| 00:04.39 | Amadeus- | seebs, you were talkign about games being cancelled the other day: https://forums.station.sony.com/vg/index.php?threads/with-heavy-hearts-we-announce-the-sunset-for-vanguard.71211/ |
| 00:05.12 | Amadeus- | it's really stupid ..basically the devs broke a bunch of stuff trying to "upgrade" the game ...and now it's not worth the time/money to fix it... so they're just going to shut it down. |
| 00:06.43 | Noshei | wow, they must have really broken it pretty good |
| 00:07.34 | Ivnedar | :( |
| 00:07.57 | seebs | Wow. |
| 00:09.55 | Amadeus- | yea |
| 00:09.57 | Amadeus- | it's really sad |
| 00:10.03 | Amadeus- | the game was fine a year ago and on auto-pilot |
| 00:10.48 | Amadeus- | they wanted to redo the network code to make raiding more viable ...but someone broke something bad and they don't have anyoen good enough to fix it (easily/quickly) I guess |
| 00:11.05 | Amadeus- | sadly, I doubt they can release the server source code even if they wanted since it uses Unreal source. |
| 00:26.23 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 00:26.23 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 00:32.54 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 00:32.54 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 00:48.02 | Amadeus- | -- |
| 00:48.18 | Amadeus- | wow, talk about trying really hard to tap into the EQ1 nostalgia: https://www.kickstarter.com/projects/1588672538/pantheon-rise-of-the-fallen |
| 01:50.49 | ZorbaBeta | they could probably release the server source, from what I understand Unreal-engine MMOs generally use the Unreal frontend for nothing more than rendering and physics. and maybe an object model. but it's very rare that a company *will* do something like that. |
| 01:56.07 | seebs | Hmm. I see occasional assertions that BananAH is why the AH had to get limited, but that seems odd, if only because I don't think the newly-added AH throttle even affected it. |
| 01:56.50 | ZorbaBeta | if anything, bananah is more efficient than our built-in auctionhouse interface |
| 01:57.27 | seebs | Interesting. |
| 01:58.19 | seebs | I am torn. I'm really having a fun crafting groove tonight in FF14. But I also wanna hop back into Rift and find out whether the guy who paid me a REX to troll someone realizes that not only did I give the guy the REX, but that the entire thing was staged. |
| 01:58.20 | Amadeus- | ZorbaBeta, what surprises me is not that they wouldn't go open source (probably embarassed to let people see ..or afraid they'll judge) ...but, am surprised they're not offering to LEASE it for use by individuals or smaller companies wanting to make a name for themselves. |
| 01:58.51 | ZorbaBeta | virtually nobody wants to lease MMO code. you need a lot of inhouse experience to keep the thing running, and then what happens when the lease expires? |
| 01:59.14 | seebs | That, and... It's a competitor to something else you're shipping, often enough. |
| 01:59.16 | ZorbaBeta | I'm not sure I've ever heard of leasing an MMO engine, to say nothing of an entire MMO. |
| 01:59.43 | seebs | I also strongly suspect that there are other motives; I think ncsoft's reason for not selling Paragon to anyone was that they knew perfectly well that anyone else could have been more successful with Paragon's staff and code. |
| 02:01.09 | Amadeus- | I dunno ..VG is pretty old and it's just standard stuff ...MMO concepts pasted onto Unreal. I imagine a couple talented coders could handle it in their spare time ....may not be any new content (hardly) (ever) ..but, as long as the players were patient for it to get up and running again, I don't think it'd be impossible if someone had the network and server infrastructure. |
| 02:01.53 | ZorbaBeta | you can't judge the difficulty of maintaining software by the complexity of gameplay. who knows what's lurking in that codebase |
| 02:02.08 | seebs | The Shadow knows! |
| 02:02.21 | ZorbaBeta | I'm not sure even the Shadow knows that one :V |
| 02:02.29 | seebs | Back in college, we used to play hunt, and it truncated nicknames. |
| 02:02.38 | seebs | So someone wanted to be Lurking Evil, and ended up as "Lurking Ev". |
| 02:02.40 | Amadeus- | ZorbaBeta, I know the codebase fairly well ;) |
| 02:02.57 | seebs | For VG? |
| 02:02.59 | Amadeus- | reversed most of it at one point or another ...other than the unreal specific stuff |
| 02:03.00 | Amadeus- | yea |
| 02:03.14 | seebs | I didn't know it was available... |
| 02:03.23 | Amadeus- | it's truly just an Unreal game with extra stuff...very straightforward |
| 02:03.38 | Amadeus- | however, I don't know how easy it would be to "fix" what they "broke" recently that caused this shutdown. |
| 02:03.40 | ZorbaBeta | the difficult part isn't that, it's the server code. MMO client code is pretty dang simple |
| 02:04.00 | Amadeus- | that's true ....you could be right. |
| 02:04.08 | seebs | Yeah, like, say you wanted to rework the rendering engine for an MMO client. Couple weeks, tops. |
| 02:05.16 | Amadeus- | I still would guess they were basically using Unreal stuff for the server as well ..just altered. I honestly don't think they ever had a "really talented" C++ coder on that team if you want the truth. I mean, they had folks and good folks ..but I don't think they ever had a prodigy on that team that would have rewritten it from what was provided. |
| 02:05.30 | Amadeus- | I think that's why it was broken recently ...someone DID try to redo some of the Unreal stuff so the network could handle raids |
| 02:05.44 | Amadeus- | the game never really handled raids because of network limittaions |
| 02:06.02 | ZorbaBeta | from what I understand, you can't really just use Unreal for the server. The requirements are way too different. But that's what I've heard, I may be wrong. |
| 02:06.45 | Amadeus- | That's probalby why the game was unplayable at launch ..haha ...all those years ago. |
| 02:07.05 | Amadeus- | game failed so bad at launch because no one could play it ...they fixed a lot of that in the next year afterwards ..but it was too late |
| 02:09.32 | Amadeus- | but, of course, I always assumed htat was client issues ..not server ...so, yea, no clue ....you might be right though ..the server code might be a mess to get a handle on and/or fix. |
| 02:10.15 | Ivnedar | unreal has a server engine too? |
| 02:10.16 | Ivnedar | hrrm |
| 02:10.44 | ZorbaBeta | I don't think they do |
| 02:10.56 | ZorbaBeta | I strongly suspect that unreal-engine MMOs are just using Unreal for the client |
| 02:11.16 | Amadeus- | well, the packet sending/receiving is built into the unreal sections of the client, I can tell you that |
| 02:11.22 | Amadeus- | it's all integrated very heavily |
| 02:11.38 | Amadeus- | within the unreal dlls ..not just the custom code I mean |
| 02:11.41 | Ivnedar | an mmo server isn't just a client with a modified config |
| 02:12.13 | Amadeus- | There are other games that have server/client models though ...I mean, "pure" unreal stuff ..like the game "Unreal" ....I thought. |
| 02:12.27 | ZorbaBeta | sure but that's easy enough to deal with. either replace the packet format code or just take that code and slap it into your server |
| 02:12.29 | Amadeus- | simple packet communication I think ...but yea, right? |
| 02:12.49 | Imhothar | for all you know the server could be written and maintained by 20 undergraduates or a single prodigy. you couldn't really tell a difference |
| 02:12.51 | Amadeus- | yea of course ...but, I'm just saying I think Unreal has something ....maybe not much. |
| 02:13.15 | ZorbaBeta | http://udn.epicgames.com/Three/NetworkingOverview.html neat looks like I've got a reading project :V |
| 02:14.05 | Amadeus- | unreal packet classes have interesting names |
| 02:14.14 | Amadeus- | just sec |
| 02:15.43 | Amadeus- | like this: http://pastebin.com/HPWaP5rh |
| 02:16.35 | Imhothar | not a surprise, really |
| 02:16.48 | Amadeus- | or maybe something a little more complex: http://pastebin.com/Cb8re9m9 |
| 02:17.34 | Imhothar | how is that more complex? |
| 02:17.43 | Amadeus- | well, not complex ...just more stuff I mean |
| 02:18.07 | Imhothar | 5 or 15 members, not really a difference |
| 02:18.42 | ZorbaBeta | Well, the protocol definitely supports MMO-ish behavior. I'm not at all convinced that the server-side code would be capable of scaling up to an MMO's worth of load |
| 02:19.00 | Imhothar | it's more interesting to see they use polymorphic types for serialization |
| 02:20.48 | ZorbaBeta | it's also using a pretty obsolete networking model for MMOs - UDP is dead for all but the highest-speed action games, and maybe even for those. everyone just uses TCP now |
| 02:21.27 | Amadeus- | yea, Vanguard was released in 2006 I think it was ...so developed starting probably in 2003-2004 or something? |
| 02:21.38 | Amadeus- | maybe even earlier ...they took FOREVER I remember that |
| 02:22.23 | Amadeus- | yea, I think they announced the name not long after EQ1 came out |
| 02:27.17 | Ivnedar | oh wow, the vanguard launch sounded terrible :( |
| 02:27.47 | Ivnedar | 150 employees at launch, general staff meeting 4.5 months after launch in the carpark, employees told they're all fired ... |
| 02:29.14 | Amadeus- | yep |
| 02:29.17 | Amadeus- | I remember taht |
| 02:30.02 | Imhothar | god bless america |
| 02:30.17 | Imhothar | now don't say it was somewhere else |
| 02:30.47 | Ivnedar | California |
| 02:31.04 | Imhothar | ... the land where treating employees like shit is patriotic |
| 02:32.54 | Imhothar | "In the beginning, (...) Doom and Duke Nukem." |
| 02:32.56 | Imhothar | I like that |
| 02:33.01 | Noshei | ZorbaBeta: I just heard that "Trion" has promised in 3.0 they will promote a better graphics engine |
| 02:33.02 | Imhothar | totally out of context but a valid quote |
| 02:33.38 | ZorbaBeta | depends on what "better" means, that's believable :V |
| 02:33.46 | Ivnedar | I assume initial aim for 3.0 is end of year |
| 02:34.01 | ZorbaBeta | hell technically every patch has had a "better" graphics engine |
| 02:34.33 | Noshei | well I get the impression that people think rift is going to go to a different engine with 3.0 |
| 02:34.48 | ZorbaBeta | out of curiosity, any idea where the root belief of that came from? |
| 02:35.30 | Ivnedar | probably people blaming gamebryo for everything, someone saying that it needs to be replaced, someone taking that as "it is being replaced", repeats it, it's now gospel |
| 02:36.03 | ZorbaBeta | sigh :P |
| 02:36.29 | ZorbaBeta | (although for what it's worth, gamebryo really is responsible for a good amount of the performance problems - I'm in the middle of gutting and revamping the main rendering code) |
| 02:37.08 | Ivnedar | There's no way you'd be able to merge Trion's customisations into latest gamebryo |
| 02:37.11 | Ivnedar | I'm assuming |
| 02:37.20 | ZorbaBeta | nope, totally impossible |
| 02:37.35 | Noshei | ZorbaBeta: you mentioned you might have a change for the flickering rocks and color changing bugs right? |
| 02:39.47 | ZorbaBeta | yep. might not be live in the next patch |
| 02:40.04 | ZorbaBeta | though it's possible I'll end up sneaking these changes in as a config file parameter - they've been really surprisingly stable |
| 02:40.33 | Noshei | it amazes me how untrusting the community seems to be of the dev team |
| 02:40.39 | Ivnedar | oooh, be really sneaky, get the patcher to enable them if it's given -dx9 as a parameter :D |
| 02:40.47 | Ivnedar | hides |
| 02:41.07 | ZorbaBeta | hah |
| 02:41.12 | ZorbaBeta | community would kill me :V |
| 02:41.20 | Ivnedar | not if they work! |
| 02:41.35 | Imhothar | they know where your house lives?! |
| 02:42.06 | Imhothar | hmm doesn't really work in English |
| 02:42.23 | Noshei | lol |
| 02:42.36 | Noshei | translated from german? |
| 02:42.47 | Imhothar | it's a joke |
| 02:43.45 | seebs | There was some feedback on the graphics changes from someone who compared frame rates and got noticeable improvements. |
| 02:44.11 | Imhothar | we have two separate words for "live" as in "be alive" and in "living at a location". it's using the second |
| 02:44.32 | Noshei | ah |
| 02:46.08 | Noshei | why is it that rifts population is so terrible |
| 02:46.18 | seebs | Is it? |
| 02:46.23 | Imhothar | show me a mmo community that isn't |
| 02:46.30 | seebs | Faeblight's been mostly better lately. I mean, we still have some jerks, but it's definitely improved. |
| 02:46.38 | Ivnedar | I assume you mean attitudes/personality, not numbers? |
| 02:46.39 | Noshei | seems like almost everything in 60 chat is people arguing about some stupid thing or another |
| 02:47.20 | Imhothar | http://xkcd.com/386/ |
| 02:47.23 | Imhothar | this is why |
| 02:47.24 | Noshei | I've never paid much attention to the community in mmo's that I've played |
| 02:47.39 | Ivnedar | They're likely the same actual people on every game :P |
| 02:48.21 | seebs | The gap between Faeblight in mid-November and Faeblight this last week has been AMAZING. |
| 02:49.39 | ZorbaBeta | seebs, I've actually heard a few people saying that the PTS is significantly faster. I'm taking this with a grain of salt because there's obviously fewer people on PTS, but . . . there really are some nice performance improvements coming up in the next major patch |
| 02:49.43 | ZorbaBeta | so it could be legitimate |
| 02:50.46 | seebs | This one had reasonably plausible straight comparisons between similar things. Showed about a 10% increase in frame rates, I think. |
| 02:51.08 | Noshei | and all those improvements go right out the door when people get 20 of those laser kegs in their dimensions! |
| 02:52.22 | Ivnedar | s'ok, that's limited to dimensions |
| 02:52.31 | Ivnedar | imagine it as an item you could use in raid :D |
| 02:52.45 | Noshei | until it is an item you can use in raids |
| 02:53.07 | Ivnedar | actually, worse, encounter designers start using those effects in raids |
| 02:53.55 | Noshei | I think 20 people using something that spawned one of those laser balls in a raid would be worse than most of what the designers would do |
| 02:54.15 | Noshei | though a couple abilities on abominus are pretty crazy on graphics |
| 02:55.05 | Ivnedar | still not a fan of the new regulos purple fire |
| 02:55.10 | seebs | I already have a lot of trouble with abilities that use very-different visual indicators. |
| 02:55.16 | Ivnedar | must remember to complain about that more :P |
| 02:55.30 | seebs | There was some hunt rift I was doing where there's wandering-AoE during the fight which have a faint purple circle on the ground. |
| 02:56.04 | Noshei | not sure what you mean Ivnedar what purple fire on regulos? |
| 02:56.14 | Noshei | or do you mean ultane? |
| 02:56.39 | Ivnedar | regulos |
| 02:56.56 | Ivnedar | a couple of hotfixes ago, new purple flames were added, coming out of the pit he's in |
| 02:57.14 | Ivnedar | extending out a few meters |
| 02:57.26 | Noshei | ah |
| 02:57.47 | Noshei | didn't even notice |
| 02:58.08 | Imhothar | sigh moving badly organized SVN code to git is a pain |
| 02:58.25 | Ivnedar | we're using a hybrid sccs/git thing now |
| 02:58.27 | Ivnedar | shudders |
| 02:59.14 | Imhothar | sccs? |
| 02:59.24 | Ivnedar | yup! |
| 02:59.57 | Imhothar | like... *the* SCCS? |
| 02:59.59 | Ivnedar | http://en.wikipedia.org/wiki/Source_Code_Control_System |
| 03:00.00 | Ivnedar | yes |
| 03:00.11 | Imhothar | .... why? |
| 03:00.37 | Ivnedar | it works, it lives with the code base, it's installed on every *nix platform we use, why not? :) |
| 03:00.51 | Imhothar | because... reasons? |
| 03:01.09 | Ivnedar | it's mostly for config management, not much code using it |
| 03:02.55 | Ivnedar | today's outage here is going well, got a list of 10+ tasks to get done in a 12 hour outage window |
| 03:03.04 | Ivnedar | we're 8 hours in and have almost done the first one |
| 03:03.06 | Ivnedar | wee |
| 03:03.34 | Noshei | lol |
| 03:03.37 | Ivnedar | solaris patch decided to break libxnet on a bunch of hosts, thanks oracle |
| 03:03.49 | Ivnedar | no xnet = no ifconfig = no network |
| 03:05.04 | Imhothar | any idea if it's possible, given two disjoint branches to somehow stitch them together but keeping their commits in chronological order? i.e. not just rebase one branch on the other |
| 03:05.53 | Imhothar | something's telling me there's a way with filter-branch and --parent-filter but... eh |
| 03:07.36 | seebs | I am not sure about chronological order, except trying to redo that by hand. |
| 03:07.42 | seebs | Probably scriptable. |
| 03:10.02 | Ivnedar | aha, worked out the issue. patch installs a new ifconfig, that's been built against a newer version of libxnet. The patch that installs the newer libxnet is neither recommended nor security, so hasn't been installed, it's also not in the dependancy list for the ifconfig patch ... |
| 04:01.46 | seebs | Wow, classy. |
| 04:13.17 | Noshei | so bored |
| 04:32.12 | seebs | I'm still at the initial stage of being-60 where it would be Very Not Fun to do randoms with pugs, so I have to wait around in case people I know happen to form a group and be looking. |
| 04:45.44 | Amadeus- | ZorbaBeta, myself and everyone in my guild is reporting a strange lag while "chatting" tonight. Like, you hit enter and your chat doesn't show up for 1-2 seconds. |
| 04:46.05 | Amadeus- | I don't really notice it otherwise ..just chatting ....and it's not just custom channels ..also things like 'guild' |
| 04:46.11 | Amadeus- | dunno if it's something new testing with last update? |
| 04:47.03 | Amadeus- | intermittent (sp?) |
| 04:49.06 | Noshei | Amadeus-: that is what we call server lag, I bet CQ was going on |
| 04:49.15 | Amadeus- | no, actually it wasn't |
| 04:49.26 | Amadeus- | honestly, it's the first time I've ever noticed it since I came back to the game like 5 months ago ..heh |
| 05:47.31 | *** join/#riftuidev Imhothar_ (~Imhothar@ppp-93-104-183-20.dynamic.mnet-online.de) |
| 05:47.40 | seebs | Oh, I see that a lot. |
| 07:11.23 | seebs | Oh, that sucks. |
| 07:11.41 | seebs | Someone I know was in the hospital for the entire duration of Mech Week, and now wants one of the mech mounts. |
| 07:11.42 | Noshei | ? |
| 07:11.44 | seebs | Bad timing. |
| 07:11.51 | Noshei | ah that sucks |
| 07:13.05 | Ivnedar | Tell them that on the bright side, they probably saved themselves a fortune in disappointing lockboxes! |
| 07:13.45 | Ivnedar | they've been back twice since haven't they? |
| 07:15.24 | seebs | I think so? |
| 07:15.55 | Ivnedar | probably still a chance at getting it then :D |
| 07:16.06 | Ivnedar | just have to notice the sale times |
| 07:17.55 | *** join/#riftuidev Natch_q (~Natch@c-88cde155.25-4-64736c10.cust.bredbandsbolaget.se) |
| 07:26.50 | seebs | Huh. Inspect.Experience() appears to just return nil at 60? |
| 07:30.27 | Noshei | shouldn't |
| 07:32.27 | Ivnedar | availability issue? |
| 07:32.46 | Noshei | ah nm, it should return nil at 60 |
| 07:32.53 | Noshei | Inspect.Attunement.Progress() |
| 07:32.57 | Noshei | that is for PA |
| 07:33.49 | Ivnedar | oh wow, when did that come in? |
| 07:33.56 | Noshei | 2.5 ish |
| 07:34.20 | Ivnedar | excellent, that fixes my soulwalk cooldown issue when spending pa :D |
| 07:35.08 | Noshei | lol nevermind |
| 07:35.30 | Noshei | 12-04-2012 is the date on the post zorba made that included it |
| 07:35.38 | Ivnedar | wow |
| 07:35.44 | Ivnedar | ok, I wonder why I've never seen that |
| 07:35.58 | Noshei | so, over a year ago, lol |
| 07:36.02 | Ivnedar | must look through api every so often |
| 11:28.28 | Repo | nkwardrobe: naifu * r154 (7 files in 2 directories): (9 lines trimmed from commit message) |
| 11:28.30 | Repo | bug fix: major fixes to the loading of equipment sets |
| 11:28.32 | Repo | bug fix: loading from wardrobe now works much better |
| 11:28.34 | Repo | bug fix: linked sets no longer cause mixed up sets during loading |
| 11:28.36 | Repo | bug fix: linked sets using wardrobe now actually work |
| 11:43.03 | Repo | nkwardrobe: naifu * r155 (2 files in 1 directory): |
| 11:43.03 | Repo | bug fix: applied new moving mechanism to get equip from wardrobe button |
| 11:44.05 | Repo | nkwardrobe: naifu V5.1.0 * r156 : Tagging as V5.1.0 |
| 11:48.10 | Repo | nkadvisor: naifu V3.4.6 * r254 : Tagging as V3.4.6 |
| 19:04.48 | *** join/#riftuidev Imhothar (~Imhothar@ppp-93-104-183-20.dynamic.mnet-online.de) |
| 19:50.36 | Amadeus- | hey Ivnedar or seebs ..what am I doing wrong here? http://pastebin.com/cqtGdRvx |
| 19:51.02 | Amadeus- | it prints "dispatch" ..but nothing else ...it's like it doesn't run the coroutine function at all |
| 19:51.18 | Amadeus- | am I just totally wrong on how all this works? |
| 19:52.12 | Imhothar | it helpds when you tell people what you expect to happen and what happens instead |
| 19:53.24 | Amadeus- | Imhothar, I expect _dispatch() to run ..and what's happening is that it's not running |
| 19:53.35 | Amadeus- | that's why I said it's printing "dispatch" but nothing else :) |
| 19:53.35 | Imhothar | you never start the coroutine |
| 19:53.46 | Amadeus- | oh, so "create" doesn't start |
| 19:53.53 | Imhothar | it helps to read the manual |
| 19:54.13 | Amadeus- | I read http://lua-users.org/wiki/CoroutinesTutorial |
| 19:54.45 | Amadeus- | it reads to me like 'create' starts it ..which is what I would expect. |
| 19:55.03 | Amadeus- | so..you're saying one creates it and then immediately 'resumes' it? |
| 19:55.10 | Imhothar | "To start the thread we use the coroutine.resume() function." |
| 19:55.16 | Imhothar | it says it right there |
| 19:56.00 | Amadeus- | ok |
| 19:56.05 | Imhothar | and just to squelch some expectations you might have: it's not "threads" in the classical tense |
| 19:56.17 | seebs | Yeah, "create" just makes the thing so it *can* be called. |
| 19:56.53 | Imhothar | Lua coroutines are not dispatched parallel threads but do collaborative multitasking |
| 19:57.49 | Amadeus- | ok ...makes more sense now. |
| 19:58.07 | Amadeus- | I still can't find that sentence "To start the thread we use the coroutine.resume() function." on that page ...lol |
| 19:58.42 | Imhothar | you call resume and it resumes your funtion where it yielded the last time until it yields again, then resume returns. it doesn't run in parallel |
| 19:58.49 | Imhothar | ctrl+f |
| 19:58.53 | Amadeus- | ok, now I see it |
| 19:58.54 | Imhothar | or cmd+f |
| 20:01.40 | Imhothar | the way you normally use coroutines in Rift is call resume() in event.system.update.begin/end until it's status shows it's finished |
| 20:02.05 | Amadeus- | yea ..I was just noticing that ...it's more clear now. I can use the /dump command in TrionDevelopmentTools as a guide |
| 20:06.42 | seebs | I really ought to make the pretty-dump in SlashPrint more available. |
| 20:06.53 | seebs | It's not at all well factored right now. |
| 20:07.07 | seebs | It produces prettier output than /dump. |
| 20:07.24 | seebs | I should also make it into a coroutine that can be called so it can do large hunks without performance warnings. |
| 20:09.26 | Amadeus- | I imagine this looks much better? |
| 20:09.26 | Amadeus- | http://pastebin.com/WepCp7ry |
| 20:09.37 | Amadeus- | (blatent copy of code from /dump ..lol) |
| 20:10.22 | Amadeus- | is 0.02 a good value to check against the watchdog? ..or is that too small? |
| 20:10.35 | seebs | I don't know much about the watchdog stuff. |
| 20:11.32 | Amadeus- | well, this function (dispatch) is causing 41% addon errors for imhobags because of watchdog performance warnings ...so trying to fix it |
| 20:11.43 | Amadeus- | doesn't happen to me ..must happen to folks with a TON of shit. |
| 21:04.24 | *** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net) |
| 22:02.08 | Amadeus- | there |
| 22:02.14 | Amadeus- | http://pastebin.com/iXGdr5pM |
| 22:02.23 | Amadeus- | Imhothar ^ ..that seems to work as I intended |
| 22:02.46 | Amadeus- | I have the value set to 0.11 rightnow for testing (so I can see it working) ...not really sure what the best value would be. |
| 22:03.17 | Amadeus- | oh..and ignore 'counter' ..forgot to remove it (old code) |
| 22:03.36 | Amadeus- | (wish things looked as good on pastebin as in uedit...) |
| 22:05.02 | Amadeus- | Ivnedar or ZorbaBeta, you see anything badly wrong with that paste |
| 22:05.04 | Amadeus- | ? |
| 22:06.04 | Amadeus- | ...I'm never quite sure how to handle removing things from a table that I don't need anymore in Lua .. setting to nil just seems like it would be inefficient in the long run because iterating the table would take longer and longer ...but, table.remove() just seems somehow 'unsafe' ...but, I'm not that good at Lua |
| 22:08.11 | Noshei | Amadeus-: why go through all that trouble with your bit at the end? |
| 22:08.28 | Amadeus- | how do you mean? |
| 22:08.52 | Noshei | you can do a single liner at the end of your addon with the Command.Event.Attach that references a function |
| 22:09.10 | Noshei | then have the rest in a named function |
| 22:09.25 | Noshei | it will attach to the event when the addon is loaded |
| 22:09.49 | Amadeus- | yea, I was just copying logic from the /dump command in the TrionDevelopmentTools for now ...can move it to it's own thing later |
| 22:09.52 | Noshei | no need to have a test to see if it has attached to the event |
| 22:10.03 | Noshei | ah |
| 22:10.37 | Amadeus- | although ..it's kindof neat having everyhing included in a single function that you want to run a in coroutine like that |
| 22:11.50 | Amadeus- | it's finally easy for me to see/understand how it all works together |
| 22:12.31 | Imhothar | setting a table value to nil is *the* way to remove it |
| 22:13.19 | Noshei | well the event bit is just a wasted if statement any time the function is called. Not really a big deal, but from an optizmation standpoint it is a wasted cycle checking it. |
| 22:14.09 | Imhothar | pair() and ipair() never gives you key/value pairs whrere the value is nil |
| 22:14.31 | Amadeus- | Imhothar, so iterating a table with 99 nils and 1 value isn't slower than a table with 1 value? |
| 22:14.47 | Amadeus- | or, true with pair()? |
| 22:15.34 | Imhothar | yeah there's no other way |
| 22:15.41 | Amadeus- | so, what's table.remove() for? |
| 22:15.53 | Imhothar | what table.remove does is move all values down a slot and set the last entry to nil |
| 22:15.53 | Amadeus- | I mean, I know what it does ..but, why is it not used? |
| 22:16.09 | Imhothar | prevents your table from having gaps when iteratirn 1..n |
| 22:16.32 | Amadeus- | wouldn't that make it faster in the long run though? Or, does having thousands of "nils" not really slow it down much? |
| 22:16.47 | Imhothar | faster than what? |
| 22:16.55 | Amadeus- | when you're iterating through it |
| 22:17.20 | Imhothar | something has to go horribly wrong to end up with a table with "thousands" of nil |
| 22:17.59 | Imhothar | setting a value to nil is just a semantic it doesn't mean it's what Lua does internally |
| 22:18.00 | Amadeus- | well, if I'm using a table as a list ....that I'm iterating to get values ...but when I'm done with a value, then I won't need it any more |
| 22:18.24 | Imhothar | as far as you are concnerned a table cannot have non-nil values |
| 22:18.35 | Imhothar | setting any table value to nil removes it from the table |
| 22:18.49 | Amadeus- | ah, ok |
| 22:19.05 | Amadeus- | so a table can't look like this: { nil, nil, "something", "else", nil, nil, 35 } |
| 22:19.19 | Amadeus- | the # of that table would be 3? |
| 22:19.31 | Imhothar | it's why you can query *any* key in a table, even if it doesn't exist. instead of throwing an error you get nil back |
| 22:19.45 | Imhothar | the # of that table is 0 |
| 22:19.58 | Imhothar | # only counts the consecutive number of non-nil elements starting at 1 |
| 22:20.29 | Amadeus- | yea, ok, that's what I need though ..I need soething that works like a queue ..but, I can't assume that it's ever empty ..I can only assume that the one thing I'm doing is finished, so remove that one. |
| 22:20.30 | Imhothar | thing is: don't care about it |
| 22:21.11 | Imhothar | *if*'you require a consecutive, ordered sequence of elements that's what table.remove is meant for |
| 22:21.20 | Amadeus- | ok |
| 22:21.21 | Amadeus- | thanks |
| 22:21.37 | Imhothar | it's a utility though, not an optimization or anything like that |
| 22:23.02 | Amadeus- | okie ..I think I see now |
| 22:30.14 | *** join/#riftuidev seebs (~seebs@home.seebs.net) |
| 22:30.14 | *** mode/#riftuidev [+o seebs] by ChanServ |