grasp things like how a 1 dimensional object can have a radius |
| 13:53.33 | Thorarin | Even when the result keeps getting higher (positive or negative), for every x I can find -x in the infinite sequence as well |
| 13:53.43 | Thorarin | So I'd still say it cancels out, but meh :P |
| 13:53.49 | Ivnedar | where? |
| 13:54.04 | Thorarin | I'm just confusing myself more :> |
| 13:54.23 | Ivnedar | +1 -2 +3 -4 ... |
| 13:54.30 | Thorarin | 1 dimension with a radius 0.o |
| 13:54.50 | Noshei | ugg, Thorarin you want to do abom's kbm file? |
| 13:54.54 | Thorarin | In 1 dimension radius is equal to length I suppose? |
| 13:54.58 | Noshei | it is going to be a HUGE pain in tha ass |
| 13:55.11 | Ivnedar | lol |
| 13:55.27 | Ivnedar | 3 mini bosses all their discrete phases and joined phases plus head? |
| 13:55.36 | Noshei | yeah |
| 13:55.45 | Ivnedar | addiction kill video looked pretty straight forward |
| 13:57.04 | Noshei | haven't watched the video yet, just reading the strat atm |
| 13:59.54 | Ivnedar | but yeah, the S2 I can accept, you have effectively two lines that it alternates between that are diverging, adding them together almost cancels each other out, but the line below the axis is a step behind the one above, so the average is slightly above the x axis, so I can accept that 1/4 makes sense |
| 14:00.01 | Ivnedar | attempting to visualise it anyway |
| 14:01.23 | Ivnedar | (plotting out the running total that is, not the values) |
| 14:03.03 | Ivnedar | and if I can accept S1 and S2, then I can see how to get to S3, even if visualising it completely breaks down |
| 14:06.43 | ZorbaBeta | I've heard that when ramanujan came up with the wacky -1/12 result, it was just offhandedly mentioned in his notes without any explanation or context |
| 14:07.46 | Thorarin | Wannabe Fermat tbh :P |
| 14:08.02 | ZorbaBeta | honestly he was in a league above fermat. the guy was inhumanly skilled |
| 14:08.35 | Thorarin | But it took longer to prove Fermat's theorem :) |
| 14:09.05 | ZorbaBeta | true. I think ramanujan explained how he'd come up with that later ;) |
| 14:09.25 | ZorbaBeta | that said, note that it's generally accepted that Fermat had not actually proven his theorem (though the jury's out on whether he thought he had) |
| 14:09.43 | Thorarin | There's a book about it right? |
| 14:10.05 | Thorarin | Some guy that thought he proved it in god knows how many pages, but there was an error somewhere |
| 14:11.01 | Thorarin | While Fermat's comment sounds like he could ALMOST fit it on the page ;) |
| 14:12.56 | ZorbaBeta | yeah, though he eventually managed to fix the error |
| 14:13.06 | Ivnedar | wiles |
| 14:13.18 | ZorbaBeta | and really that's the reason people don't think fermat proved it - the eventual proof was *vastly* larger than could fit on a page, and relied on entire schools of mathematics that had not been invented yet |
| 14:13.51 | ZorbaBeta | either Fermat came up with a clever trick that we've never been able to reproduce in years, invented a vast swath of mathematics and then didn't bother to write it down, was trolling, or was wrong. chances are good it was the last one. ;) |
| 14:14.12 | Ivnedar | second last one is plausible too :P |
| 14:14.17 | ZorbaBeta | plausible, yes :) |
| 14:14.46 | Noshei | well good thing he didn't post on the trion forums, he would have gotten banned for trolling |
| 14:15.21 | Ivnedar | ooh, he posted his proof on the forums and the post was removed as trolling since it predated the visible infraction system! |
| 14:15.46 | Ivnedar | the moderator didn't have a thorough enough understanding of mathematics |
| 14:15.56 | Ivnedar | I believe this has since been rectified |
| 14:18.02 | Noshei | Thorarin: didn't you say you had an addon you made to track when a boss casts something, how often etc |
| 14:19.34 | Ivnedar | http://pastebin.com/N6B5XKaR |
| 14:19.37 | Ivnedar | that's the one I use mostly |
| 14:19.57 | Ivnedar | It assumes the boss you want to track is in the TrackedBosses list and focused |
| 14:20.36 | Ivnedar | it'll be rather noisy at first startup as it records and announces every buff and ability cast it sees as it fills its cache :P |
| 14:21.41 | Ivnedar | It could be fixed up to retrieve the specifier by a unit lookup instead of relying on the focus target I guess if you want to track multiple bosses at once |
| 14:21.53 | Ivnedar | though the last cast timers would need to be unitid indexed then |
| 14:23.55 | Noshei | can you post the toc as well |
| 14:24.45 | Ivnedar | http://pastebin.com/U7hR5c7R |
| 14:25.54 | Ivnedar | other than adding a new boss every now and then, I haven't really touched it since regulos |
| 14:27.31 | Noshei | hows it output? |
| 14:27.55 | Noshei | other than the chat output, just the savedvariable file? |
| 14:28.01 | Ivnedar | yeah |
| 14:28.10 | Ivnedar | I look in the saved var for the buffs later on |
| 14:28.21 | Ivnedar | and dump the cast stuff to /log |
| 14:28.29 | Ivnedar | 09:58:43: [Buff Log] Breaker X-1 "Onyx" (u8000000373039DB1) casts: Fission Burst (nil) -> Ivnedar HP: 73.22 Last: 10, From Start: 175 |
| 14:28.35 | Ivnedar | 09:58:48: [Buff Log] Breaker X-1 "Onyx" (u8000000373039DB1) casts: Quantum Disruptor (nil) -> Finella HP: 72.25 Last: 10, From Start: 179 |
| 14:28.40 | Ivnedar | 09:58:53: [Buff Log] Breaker X-1 "Onyx" (u8000000373039DB1) casts: Fission Burst (nil) -> Markai HP: 71.23 Last: 10, From Start: 185 |
| 14:28.43 | Ivnedar | etc |
| 14:29.24 | Ivnedar | so with enough attempts, I get a pretty good idea ability cooldowns, whether things are %hp based or time from start based, etc |
| 14:29.39 | Ivnedar | could probably record say and notify events I guess |
| 14:29.50 | Ivnedar | I wonder how much more would be needed to record a whole encounter |
| 14:32.24 | Ivnedar | the (nil) is the abilitynew id if it exists |
| 14:32.30 | Ivnedar | most raid mechanics don't have them, random ones do |
| 14:32.45 | Ivnedar | I was hoping to use it so I didn't need to translate every ability |
| 14:41.35 | Thorarin | Hmm, so Fermat invented trolling AND a bunch of mathematics :) |
| 15:07.18 | Thorarin | So much fun to integrate two separately developed pieces of code :< |
| 15:08.05 | Thorarin | Turns out the other developer didn't bother to test anything, so that was bug number 5 that I fixed on his side when assuming it was on my side :P |
| 17:52.22 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 17:52.23 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 17:55.07 | seebs | I think the reason I'm confused by the -1/12 thing is that it's a sum of a series of all positive numbers. |
| 17:55.29 | seebs | Oh, and Amadeus-, it's not that update.begin ignores the watchdog. It's that accumulating things does less work. |
| 17:56.09 | seebs | So, as an example, my bag addon used to do savedvar[my_name] = Inspect.Item.Detail(Inspect.Item.List(Utility.Item.Slot.All())) on every bag event. |
| 17:56.30 | seebs | So if there were multiple inventory/slot/item events in a frame, I would generate that very large table on every event. |
| 17:56.53 | seebs | When I switched to a queue, and just do it once per frame at most, the cost was noticably reduced. |
| 19:07.31 | *** join/#riftuidev Imhothar (~Imhothar@ppp-62-216-194-156.dynamic.mnet-online.de) |
| 22:02.45 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 22:02.46 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 22:03.55 | Noshei | lol ZorbaBeta looks like simon is trying to make your job even harder |
| 22:04.16 | Noshei | that hialol laser thing for dimensions, wow |
| 23:44.18 | Ivnedar | lol, love daglar's response to rift dev team size thread |