| 00:02.20 | *** join/#riftuidev Snstr (~darki@p54B36497.dip.t-dialin.net) |
| 00:13.18 | *** join/#riftuidev Aurorus (~Aurorus@unaffiliated/aurorus) |
| 01:32.58 | *** join/#riftuidev Kraps (~chatzilla@ool-4572cb11.dyn.optonline.net) |
| 02:04.21 | lXl | http://news.softpedia.com/news/EVGA-SR-X-Dual-Socket-LGA-2011-Motherboard-Picture-Preview-239810.shtml WANT THIS |
| 02:11.10 | *** join/#riftuidev Aurorus (~Aurorus@unaffiliated/aurorus) |
| 04:12.09 | *** join/#riftuidev dolby (~dolby-wow@MMOI/Administrator/pdpc.silver.dolby) |
| 04:12.09 | *** mode/#riftuidev [+o dolby] by ChanServ |
| 05:59.45 | *** join/#riftuidev RagebeastRT (~chatzilla@68.170.244.180) |
| 06:00.25 | RagebeastRT | Anyone know where I can find a buff detail identifier list? nothing on the wiki or the api docs |
| 07:02.22 | *** join/#riftuidev Sunspots (~Sunspots@c213-89-195-33.bredband.comhem.se) |
| 07:19.22 | *** join/#riftuidev Kreiri (~kreiri@109.122.11.176) |
| 08:34.17 | *** join/#riftuidev Dramiel (~Adium@134.140.broadband5.iol.cz) |
| 12:11.43 | Kreiri | Zorba: is bug with buff durations becoming equal to their remaining time on teleport an API bug or buff system bug? |
| 13:01.37 | *** join/#riftuidev TiuTalk (~Thiago_Be@189.60.225.176) |
| 13:01.45 | TiuTalk | hiho, anyone there? |
| 13:02.11 | Dramiel | nope |
| 13:02.18 | TiuTalk | :P |
| 13:03.03 | TiuTalk | i'm working on an addon and i need it to make the player cast a spell, in game I have "cast <spell name>" but this doesn't work inside the addon code |
| 13:03.23 | TiuTalk | any tip? |
| 13:04.01 | Dramiel | you can cast spells only with secure frames or something like that |
| 13:04.10 | Dramiel | and it will be only to response to click or something |
| 13:04.59 | TiuTalk | yeah, i have the click event |
| 13:05.07 | TiuTalk | i just don't know what to put in there |
| 13:20.37 | Dramiel | frame:SetSecureMode("restricted"); frame.Event.LeftClick = "cast something"; |
| 13:20.44 | Dramiel | or something like that |
| 13:27.11 | *** join/#riftuidev Dramiel1 (~Adium@88.100.140.134) |
| 13:30.00 | *** join/#riftuidev Dramiel (~Adium@134.140.broadband5.iol.cz) |
| 13:35.53 | *** join/#riftuidev Sunspots (~Sunspots@c213-89-195-33.bredband.comhem.se) |
| 14:08.42 | *** join/#riftuidev sylvanaar_work (~sylvanaar@pdpc/supporter/active/sylvanaar) |
| 15:36.37 | *** join/#riftuidev Kraps_w (~chatzilla@static-64-61-105-166.isp.broadviewnet.net) |
| 16:27.39 | *** join/#riftuidev Kreiri (~Miranda@109.227.120.47) |
| 18:36.51 | *** join/#riftuidev TiuTalk (~Thiago_Be@189.60.225.176) |
| 18:46.38 | TiuTalk | Hiho |
| 18:47.23 | TiuTalk | to have a restricted frame my whole context must be restricted? |
| 19:18.58 | *** join/#riftuidev RagebeastRT (~chatzilla@68.170.244.180) |
| 19:19.55 | RagebeastRT | Anyone alive that could help me with a string function? split/join isnt in the api so would like a routine that emulated them |
| 19:30.07 | seebs | Depends on how fancy you need them to be. |
| 19:30.16 | seebs | I have a split buried in LibBaggotry, I think. |
| 19:32.10 | ShaunO`work | i use the explode function provided on the string splitting lua wiki |
| 19:32.27 | ShaunO`work | http://lua-users.org/wiki/SplitJoin |
| 19:33.57 | RagebeastRT | I want to be able to concat macro strings with a delimiter into a string and then split them back at a later time |
| 19:34.08 | RagebeastRT | I see that Shaun, doesnt work in this api |
| 19:35.37 | RagebeastRT | ah, didnt scroll far enough |
| 19:36.47 | RagebeastRT | what would be best to use on that page for "cast group## Ability; cast [ctrl] group## ability;target group##; cast [alt] focus ability" |
| 19:37.14 | RagebeastRT | not that would logically work, just an example of a string |
| 19:39.58 | ShaunO`work | just implement whatever you understand the best and tweak it if it's not desirable |
| 19:40.03 | ShaunO`work | *change it |
| 20:02.11 | RagebeastRT | not having any luck. I can get it to return an index so I know its being tabled, I get 3 indexs with 3 different macros but not getting the macro itself |
| 20:02.25 | RagebeastRT | using the Function: Split a string with a pattern, Take Two example |
| 20:04.19 | ShaunO`work | i used the php style explode function and it worked fine, only thing that caught me out was the 1-based index, i only just started using lua in the weekend so can't go into detail |
| 20:13.30 | RagebeastRT | I tried it, same rsults |
| 20:13.33 | RagebeastRT | t = explode(",", "cast group## Ability, cast [ctrl] group## ability,casst [alt] focus ability") |
| 20:13.50 | RagebeastRT | <PROTECTED> |
| 20:14.07 | RagebeastRT | I get results of 1, 2, 3 |
| 20:16.10 | ShaunO`work | it |
| 20:16.38 | ShaunO`work | is that correct syntax? i dont think it returns pairs of values, like i said though i know nothing about lua |
| 20:16.44 | ShaunO`work | but i wouldnt imagine you could call pairs() on it |
| 20:16.56 | ShaunO`work | there wont be a key |
| 20:17.02 | ShaunO`work | just an indexed array of values? |
| 20:17.26 | RagebeastRT | it returns a table, I am just printing out the contents of a table |
| 20:17.51 | ShaunO`work | what if you just print(t[1]) or something do you get the value |
| 20:18.23 | ShaunO`work | i think you need to use ipairs() in this context |
| 20:19.03 | ShaunO`work | i think pairs() is only valid in an associative array but its just a guess |
| 20:21.03 | RagebeastRT | if I try ipairs, it only returns a 1, no text |
| 20:21.36 | RagebeastRT | and Seebs, tried your split, it only returned the first string, no consecutive strings |
| 20:23.10 | RagebeastRT | had a wrong value Shaun. It returned all 3 but no strings, just $s |
| 20:23.11 | seebs | hmm |
| 20:23.14 | RagebeastRT | #s |
| 20:23.21 | seebs | That's interesting, 'cuz it obviouslt works for me. |
| 20:23.57 | RagebeastRT | Seebs, when I tried to access the result with pairs, it said expecting table, got string |
| 20:24.02 | seebs | hmm |
| 20:24.14 | RagebeastRT | so I just printed out t and it printed the first string and then stopped |
| 20:25.24 | seebs | ohh |
| 20:25.39 | seebs | ary = [ lbag.strsplit(s, p) ] |
| 20:25.47 | seebs | It returns multiple values, not an array. |
| 20:26.01 | seebs | In lua, functions can have multiple returns: |
| 20:26.06 | seebs | x, y = twovalues() |
| 20:26.09 | seebs | and if you just do: |
| 20:26.11 | seebs | x = twovalues() |
| 20:26.22 | seebs | it just gets the first value. |
| 20:26.23 | ShaunO`work | D: |
| 20:26.52 | ShaunO`work | madness |
| 20:27.20 | RagebeastRT | im lost |
| 20:27.53 | RagebeastRT | so I guess t1, t2, t3 = bag.strsplit(s,p) woudldnt work? |
| 20:27.59 | seebs | That would! |
| 20:28.03 | ShaunO`work | surely that syntax wouldn't be used in this case seebs because the amount of items split from the string is non-deterministic |
| 20:28.08 | seebs | That would get you the first three values. |
| 20:28.15 | RagebeastRT | doesnt work :P |
| 20:28.18 | ShaunO`work | there must be a way to iterate over the returned collection |
| 20:28.35 | seebs | There may be, but I just do ary = [ thing_returning_multiple_values ] when I want that. |
| 20:28.36 | RagebeastRT | attempt to index global 't' |
| 20:29.01 | seebs | That sounds like you wrote t[1] or something. |
| 20:29.07 | RagebeastRT | no |
| 20:29.09 | RagebeastRT | sec |
| 20:29.11 | ShaunO`work | in my own code i used, local SplitValues = explode(" ", "test words"), print(SplitValues[1]) -- = "test", print(SplitValues[2]) = "words" and it worked first time |
| 20:29.52 | ShaunO`work | i would assume i could iterate over SplitValues with a counter and just do SplitValues[i] to the length of the array |
| 20:30.16 | RagebeastRT | I dont want words, want the who sections of string |
| 20:30.16 | RagebeastRT | seebs I had . instead o f comma |
| 20:30.16 | ShaunO`work | yeah was just en example |
| 20:30.21 | RagebeastRT | anyhow, fixed that and just printed 1st one, not other 2 |
| 20:31.14 | RagebeastRT | and shaun, your example is the same as for a,b in pairs(result) do print (b) end |
| 20:31.35 | RagebeastRT | I was print(a,b) just to see how many reutrned |
| 20:32.52 | RagebeastRT | try using your example but make it "test words1, test words 2" and see what results are |
| 20:33.00 | RagebeastRT | and use , as delim |
| 20:33.13 | ShaunO`work | im at work, no lua to test with |
| 20:33.48 | ShaunO`work | oh wait an online repl exists :D |
| 20:35.17 | ShaunO`work | works fine |
| 20:35.23 | RagebeastRT | grrr |
| 20:35.24 | ShaunO`work | http://repl.it/BsC# |
| 20:36.17 | ShaunO`work | hit the play button up at the top in the middle in between the two and see the output |
| 20:36.37 | ShaunO`work | err |
| 20:36.46 | ShaunO`work | in the loop it should just be print(Value) |
| 20:36.55 | ShaunO`work | still works fine |
| 20:38.53 | ShaunO`work | not sure why you're having trouble, put your code in the repl and if it doesnt work, paste the link in here to have a look at http://repl.it/BsC/1# |
| 20:39.47 | RagebeastRT | woot it works! |
| 20:40.12 | RagebeastRT | no idea why print (a, b) didnt but t[a] did |
| 20:40.36 | ShaunO`work | what is the syntax of the print command? |
| 20:40.40 | ShaunO`work | what are the extra parameters? |
| 20:40.50 | ShaunO`work | does it just concatenate the string |
| 20:40.53 | ShaunO`work | err the parameters |
| 20:41.08 | RagebeastRT | for a, b in ipairs(t) do print (a, b) end |
| 20:41.11 | ShaunO`work | i need coffee badly |
| 20:41.15 | RagebeastRT | same :P |
| 20:41.19 | ShaunO`work | what does print(a, b) do |
| 20:41.30 | RagebeastRT | just prints the index, value |
| 20:42.03 | RagebeastRT | same as your example, you use index, value, I used a, b same thing |
| 20:42.24 | ShaunO`work | sorry i was more meaning the actual print command itself |
| 20:42.32 | ShaunO`work | what does it do when it receives multiple arguments |
| 20:42.37 | ShaunO`work | is it just the same as print(a .. b) |
| 20:42.46 | RagebeastRT | and.... |
| 20:42.59 | RagebeastRT | for a, b in pairs(t) do returns same as a,b in ipairs(t) |
| 20:42.59 | seebs | I think it puts spaces in between them or something. |
| 20:43.12 | ShaunO`work | ah yep |
| 20:43.12 | RagebeastRT | think the comma is a tab in the print function |
| 20:43.14 | seebs | That'll be true for any table with all numeric keys. |
| 20:43.16 | ShaunO`work | i see i see |
| 20:44.00 | RagebeastRT | soooo now I can get back to work again after this fine little tutorial :P |
| 20:44.09 | RagebeastRT | well, after a smoke |
| 20:44.11 | seebs | ipairs() shows integer keys only, pairs() shows all keys. |
| 20:44.35 | RagebeastRT | I use just pairs throught all my code, seemed to work fine |
| 20:45.04 | RagebeastRT | in inspect.unit.lookup, abilities, details...everything |
| 20:45.39 | RagebeastRT | hell, when I first started it was for = 1, 20 :P |
| 20:46.02 | RagebeastRT | old basic, for i = 1 to num :P |
| 20:48.56 | seebs | That's right for most of that stuff, because most of those are tables that aren't array-like. |
| 20:49.17 | seebs | Here's the thing. In many languages, there's a distinction between arrays (numeric keys in order) and tables (or hashes or whatever), where keys can be anything. |
| 20:49.33 | seebs | In Lua, they're all just "tables"; the array functionality is just special cases of the table functionality. |
| 20:50.22 | RagebeastRT | last time I coded was visualbasic 5 |
| 20:50.42 | seebs | I have long since lost count of the number of languages in which I can program well enough to get stuff done. |
| 20:50.51 | RagebeastRT | since rift is just starting with the addon and werent any available for what I wanted, decided to try it out |
| 20:50.52 | seebs | I have the problem where I can't remember which one I'm in sometimes. |
| 20:50.58 | RagebeastRT | I bet |
| 20:51.06 | seebs | The rift addon API is one of the best I've seen. |
| 20:51.12 | ZorbaTHut | kreiri, I haven't looked into it yet in depth (it's on my todo list) but I'm guessing it's behaving-as-intended, in the sense that it's not going to get fixed. It's probably just an artifact of how we deal with teleports. |
| 20:51.18 | RagebeastRT | zorba! |
| 20:51.23 | seebs | WoW addons, I felt like I was constantly fighting to overcome the API. |
| 20:51.43 | seebs | Rift addons, the only problem is that the API lead is a narcissistic idiot with no sense of humor, who reacts badly to being teased. |
| 20:51.46 | RagebeastRT | please add unit.detail.los .aggro .range pretty pretty pls :P |
| 20:52.05 | RagebeastRT | needed badly for Rift Healer |
| 20:52.43 | Kreiri | it causes existing buffs which were filtered with duration filters to not match them after teleport |
| 20:52.58 | ZorbaTHut | rage, I actually can't think of a reason not to add those, so they might come as part of 1.7 (no promises - all depends on my available time) |
| 20:53.19 | ZorbaTHut | .range was on hiatus for a while as I ran some stuff past Design to make sure it was okay, but I got approved, so |
| 20:53.26 | ZorbaTHut | seebs, :P |
| 20:53.45 | ZorbaTHut | kreiri, yeah, but, honestly, if it got "fixed", it would be via me doing a bunch of code to match the new buffs up with the old buffs |
| 20:53.53 | ZorbaTHut | and at that point, I may say "just write that code yourself" |
| 20:54.02 | ZorbaTHut | as I said, I'm gonna look into it, I just suspect it'll end up being as-designed. |
| 20:54.05 | Kreiri | 'kay |
| 20:54.26 | RagebeastRT | that will make all the difference in the addon. Right now I have to have defaults frames open just to see if there is a los/range issue and to see who has main aggro |
| 20:54.31 | ZorbaTHut | I agree it's a bit annoying but oh man you have not seen my todo list :V it's like three pages long and I don't run my monitor at a low resolution |
| 20:55.42 | ShaunO`work | lol programming C# all day its like a luxury car compared to lua but lua is fine compared to perl |
| 20:55.57 | ZorbaTHut | rage, technically you could figure out "main aggro" by just finding the boss entity and seeing who it's targeting |
| 20:56.15 | ZorbaTHut | and, I mean, if nobody in your party is targeting the boss, you have bigger problems than who the boss is fighting |
| 20:56.47 | seebs | Lua is actually one of my favoritest programming languages. I like it a lot better than most of the C-derivatives. |
| 20:56.50 | RagebeastRT | using it in testing in warfronts though where I will get the highest overhead to test |
| 20:57.07 | ZorbaTHut | ah, yeah, I can see how that'd be a little sketchy |
| 20:57.09 | seebs | It's a small, clear, language that can do what I want. |
| 20:57.24 | RagebeastRT | could you explain using LibUnitChange over event.unit add/remove/avail? I see alot of cpu usage and need to cut it down |
| 20:57.32 | ZorbaTHut | seebs, I like it for the same reason. it also doesn't get in my way. |
| 20:57.38 | Kreiri | I guess when people want to see first five units in boss agrro table they want too much |
| 20:57.41 | ZorbaTHut | rage, I thinks someone posted a LibUnitChange example on the forums in the last few days |
| 20:57.51 | Kreiri | spoiled WoW babies etc |
| 20:57.54 | RagebeastRT | read it, didnt make sense to me when to actually call it |
| 20:57.58 | ShaunO`work | yeah i just got thrown off by a few things |
| 20:57.58 | seebs | Yeah. I basically never find myself fighting with the language. I may have to implement stuff, but I rarely find myself annoyed by it. |
| 20:58.00 | RagebeastRT | new to lua |
| 20:58.10 | ZorbaTHut | during setup - you call it when you want to start monitoring a unit |
| 20:58.11 | ShaunO`work | the worst so far being the lack of native a switch..case syntax |
| 20:58.12 | ZorbaTHut | just once |
| 20:58.18 | ZorbaTHut | and then it'll inform you when the unit changes |
| 20:58.27 | seebs | LibAccounts was pretty fun to write, and now LibBaggotry can give the *useful* answer "what stuff is had by characters I can get stuff from". |
| 20:58.34 | ZorbaTHut | seebs, fancy garbage collection is a bit tough, but that'll be improved with ephemerons |
| 20:58.35 | seebs | Except that I haven't added an "expose stuff across accounts" feature yet. |
| 20:58.46 | ZorbaTHut | shaun, to be honest, I never use switch/case even in C++ |
| 20:59.00 | ZorbaTHut | but in Lua you can accomplish it with a jump table |
| 20:59.04 | RagebeastRT | but I can only access the unit.details in that avent function right? |
| 20:59.17 | RagebeastRT | like I couldnt |
| 20:59.17 | seebs | I use it occasionally in C. But only occasionally. Mostly because gcc has a warning for "you're switching on an enum but didn't handle all the cases." |
| 20:59.18 | RagebeastRT | err event |
| 20:59.31 | ZorbaTHut | rage, right now, you can access unit.details everywhere, as long as you're requesting data about a unit that is known to exist |
| 20:59.50 | ShaunO`work | havent tried being fancy yet but the two cases so far were iterating over an itemdetail collection to run specific code dependent on the key, and the other being registering an event handler for a keypress style event |
| 20:59.59 | ShaunO`work | but its just sugar |
| 21:00.02 | ShaunO`work | no harm |
| 21:00.06 | RagebeastRT | I tried to in certain functions. Unit.Lookup worked fine but details failed everytime unless I was in an event function |
| 21:00.13 | seebs | The next big function for LibBaggotry is a bit of UI hookery. It's going to acquire the ability to display filters as windows of editable criteria and checkboxes and stuff. |
| 21:00.38 | seebs | Which means it'll have to have a table of recyclable windows. |
| 21:00.56 | RagebeastRT | table.insert(Event.Unit.Add,{rhrUnitDetails, "RiftHealer", "rhrUnitDetails"}) I could only access .details in that function |
| 21:01.00 | ShaunO`work | i keep using a . instead of a : though |
| 21:01.08 | seebs | But eventually it'll be able to do stuff like save your preferred set of filters, display them graphically, and so on. |
| 21:01.32 | ZorbaTHut | I'm sort of surprised that nobody's made a onebag-type addon yet. especially given how quickly DPS meters showed up |
| 21:01.50 | RagebeastRT | I would rather see a mouseover quantity in bags then a 1 bag |
| 21:01.53 | ShaunO`work | i think there's a few limiting factors in the api to make a seamless experience |
| 21:02.04 | ZorbaTHut | shaun, there probably are, but until someone tells me what they are I can't fix 'em :P |
| 21:02.07 | ShaunO`work | the most major is the tooltip |
| 21:02.17 | RagebeastRT | like mousing over witchweave cloth and it will tell me total in all bag, total in bank, total in alts |
| 21:02.17 | ZorbaTHut | and a few of the limiting factors mentioned have been actually solvable |
| 21:02.22 | ZorbaTHut | yeah, tooltip needs some love |
| 21:02.46 | RagebeastRT | I miss bagnon more then any other addon from wow |
| 21:02.58 | ShaunO`work | but i haven't tried to implement the functionality of onebag though so i dont know what the other hurdles are but i think one of them was not being able to drag onto native rift ui elements |
| 21:03.07 | ShaunO`work | ie drag an 'item' onto the actionbars |
| 21:03.15 | ZorbaTHut | that actually should work fine, via Command.Cursor() |
| 21:03.26 | ShaunO`work | i'll have a play around tonight |
| 21:03.30 | ZorbaTHut | in fact that's why Command.Cursor() exists :) |
| 21:03.46 | ShaunO`work | wasn't really my interest but i do miss not having an all in one bag |
| 21:03.55 | RagebeastRT | we cannot insert strings into cursor though right? |
| 21:04.02 | ShaunO`work | my main interest is gear comparisons and auction house tracking |
| 21:04.41 | ZorbaTHut | rage, "cursor" is literally the item held in your cursor. I'm not sure what inserting a string would do |
| 21:04.46 | Kreiri | "details failed everytime unless I was in an event function" - according to Zorba, Inspect.Unit.Detail("player") should always have at least player's name. I can name at least two cases when it returns nil. Not even an empty table, just nil. |
| 21:04.53 | ZorbaTHut | the tooltip isn't going to be modifiable ever, but there will eventually be some way to figure out where it is |
| 21:04.58 | ZorbaTHut | kreiri, yeah, that's on my todo list |
| 21:05.03 | ShaunO`work | oh never modifiable :O |
| 21:05.05 | RagebeastRT | nevermind, I am thinking tooltip when I see cursor |
| 21:05.25 | ZorbaTHut | shaun, the intended solution is "put your extra tooltip data in a little augmented segment next to it" |
| 21:05.30 | ShaunO`work | not so bad though, easy enough to add a side-by-side tooltip |
| 21:05.53 | RagebeastRT | oh, tooltip location would work if you could setpoint to it |
| 21:07.06 | ZorbaTHut | rage, yep, although it'll be a *little* funky to work with since it may be on the edge of the screen |
| 21:07.40 | RagebeastRT | I think we could code a fix for that if we just had it to work with |
| 21:07.54 | RagebeastRT | if frame > uiparent:getright and such |
| 21:08.37 | ZorbaTHut | yeah, it's definitely solvable |
| 21:08.59 | RagebeastRT | and now...frame:Drag(true/false) pls |
| 21:09.14 | ShaunO`work | but yeah i quite like the rift addon api its decent to work with, works well and best of all is fasstt |
| 21:09.16 | RagebeastRT | im TinnerKB on the forums... but all that I posed came down to the drag problem |
| 21:09.23 | RagebeastRT | err posted |
| 21:09.50 | RagebeastRT | oh and you cannot close the main window in combat, so you are stuck with it til after combat if there is a problem |
| 21:10.50 | RagebeastRT | I can see where that would be a prob for you though, as someone could stack 20 frames on top of each other, setvisible(false to all but the one with lowhealth |
| 21:11.12 | ZorbaTHut | *nod* yeah, I spent like a week coming up with exploits that could be done with mouseover before I started coding it |
| 21:11.24 | seebs | Yeah, I have been bitten a couple of times by Inspect.Unit.Detail("player") yielding nil. |
| 21:11.30 | RagebeastRT | frame:Drag would be nice if it was hardcoded |
| 21:11.31 | ZorbaTHut | though keep in mind that, while a restricted frame must have restricted parents, it *doesn't* need restricted children |
| 21:11.40 | RagebeastRT | that way it couldnt be exploitable |
| 21:11.47 | ZorbaTHut | using that you should be able to visually rearrange whatever you like |
| 21:11.48 | RagebeastRT | yeah tried that |
| 21:11.58 | ZorbaTHut | just can't move the important bits around |
| 21:12.00 | RagebeastRT | I could move my window, the click frames stayed where they were |
| 21:12.08 | ZorbaTHut | *nod* yeah |
| 21:12.14 | RagebeastRT | so kinda defeated the purpose :P |
| 21:13.19 | ZorbaTHut | I can't offhand come up with an exploit for "if you are currently dragging a frame that is drag-hooked via the official API, then all existing restricted frames move to their new locations as if you were out of combat" |
| 21:13.39 | ZorbaTHut | since it relies on the user making a drag motion which isn't the kind of thing that could be part of automation |
| 21:13.46 | ZorbaTHut | but I'll have to think about that one |
| 21:13.48 | RagebeastRT | not if it was hardcoded and could only be true/false set |
| 21:14.30 | seebs | It isn't exactly exploitable, but... Hmm. |
| 21:14.36 | ZorbaTHut | it'd be a little more thorough than that - it would also let you pass it a table that it could store coordinates in, for use with SavedVariables, and it'd have some way to distinguish the "grab" handle from the "move" frame - but yeah, that'd be the gist of it |
| 21:14.44 | seebs | Imagine that I have a thing where every time you drag, all the windows are shuffled in some elaborate way. |
| 21:14.45 | RagebeastRT | if frames moved to their setpoint in combat it would be fine too, just hardcode SetpOint to not work in combat |
| 21:14.55 | seebs | That's not really exploitation, but it's awfully flexible. |
| 21:16.11 | ZorbaTHut | seebs, yeah, it's moving kind of uncomfortably close to that line of "automation" . . . but I can't think of a way that it could actually be used to automate |
| 21:16.32 | ZorbaTHut | I'd probably lock mouse input out of the addon system while dragging was occuring, just for safety's sake |
| 21:16.34 | seebs | I think... There's a category of things which isn't really "automation" but is fuzzy. |
| 21:16.39 | seebs | Like, okay, here's an example. |
| 21:16.55 | seebs | Say I want a raid-rebuffer. I know I'm gonna have to rebuff a bunch of people, but it's hard to see which ones. |
| 21:17.10 | ZorbaTHut | hmm, you could definitely set it up to do drag-click-drag-click to choose abilities, though, and with a 1.5s gcd that's probably a bit exploitable. |
| 21:17.11 | seebs | Every time I drag a window, all the frames that need my buff are moved to the left, and the others to the right. I can now click down the list. |
| 21:17.25 | ZorbaTHut | mmm, good point there |
| 21:17.33 | RagebeastRT | think of disabling SetPoint in combat instead of dragging. We could drag main window, everything already attached moves with it but you couldnt code a frame to move otherwise |
| 21:17.48 | ZorbaTHut | rage, well, it already is disabled for restricted frames |
| 21:17.50 | seebs | D'oh. I forgot. I am in favor of addons. I should be trying to get the ability to automate the game so I can finally compete in content that's too hard for me. |
| 21:18.13 | seebs | I feel like I am not living up to the expectations of my fans. |
| 21:18.17 | RagebeastRT | yeah but so is the parent |
| 21:18.20 | ZorbaTHut | the problem is that a restricted frame can be SetPoint() to a non-restricted frame. if *that* frame moves, then the restricted frame just doesn't move along with it, which works fine for now |
| 21:18.35 | ZorbaTHut | but if I were going to relax the restriction then that gives a whole lot of power |
| 21:18.43 | ZorbaTHut | in retrospect perhaps I shouldn't have allowed that |
| 21:19.17 | RagebeastRT | well my healer addon is restricted from context all the way down to the click frames |
| 21:19.46 | seebs | hmm |
| 21:20.02 | RagebeastRT | like 4 levels of restricted not to mention the frames[20] that are restricted |
| 21:20.06 | seebs | Possibly upon restriction, a frame which is setpoint to a nonrestricted frame should be resetpointed so that it's setpointed absolutely to the base window. |
| 21:20.22 | ZorbaTHut | I don't want to go mucking about with people's setpoints like that |
| 21:20.24 | seebs | Oh, but that doesn't help with the case where you setpoint later. |
| 21:20.25 | seebs | Yeah. |
| 21:20.40 | seebs | Okay, here's my proposed addon: Target Pong. |
| 21:20.46 | RagebeastRT | I got it. make context, make frame, set frame:Grandfather :P |
| 21:20.55 | ZorbaTHut | alright here's a crazy overcomplicated solution - if you move a restricted frame via dragging, then all the other restricted frames which are attached to it are allowed to move along with it, as are restricted frames attached to those, ad infinitum |
| 21:21.06 | ZorbaTHut | but a single non-restricted frame in the middle breaks the chain and prevents further updating |
| 21:21.15 | ShaunO`work | just a quick one, does the api allow chat reading and writing? |
| 21:21.19 | RagebeastRT | sounds like a coders problem |
| 21:21.24 | Kreiri | no |
| 21:21.25 | seebs | Your target window moves around the screen bouncing off other unit frames, and there's a thing glued to the left side of the screen which tracks your mouse vertically. This is TOTALLY a good addon. |
| 21:21.29 | ZorbaTHut | that way you get the ability to "drag" in combat but not anything else |
| 21:21.31 | Kreiri | not yet |
| 21:21.31 | seebs | Shaun: Not yet. |
| 21:21.33 | ShaunO`work | ty |
| 21:21.43 | ZorbaTHut | shaun, not yet, though I'm finally getting some of the infrastructure in place. |
| 21:21.47 | seebs | Hmm. That seems reasonably sane. It's well-defined, anyway. |
| 21:22.03 | ZorbaTHut | chat writing is going to be a fascinating thing to make nonabusable |
| 21:22.09 | ShaunO`work | :D |
| 21:22.19 | RagebeastRT | oh yeah, here comes gold sellers |
| 21:22.34 | ShaunO`work | at the very least a restricted method of external plugin communication |
| 21:22.59 | RagebeastRT | actually dont think chat interaction will really change the spam, there are 3rd party mods that will do the same |
| 21:23.14 | ShaunO`work | well, reading chat = bots |
| 21:23.18 | ShaunO`work | intelligent bots |
| 21:23.23 | ZorbaTHut | yeah, an addon channel is something I really, really want to design |
| 21:24.01 | ShaunO`work | just opens up a lot of possibilities |
| 21:24.01 | ZorbaTHut | I'd also love to add some limited serverside storage, though I have no idea what the API for that will end up being |
| 21:24.07 | ShaunO`work | serverside storage :O |
| 21:24.09 | ShaunO`work | that'd be hard core |
| 21:24.40 | Kreiri | just imagine an Eliza bot in Level 50 channel. |
| 21:24.41 | seebs | I have proposed a pretty trivial serverside storage API. |
| 21:25.07 | seebs | Command.Storage.Stash("key", "value"); Inspect.Unit.Storage(unit, key) => value |
| 21:25.19 | ShaunO`work | sounds scary |
| 21:25.28 | Kreiri | space issues |
| 21:25.32 | ShaunO`work | i'd hate to think of the cpu cycles on the server side in large collections |
| 21:25.34 | ZorbaTHut | yeah, that's roughly what I'd be going for, but it brings up questions about how space is organized and how running out of space is handled |
| 21:25.35 | ShaunO`work | looking up keys |
| 21:26.05 | seebs | If value has to be a string-or-number-or-boolean (no functions, no tables), and there's a cap, it's not TOO bad. |
| 21:26.27 | seebs | Pop cap per server is roughly 2k, figure 90% of them are whichever faction the forum whiners say is on top, so 1800. |
| 21:26.45 | ShaunO`work | i wouldnt worry about the space personally, i'd worry about key lookups |
| 21:26.48 | seebs | 1800 players, say they have 1MB of storage space each, that's almost 2GB. That's trivial in modern hardware. |
| 21:26.58 | ShaunO`work | and how data is transferred back and forth |
| 21:27.09 | ZorbaTHut | I think our server techs would object to a claim that 2GB is trivial :P |
| 21:27.25 | ShaunO`work | it'd all have to be stored in memory |
| 21:27.28 | seebs | Heh. |
| 21:27.52 | seebs | Well, you could set the numbers a fair bit lower. At 100K or so, 200MB is pretty small on modern hardware. |
| 21:27.57 | RagebeastRT | back to my original question Zorba, will libunitchange linrary reduce the overhead I am getting. In a 20 man warfront, I get 50% or more cpu spikes. I had to throttle the event funtion to .5 secs and still spikes |
| 21:28.05 | ZorbaTHut | realistically you'd be looking at a *much* smaller amount per player - I'd probably start with 1k just for load testing |
| 21:28.12 | ZorbaTHut | rage, the real question is, what are you doing during each of those events? |
| 21:28.12 | RagebeastRT | grr, cant type |
| 21:28.24 | RagebeastRT | very simple function |
| 21:28.26 | ZorbaTHut | libunitchange on its own won't improve your CPU, but it provides some premade tools that can be *used* to improve your CPU |
| 21:28.31 | Kreiri | use pastebin |
| 21:28.53 | RagebeastRT | I check for role, class and name. I then only update the frames in the table I made |
| 21:29.06 | RagebeastRT | if 2 ppl, table is 2 sizes |
| 21:29.15 | ZorbaTHut | I *suspect* you're doing more than you think you are :) |
| 21:29.49 | Kreiri | are you calling .details every .5 sec? |
| 21:29.56 | ShaunO`work | it'd do a key value(strings only) store with a fixed size value field and a fixed size array length, force all key lookups on the client side by caching the data on login and sending update requests to the server to be queued with a callback to say its done |
| 21:29.58 | RagebeastRT | I actually made it table=oldtable and if no group member changes it wouldnt do anything and still it gets high |
| 21:30.09 | ZorbaTHut | seriously, post your code on pastebin |
| 21:30.14 | Kreiri | you only need to get unit's details when it's added |
| 21:30.21 | RagebeastRT | sec |
| 21:30.33 | Kreiri | if it's only class, role and name |
| 21:30.50 | Kreiri | also, I suspect you are calling .detail for every frame |
| 21:31.19 | Kreiri | make a table of unit ids or specs you are interested in and call details on that |
| 21:32.04 | RagebeastRT | ifyou can make sense of it..http://pastebin.com/CFBNV7P7 |
| 21:32.05 | Kreiri | it's even in docs: details = Inspect.Unit.Detail(units) -- table <- table |
| 21:32.23 | RagebeastRT | there is a .5 sec throttle when out of combat |
| 21:32.33 | RagebeastRT | in combat the function doesnt even run |
| 21:33.07 | RagebeastRT | Kreiri, it was and then I added the throttle |
| 21:33.16 | Kreiri | "for v, k in pairs(rhr.GroupTable) do details = Inspect.Unit.Detail(k)" |
| 21:33.54 | ZorbaTHut | yeah I'd definitely turn that into event-based instead of polling. it's a little more work but it results in far faster addons |
| 21:34.19 | Kreiri | seriously, try making a table of unit specs and call inspect.unit.details(unitspecs) |
| 21:34.23 | RagebeastRT | would love to but not sure how |
| 21:35.13 | RagebeastRT | most of the code on top of if same(rhr.GroupTable, rhr.OldTable) == false is just quick checks for raid, group, group has pet, then solo |
| 21:36.56 | ZorbaTHut | rage, here's a suggestion: grab Trion Development Tools from the forum, install it, run it, type "/eventlog Unit" |
| 21:37.02 | ZorbaTHut | then run around doing things for a minute |
| 21:37.23 | Kreiri | register changes of group01, group01.pet,..., group20, group20.pet in lib.unit.change |
| 21:37.28 | RagebeastRT | wont that just grab only player, player target and player mouseover? |
| 21:37.32 | ZorbaTHut | all of those events are triggered when things change. you generally don't need to do much updating when things *haven't* changed, all you need to do is catch the right events when things actually happen |
| 21:38.17 | ZorbaTHut | most of the event system works with unit IDs ("u1234567812345678") not specifiers ("player.pet"), but that's where libunitchange comes in - it can notify you when a specifier starts referring to a new ID. And then you can set up a little lookup table of some kind so you can instantly determine whether an event is something you care about. |
| 21:38.26 | RagebeastRT | Kreiri, thats what my initial question was, I just dont know where to use the return results. Right now I am using the event add funtion |
| 21:38.54 | Kreiri | a moment, I post a snippet from KBB on pastebin |
| 21:39.00 | RagebeastRT | ok this.... |
| 21:39.14 | RagebeastRT | for a, b in pairs(rhr.UnitsTable) do only looks at group01,,group20 |
| 21:39.28 | RagebeastRT | nothing else, no .target, .pet or anything |
| 21:39.31 | Kreiri | http://pastebin.com/BYcA3QxL |
| 21:39.46 | RagebeastRT | so I am limiting what units I am working with |
| 21:41.15 | RagebeastRT | Kbb.groups is a preset table like what my UnitsTable is right? |
| 21:41.25 | RagebeastRT | just group01...grouo20? |
| 21:42.03 | Kreiri | KBB.units = {["player"] = true, ... |
| 21:42.12 | Kreiri | you get the idea |
| 21:42.31 | Kreiri | Kbb.groups is the table of my bar groups |
| 21:42.39 | RagebeastRT | I meant units |
| 21:42.57 | RagebeastRT | but your units is a 2 column table, mine is just 1 |
| 21:43.24 | RagebeastRT | I use the Lookup to see if the loop unit is true |
| 21:44.59 | RagebeastRT | so I guess the question I am asking is: Does the libunit change return just 1 unit and I can quesry and change based on that unit or what |
| 21:45.56 | RagebeastRT | if I have a group01 through 07 and 08 adds, I would only want to get details from him. And if group04 left, I could then table.remove him |
| 21:46.48 | RagebeastRT | In a lfg dungeon, I dont have a prob. Its big raid warfronts and full publics |
| 21:52.54 | *** join/#riftuidev lXl (~Main@cpe-173-172-35-32.tx.res.rr.com) |
| 22:34.59 | *** join/#riftuidev Torhal (~textual@67.58.77.206) |
| 22:35.39 | *** join/#riftuidev Torhal (~textual@67.58.77.206) |
| 22:35.57 | *** join/#riftuidev Torhal (~textual@67.58.77.206) |
| 22:35.57 | *** mode/#riftuidev [+v Torhal] by ChanServ |
| 23:03.39 | *** join/#riftuidev Aurorus (~Aurorus@unaffiliated/aurorus) |
| 23:03.49 | *** join/#riftuidev Torhal (~callahan@67.58.77.206) |
| 23:03.49 | *** mode/#riftuidev [+v Torhal] by ChanServ |
| 23:21.12 | *** join/#riftuidev MentalPower|Work (~chatzilla@WoWUIDev/Norganna/Administrator/MentalPower) |
| 23:24.53 | *** join/#riftuidev Kraps (~chatzilla@ool-4572cb11.dyn.optonline.net) |