IRC log for #riftuidev on 20130712

04:21.48NosheiZorbaGama: is there a way to tie a custom event to a function?  So that when a function is triggered it causes an event to trigger.  At least I think that is what I want to do.
04:22.23NosheiSo I'm making a custom scrollbar, and I want an event to fire any time the :SetPoint() function happens for a specific frame.
04:23.29Nosheianyone got any ideas here?
04:24.20seebsI don't think you can directly do that, unless there's an event already triggered by that. Or maybe if you're allowed to override the SetPoint member.
04:24.59seebse.g., local f = frame.SetPoint; frame.SetPoint = function(…) my_hook(); return f(…) end
04:29.28Nosheithat would work
04:32.57Nosheiugg, maybe it wont work.
04:34.18Nosheican you do SetPoint(0,.2, someFrame, .3,1, -27, 52)?
04:37.50Nosheiugg, you can
04:54.00Nosheiseebs: how to handle all of the parameters though?
05:28.26*** join/#riftuidev Adelea (~martin@109.224.129.157)
05:29.00Adeleamorning
06:38.32NuSueymorning
07:49.59*** join/#riftuidev Kheldar (~Kheldar@217.108.40.234)
08:10.40*** join/#riftuidev Solsis01 (~Solsis00@ip98-186-89-123.fv.ks.cox.net)
10:19.52IvnedarI'm curious who the plat goes to when using brevanic portal
10:49.40NuSueyIvnedar: to the brevanit porta overseers!
10:49.42NuSuey*portal
10:58.11*** join/#riftuidev Kheldar_ (~Kheldar@ANantes-557-1-195-193.w2-8.abo.wanadoo.fr)
11:14.33IvnedarAnd if Defiants are from the future, why don't they have access to 3.0 souls yet? :P
11:31.45IvnedarHrrm, currency gadget could be tweaked to track FES/week couldn't it?
11:32.03Ivnedarif you can track when weekly reset happens easily
12:38.16ZorbaGamanoshei, no custom frame events yet. it'll happen eventually - it's actually a lot easier with the new event system than it was with the old
12:48.52KheldarAny way to display sub-frames like the rift frames easily ? i mean for exemple frames like the listing of needed material for a craft
16:41.47*** join/#riftuidev seebs_ (~seebs@home.seebs.net)
16:41.47*** mode/#riftuidev [+o seebs_] by ChanServ
17:45.04*** join/#riftuidev Noshei (~quassel@c-24-9-171-99.hsd1.co.comcast.net)
18:21.10Nosheidang it
18:22.01NosheiI hate it when I do write something up get it working and then have to come back and completely rewrite it because something else I did later doesn't work with the way I did the first thing
19:07.55Adeleahaha Noshei, yes indeed.
19:08.13AdeleaThat and copy and paste, and you forget to update 1 character in the paste =/
19:09.18Nosheioh I do that all the time
19:09.34Adeleaso do i, so youd think i would learn,  but oh no
19:09.46Nosheior I will copy a frame and its config then only change the name on the line where I create the frame, dont that plenty
19:10.15Adeleajust spent an hour and half trying to work out why something wasnt working. then finally thought to check my source.. and instead of x=data.x, z=data.z, i had x=data.x, z=data.x
19:11.53Nosheilol
19:12.19Adeleaid copied my data from somewhere else, so was sure it was good, so no need to check =/
19:12.45Nosheiya know, making a scrollbar from scratch is a pain in the ass
19:13.21Adeleawith a slider ppl can drag up and down ?
19:13.32Nosheiyeah
19:14.15Adeleaheh, ive done vertical scroll bar like things, but more as a 'here you are in the list', but always stayed away from making it grabable
19:14.30NosheiI'm working on drag support right now.  I can drag the scrollbar so far, but now I need it to also update the frame position.
19:14.36seebs_I just had a really awful idea. A lua-preprocessor meta-addon.
19:14.47seebs_Which basically loads hunks of data somehow and then does string processing on them before loading them.
19:17.27Adeleawould make localization simpler
19:19.22seebs_I had been thinking of wanting to be able to do things like, say, generically prefix whole hunks of code with "local foo = bar" type stuff so I don't have to duplicate that code in so many files.
19:21.36Adelea#ifdef would be nice
19:25.29NosheiGAAAAHHHH!
19:25.42Nosheimy brain is not working!!!
19:25.48seebs_I fear C's macro stuff. But I really want a way to clone a hunk of local definitions.
19:30.26Ivnedaryawns
19:30.27Ivnedarmorning!
19:32.55Adeleaif you say so!
19:34.33Ivnedar5:30am is too early to be at work on a saturday :(
19:34.48Adeleas/early.*/early/
19:34.55Ivnedaryeah, tell me about it
19:35.12Ivnedarespecially since I normally start 10am weekdays :P
19:35.45Ivnedarpretty sure I'm only awake because body's pumping adrenaline trying to work out what's going on
19:36.12Adelea"its the end of the word - engage flight or fight!"
19:36.18Adelea*world
19:38.33Ivnedarnow to twiddle thumbs for 20 mins before I can bring the world down!
19:40.48Adeleacurrent least favourite thing in the whole world : intermittent non-reproducable bugs
19:43.06IvnedarMine is having a 'bug' fixed by a vendor that had been accepted as how things worked and used a lot
19:43.39Adeleaoh dear
19:44.56IvnedarTrying to sync passwords between two organisations that have differing password policies is 'fun' :)
20:02.37Ivnedarwasn't me!
20:06.30*** join/#riftuidev ChanServ (ChanServ@services.)
20:06.30*** mode/#riftuidev [+o ChanServ] by orwell.freenode.net
20:38.22Nosheigah I'm stuck!
20:40.21NosheiSo I'm making a custom scrollbar and I'm trying to get the display area to scroll correctly when the use drags the scrollbar around.
20:41.26NosheiSo For the display I have a maks and a frame, the frame being bigger than the mask.  For the scrollbar I have the bar and a background frame.
20:42.41NosheiSo when the user grabs and drags the scrollbar it will change position by changing the point it attaches to background at using an ordered pair
20:43.43NosheiI am also doing the same thing for the display area, the frame will change where is it attached to the mask by use of an ordered pair.
20:44.34Nosheithe problem I am having is that when the user drags and drags the scrollbar I can easily figure out where to move the scroll bar to, but I can't seem to figure out where to move the display frame to.
20:45.13Adeleathere is a 1:1 correlation between the scrollbar and display frame?
20:45.27Adeleain as much as full left on scroll = full left on display, and same for right?
20:45.34NosheiThe issues here are really 2 that I see, on the scrollbar I am moving the smaller of the 2 frames.  On the display I am moving the bigger of the 2 frames.  Also I am moving the display frame the oposite direction of the scrollbar.
20:45.48Nosheiup/down scroll
20:47.13Adeleasurely you just need to work out the fraction up/down the scroll bar is, and then set display frame to the same fraction, or 1-it if youre going in reverse?
20:48.22Nosheithat is the obvious solution, but sadly isn't the correct solution.
20:48.39Adeleahow so?
20:48.48Nosheithe reason for that is because of the first issue I noted.
20:50.42NosheiI've got the current code up at http://rift.curseforge.com/addons/vendkp/files/3-r20130712204704/ if anyone wants to take a look at it
20:51.13Nosheithe revelant code is in vKPUtility.lua and vKPSettingsTab.lua
20:53.04Adeleaok, addon installed - what do i need to do to see what y ou are talking about?
20:53.36Adeleaok go to Settings i guess?
20:54.00NosheiI don't have it implemented yet, as far as the display area updating really.
20:54.12Nosheistrill just trying to figure out how to go about it.
20:54.38Baananooffset of frame (TOP frame relative to TOP mask) = -offset of handle (TOP handle relative to TOP scrollbar) * (height of frame - height of mask) / (height of scrollbar - height of handle) ?
20:54.45Adeleawhat is it you want to happen? the vertical scroll on settings to be smooth ?
20:55.44NosheiBaanano: I think that is what I need
21:02.21Nosheidang it, not quite right, ends up being slightly greater than it should be
21:02.55Baananonote hHandle = hMask * hScrollbar / hFrame
21:05.06NosheiI went with hHandle = 1-(hMask/hFrame)
21:18.47Nosheioh nevermind that was somehting different
21:19.07NosheiI am doing the height of the handle the way you said
21:38.36Nosheiany ideas Baanano it is close, but just a bit off
21:44.54*** join/#riftuidev Cairenn (~Owner@MMOI/Administratrix/Cairenn)
21:44.54*** mode/#riftuidev [+o Cairenn] by ChanServ
21:50.31Nosheiah nope, your bit is right, I'm not moving the scrollbar to the right place it seems.
21:57.34NosheiI think maybe I might be running into a rounding error, but 4% seems pretty high for a rounding error
22:01.03*** join/#riftuidev Kheldar (~Kheldar@ANantes-557-1-195-193.w2-8.abo.wanadoo.fr)
22:01.53Nosheiyeah I'm pretty sure at this point that it is a rounding error somewhere, because everything looks good.
22:31.53NosheiGAH
22:34.30Nosheieither the display is moving to far or the handle isn't moving far enough on each change, but one of them isn't moving hte correct amount.  Like if I move the scrollbar to 25% of the heigth the display is 1.7% further down than it should be.
22:34.54seebshmm
22:35.09seebs*thinks*
22:35.19Nosheiall of the math appears to be correct as well
22:35.27seebsOkay, this one's easy to get wrong and very prone to fenceposts.
22:35.52seebsUnderlying problem: The handle goes from top-aligned-with-top to bottom-aligned-with-bottom.
22:36.31seebsAnd the same thing is probably true of whatever's being scrolled, except that sometimes people intentionally allow scrolling so that at least some of the window is showing things outside the range of the thing scrolled, such as when editing.
22:37.21seebsBut basically, the center of the handle should be exactly centered at a 50% scroll, but half a handle-height low at 0% and half a handle-height high at 100%, or so.
22:39.55NosheiI'm actually positioning everything based solely off the top of the frames
22:40.26Nosheiwell it is percent based where I attach to
22:40.45Nosheihere take a look if you want http://rift.curseforge.com/addons/vendkp/files/4-r20130712223805/
22:41.33Nosheithe revelant code is in vKPUtility.lua lines 39-75
22:41.53Nosheieverything is displayed on the settings tab
22:42.27Nosheiand this is specifically when you drag the scrollbar, not when you use the mousewheel.
22:43.47seebsohh?
22:43.59seebsSo if you use the mousewheel, it ends up where you expect?
22:44.16Nosheiyeah, but that is handled by a different function
22:44.21seebsHmm.
22:44.38Nosheicause the mouse wheel is easy, since I define how far to move
22:44.43seebsThat sounds oddly familiar.
22:45.10Nosheiwhere as with dragging the scrollbar I have no control over how far it is moved.
22:45.15seebsHeyyyy.
22:45.17seebsThere's a thought.
22:46.18seebsDoes what happens vary with the way you move? Like, if you wiggle the mouse a bit, or move it slower or faster?
22:46.29Nosheinope
22:46.52seebsSo the "line" is the vertical line, and the "box" is the scrollbar handle?
22:47.05seebsOh. Hmm.
22:47.14Nosheiyeah, the box is actually the orange outline frame
22:48.01seebsDon't have game up, so I'm just looking at code. Hmm.
22:48.13Nosheiah ok
22:48.27seebsOkay, you are storing the location in the box as Scrollbar.Mouse.y.
22:48.39seebsWait.
22:48.52seebsvKP.Tab.Settings.Display.Scrollbar.Mouse.y = ((mouse.y - top)/vKP.Tab.Settings.Display.Scrollbar.line:GetHeight())
22:48.58Nosheiso that value is how far the mouse is from the top of the box, relative to the length of the line
22:49.24seebsHmm.
22:49.37Nosheiso when the mouse moves I subtract that value from the value of how far the mouse if from the top of the line based on the top of the line
22:49.59Nosheithats what happens on like 48 in vKPUtility.lua
22:50.35seebsOkay. So that's trying to figure out where on the line you want the corresponding part of the box (the one you were previously over) to be.
22:50.55Nosheiyes
22:51.03seebsI am pretty sure that, in that case, Mouse.y should be showing up again somewhere.
22:51.18seebsBecause you are now going to set the box's new location, and where you set it has to be modified in some way by Mouse.y.
22:51.34Nosheiyeah it does, movePercent = mouseLinePercent - vKP.Tab.Settings.Display.Scrollbar.Mouse.y
22:52.04Nosheiand movePercent is the y value used in the SetPoint function
22:52.17NosheimouseLinePercent = ((ycoord - lineTop)/lineHeight)
22:52.21seebsNo, I mean somewhere after that. Okay, say for the sake of argument the handle is 20% of the height of the line, and you click on the very bottom of it.
22:52.36seebsSo Mouse.y should now be 0.2, I think.
22:52.47Nosheiyes
22:52.50seebsNow you move the mouse so the mouse is precisely over the middle of the line.
22:52.57seebsMovePercent would be 0.3, right?
22:53.05Nosheiyes
22:53.34seebsAhh, I see. And then you move the *top* of the box to 0.3.
22:53.47seebsSo it should be the corresponding place.
22:53.47Nosheiyeah
22:54.15seebsHmm.
22:54.39Nosheiexactly
22:55.01seebsOkay. So the box should be where you want it. Now let's see.
22:55.08seebsvKP.Tab.Settings.Display.Body:SetPoint("TOPLEFT", vKP.Tab.Settings.Dis
22:55.09seebsplay, 0, movePercentdisplay)
22:55.27seebsSo you're setting the body to an offset of movePercentDisplay.
22:55.36Nosheiyes
22:55.37seebsocal movePercentdisplay = -movePercent * (bodyHeight - maskHeight)/(lineHeight - boxHeight)
22:56.19seebsSo in our example, that's -0.3 * (bodyHeight - maskHeight), which is the amount of scrolling room we have, divided by (lineHeight - boxHeight).
22:56.28seebsI'm not sure I understand that lineHeight - boxHeight part.
22:57.26seebsAssuming maskHeight is 100 and bodyHeight is 500 (so the handle should be 0.2x the height of the line), we then have lineHeight = 100, boxHeight = 20.
22:57.29NosheiboxHeight = lineHeight*(maskHeight/bodyHeight)
22:58.09seebsSo lineHeight - boxHeight, for a 100-pixel tall mask and line, and 20-pixel tall box, would be 80.
22:58.23seebsSo it's -0.3 * (500 - 100) / (100 - 20)
22:58.30seebs= 0.3 * 400 / 80
22:58.37seebs= 0.3 * 5
22:58.52Noshei-1.5
22:58.53seebsEr -0.3 * 5
22:59.43Nosheiwhich since the body is 5 times the mask the offset difference should 5 times greater for the body than for the box
22:59.53seebsSo we're trying to set things to be 30% of the way in, and we then move the displayed thing to be -1.5 display heights in.
23:00.12seebsSo that's 150 pixels, which is indeed 30% of the total height of the displayed thing, I think.
23:00.28Nosheiit is
23:00.45seebsBut I think that's not quite what you want. Because what you really probably want is 120 pixels, which is 30% of the 400-pixel travel range.
23:01.08seebsAt movePercent = 100, you want the top to be up by 400, not 500.
23:01.28Nosheitrue
23:02.36seebsSo I think movePercentdisplay should just be -movePercent * (bodyHeight - MaskHeight) / maskHeight. So it'd be -1.2.
23:03.31seebsBasically, I think you're double-correcting for the fact that the handle (and window contents) don't move by the whole range of the line (or contents).
23:04.02seebsSo the larger the handle is, the larger the apparent error is likely to be, also the larger movePercent is, until it reaches the end and you end up pegged at the tail end of the window.
23:08.05Nosheistrangly that doesn't work
23:09.46Nosheiactually I tried  -movePercent * (bodyHeight - MaskHeight) / linekHeight.  Which works out to -1.2 above, but doesn't work
23:10.17seebsHmm.
23:10.38seebsOccurs to me.
23:10.45seebsI don't know what the units are.
23:11.38seebsLike, if I do a:SetPoint("TOPLEFT", b, 0.5, 0.5), I don't know whether that puts the top left corner of a in the middle of b, or puts the top left corner of A at an offset to B equal to half of A's size. I assume it's the former.
23:12.54seebsHmm. .Display is the mask, so it's not that they're different heights. So if the display's height is 100, SetPoint("TOPLEFT", .Display, -1.5, 0) should yield a -150 px offset.
23:13.12seebsThis is the sort of thing where I'd probably use a lua interpreter in game to run specific :SetPoint operations to see what they produce.
23:13.41Nosheian ordered pair is always relative to the frame that it is in reference to for SetPoint
23:14.38Nosheiso if you had frame:SetPoint(a,b,mask,x,y) then a,b is in reference to frame and x,y is in reference to mask
23:18.39Nosheiwhat I don't get is that by doing  -movePercent * (bodyHeight - MaskHeight) / linekHeight it only moves the doby about half the distance that it should
23:21.02Nosheithoughts seebs?
23:24.16seebsWell, I guess the first thing I'd do is: Do some of this manually.
23:25.03seebsSo, say. Make the variables visible in a global namespace, and try stuff like /script body:SetPoint("TOPLEFT", display, 0, 0), verify that it looks like it's at the top.
23:25.19seebsThen do (…, display, 0, -0.5) and verify that it is half-a-frame up.
23:27.09Nosheiyeah I've done that and it positions to where it should be.
23:27.29NosheiI've added in some colored frames so it is easy to see it move
23:27.49seebsHuh, that's… really odd.
23:29.08seebsAt that point, I can't figure what else can be up. If you aren't getting the same behavior from SetPoint in this handler that you are when running it directly, I'm not sure what to make of that.
23:33.20*** join/#riftuidev Fisker (~ballmer@2a01:4f8:d12:1104:d1e:d1e:d1e:d1e)
23:33.23Nosheihttp://i.imgur.com/whwkLyL.png
23:33.31Nosheihere you can see what it is going
23:34.11Nosheiso I positioned the first 3 based on the location of the body relative ot the mask
23:34.43Noshei0,0 then 0,-0.5 then 0,-1
23:36.08Nosheithe 4th one is like 2 px before you hit the bottom of where the box can scroll to.
23:40.16seebsI don't understand what I'm seeing. What are red, orange, purple, and green boxes?
23:40.36seebsOh, that's the content of the body window.
23:40.45Nosheitheir height is 25% of the body height
23:40.47Nosheiyes
23:40.53Nosheiso i can see it scrolling
23:41.02seebsThat looks right to me?
23:41.22Nosheilook at the position of the scrollbar handle in the third one
23:41.47seebsHmm. It looks really close to the bottom, and what we see appears to be really close to the bottom, I guess?
23:41.47Nosheithe bottom of the box should be even with the bottom of the line
23:42.02seebsHmm.
23:42.13seebsWell, the middle one looks really close to exact.
23:42.24seebsThe 50% one seems to be basically exactly lined up.
23:42.30Nosheiit is about 1.7% off of where it should be
23:42.43seebsThe 99% one, though, does look like it's a little higher than I expect.
23:42.57seebsHmm.
23:43.06Nosheiyeah and that is where the problem is
23:43.08seebsWell, hang on.
23:43.20seebsWhat are lineHeight and maskHeight here?
23:43.32Nosheiugly numbers
23:43.43seebsIt might be you want /maskHeight, not /lineHeight, because it looks to me like the line is about 3% shorter than the mask.
23:44.12seebsWhich might fit with a 1.5% error at 50%, and a 3% error at 100%.
23:44.25Nosheiline = 378.999969   mask = 390.999969
23:44.58Nosheithat is when I use -movePercent * (bodyHeight - maskHeight)/(lineHeight - boxHeight)
23:45.09seebsIn that case, I'd try
23:45.18seebs-movePercent * (bodyHeight - maskHeight) / maskHeight
23:45.26Nosheiif I use -movePercent * (bodyHeight - maskHeight)/lineHeight it only scrolls about half of the distance it should do
23:45.31seebsBecause what I suggested previously was -lineHeight
23:46.09seebs*thinks* Okay, with that setup, where the contents are 2x the height of the mask, you should expect to see a movePercentdisplay of -1 when you are at the bottom, I think.
23:46.15seebsAnd -0.5 when you are halfway through.
23:46.31Nosheiyes
23:46.59Nosheiso when the scrollbar is at 50% with -movePercent * (bodyHeight - maskHeight)/maskHeight I'm only at -.25
23:47.09seebs*thinks*
23:47.15seebsWell, *thinks*
23:47.50seebsbodyHeight - maskHeight should be about 780 - 390, I think. Because bodyHeight is about twice maskHeight.
23:48.01Nosheiand -movePercent * (bodyHeight - maskHeight)/lineHeight gives about the same results, but 1.5% off at 50% and 3% off at 100%
23:48.16Nosheiyes
23:48.17seebsSo it should be very close to -movePercent * (390) / 390
23:48.23seebsWhich should be very close to -movePercent
23:48.31Baananowhy don't you create two extra invisible frames? handleRail that would go from scrollbarTop to scrollbarBottom - handleHeight, and frameRail that would go from maskBottom - bodyHeight to maskTop, then you only have to keep the invariant %offset(body relative to bodyRail) = 1 - %offset(handle relative to handleRail)
23:48.35seebsSo it shouldn't be coming out to -.25 at from a movePercent of .5
23:48.45seebsUnless some other calculation in here is totally wrong.
23:53.41NosheiEUREKA!!!!!!!!!!!!!!!!!!!!!!!!!!
23:53.57Noshei-movePercent * (bodyHeight/maskHeight)
23:55.32Nosheibecause movePercent is already relative to the height of the box and line and the size of the box is relative to the difference between the body and mask
23:59.27*** join/#riftuidev Fisker (~ballmer@2a01:4f8:d12:1104:d1e:d1e:d1e:d1e)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.