**** BEGIN LOGGING AT Wed Oct 26 02:59:57 2011
Oct 26 03:09:30 <ka6sox> rwhitby, did that "work"?
Oct 26 03:57:22 <rwhitby> ka6sox: repair added to git and tested once
Oct 26 03:57:45 <rwhitby> I made it remember which partitions needed repairing, so you could still do stuff on others and come back to it
Oct 26 04:04:43 <rwhitby> ka6sox: I also now allow you to recheck a checked filesystem
Oct 26 04:09:11 <rwhitby> and media resize seems to work
Oct 26 04:19:14 <rwhitby> ka6sox: releasing that as 0.0.7
Oct 26 04:19:23 <rwhitby> unfortunately, no feedback from the resizefat
Oct 26 04:30:25 <ka6sox> bummers...still
Oct 26 04:30:33 <ka6sox> is it quitting the same way?
Oct 26 04:30:49 <rwhitby> no, it's no longer quitting
Oct 26 04:31:08 <rwhitby> all the percentage messages are coming as a bunch at the end - you can see it with ls-monitor
Oct 26 04:32:05 <ka6sox> okay, so now I should see what is terminating the individual strings
Oct 26 04:32:23 <ka6sox> and not being passed up till the end.
Oct 26 04:45:05 <ka6sox> rwhitby, you aren't going to like it even if you are successful in getting it to passing % up.
Oct 26 04:45:33 <ka6sox> because it appears to "hang" at 100% for a while, status is dwait and then it "completes"
Oct 26 04:45:44 <rwhitby> that's fine
Oct 26 04:46:18 <ka6sox> okay its not writing over the same line, so not just CR
Oct 26 04:46:42 <ka6sox> btw I am using resizefat -v
Oct 26 04:47:03 <ka6sox> but next I'll just do plain resizefat and see what it says (or not)
Oct 26 04:48:09 <ka6sox> btw...it does a fsck first.
Oct 26 04:52:45 <ka6sox> rwhitby, http://pastebin.com/cBrWtaB9
Oct 26 04:52:48 <ka6sox> our "options"
Oct 26 04:55:17 <ka6sox> so we could use this tool to repair too.
Oct 26 04:59:06 <rwhitby> I already do :)
Oct 26 04:59:32 <ka6sox> kewl
Oct 26 05:07:48 <ka6sox> okay all lines have EOL.
Oct 26 05:07:56 <ka6sox> no different
Oct 26 05:09:41 <ka6sox> so apparently it waits for a status change to dump it?
Oct 26 05:09:55 <ka6sox> ie watches std err?
Oct 26 05:41:10 <ka6sox> rwhitby, looks like I can fix errors in /media/internal.
Oct 26 05:41:18 <ka6sox> with 0.0.7
Oct 26 05:41:28 <rwhitby> oh, you caused some?
Oct 26 05:41:38 <ka6sox> yup
Oct 26 05:41:50 <rwhitby> any suggestions on the flow?
Oct 26 05:42:18 <ka6sox> if it errors on test, then offer to repair.
Oct 26 05:42:42 <rwhitby> yep, that's the current flow
Oct 26 05:42:50 <ka6sox> yup
Oct 26 05:42:51 <ka6sox> thats it
Oct 26 05:43:23 <ka6sox> then take you back to previous step
Oct 26 05:43:32 <ka6sox> however....
Oct 26 05:43:43 <rwhitby> looks like resize value is quantized, probably by cluster size or something
Oct 26 05:43:54 <rwhitby> so you don't always get what you ask for
Oct 26 05:43:57 <ka6sox> the rule of thumb is that you want a clean run *after* a fix.
Oct 26 05:44:46 <rwhitby> yep, it sets you up to do a check after a successful repair
Oct 26 05:45:59 <ka6sox> why did I miss that...(meaning I just went with it.
Oct 26 05:47:47 <ka6sox> he he
Oct 26 05:51:45 <ka6sox> let me introduce som  more errors.
Oct 26 06:00:10 <ka6sox> that was interesting
Oct 26 06:00:37 <ka6sox> I created an error the first time
Oct 26 06:00:58 <ka6sox> the second time storaged called fsck.vfat before remounting it.
Oct 26 06:01:40 <ka6sox> so, that means if you can get ugly with it,then it will fsck.
Oct 26 06:06:04 <ka6sox> rwhitby, I can't get it to do it again...I mean, I could reduce the size of the partition below the size of the FS and do it that way...
Oct 26 06:06:09 <ka6sox> (manually)
Oct 26 06:06:40 <ka6sox> but that should make tailor abort.
Oct 26 06:08:35 <ka6sox> okay first time was a fluke
Oct 26 06:08:43 <ka6sox> I confused storaged.
Oct 26 06:10:14 <rwhitby> adding ext3fs resize
Oct 26 06:12:42 <ka6sox> kk
Oct 26 06:21:14 <ka6sox> driving north
Oct 26 16:59:09 <elpollodiablo1>  /join #webos-internals
Oct 26 23:22:53 <elpollodiablo1> dtzWill: you around
Oct 26 23:24:58 <dtzWill> elpollodiablo1: yep :)
Oct 26 23:25:49 <elpollodiablo1> you have some time to talk through some thing about Xserver multi touch controls?
Oct 26 23:26:39 <dtzWill> sure, but don't have access to my normal webos dev stuff atm--but i can certainly chat. that work? :)
Oct 26 23:26:52 <elpollodiablo1> I have the idea all figured out just cant quite implement it
Oct 26 23:27:03 <elpollodiablo1> yeah that is fine, more of a brain picking conversation
Oct 26 23:27:31 <elpollodiablo1> so sdl have event.XXXXX.which
Oct 26 23:27:34 <dtzWill> sure thing, and good because X11 input is halted on lack of good ideas and people championing them :)
Oct 26 23:28:10 <elpollodiablo1> not sure if you area aware of that, but it tells you which event it came from, so one finger .which =0 2nd finger .which =1 and so on
Oct 26 23:28:45 <dtzWill> yep yep, although i'd forgotten which field it was exactly
Oct 26 23:29:46 <elpollodiablo1> ok good, so I tried to say if .which = 0 then make fingerDown[0]=1
Oct 26 23:30:16 <elpollodiablo1> basically making a bool array in C with 0 and 1 to test in an if statement
Oct 26 23:30:59 <elpollodiablo1> first off is that an acceptable approach for keeping track of which fingers are down?
Oct 26 23:31:50 <dtzWill> an array of bools? reasonable enough, sure. just make sure don't index unchecked into it with 'which', but other than that seems fine
Oct 26 23:32:18 <elpollodiablo1> why dont index it with which?
Oct 26 23:32:39 <dtzWill> fingerDown[event.XXXX.which]
Oct 26 23:32:48 <elpollodiablo1> yeah that is what I was trying
Oct 26 23:32:52 <dtzWill> anyway, i'm just saying make sure no array out of bounds s'all :)
Oct 26 23:33:20 <elpollodiablo1> oh alright, the array was giving me some issues to begin with
Oct 26 23:33:23 <elpollodiablo1> but anyway
Oct 26 23:34:29 <elpollodiablo1> so when I get the .which=1 (2nd finger) I set finger down, but it acts as if right click is still down
Oct 26 23:34:55 <elpollodiablo1> which only causes isses for on the desktop most apps take the right click input
Oct 26 23:36:06 <elpollodiablo1> so I was thinking virtually unclicking the first finger(left mouse button) but setting mouseState &= ~buttonState
Oct 26 23:36:26 <elpollodiablo1> where buttonState equals a left mouse click
Oct 26 23:36:40 <elpollodiablo1> but =by
Oct 26 23:37:15 <elpollodiablo1> so it basically removes the left click bits from mouseState
Oct 26 23:38:55 <elpollodiablo1> I am not very familiar with how bitwise opperators work. so programatically should that work?
Oct 26 23:42:51 <dtzWill> well so lets separate programmatic implementation and your ideas first
Oct 26 23:43:08 <dtzWill> the big question to answer is: what should happen when?
Oct 26 23:44:08 <dtzWill> as events we have what fingers are 'down' when, and we also have the events of the fingers being lifted
Oct 26 23:45:00 <dtzWill> for example it might make sense to not map the first finger to left button at all--but rather, fire a "left mouse down, left mouse up" in succession when the user puts a finger down and immediately (within say 500ms) without doing anything else
Oct 26 23:45:52 <dtzWill> leaving the ability for right-click to be 'one finger down, followed by another finger down, followed by both fingers being lifted", maybe
Oct 26 23:45:53 <elpollodiablo1> ok
Oct 26 23:45:59 <dtzWill> idk what's best; taht's why i haven't done this yet xD
Oct 26 23:46:12 <elpollodiablo1> lol it can be done multiple ways
Oct 26 23:46:36 <dtzWill> but things we'd want to support are left click, double click (left), drag (left) and right-click
Oct 26 23:46:43 <dtzWill> right-click drag would be nice, but not mandatory
Oct 26 23:46:56 <dtzWill> any usable scheme that gets those without spurious events sounds good to me.
Oct 26 23:47:38 <dtzWill> personally i'd be inclined to just look at whatever splashtop or similar does, as a)they probably put more thought into it than I have or care to, and b)consistency for users is an automatic point boost
Oct 26 23:47:45 <rwhitby> yep
Oct 26 23:48:14 <dtzWill> once we've sorted out what makes sense, implementing it something to worry about.  while it's not always the case for this i'm not worried about designing somehting that's impossible to implement
Oct 26 23:48:14 <elpollodiablo1> I can get all those working but I get one or two unwanted events mainly with leftclick still occuring while trying to right click.
Oct 26 23:48:26 <elpollodiablo1> I was trying to follow scummVM
Oct 26 23:48:37 <dtzWill> elpollodiablo1: oh, okay. is what they do documented somewhere?
Oct 26 23:48:51 <elpollodiablo1> I am looking at the source
Oct 26 23:49:03 <elpollodiablo1> does that count?
Oct 26 23:49:41 <dtzWill> if it's well-written, yep! :)
Oct 26 23:50:18 <elpollodiablo1> it has pretty good comments in it
Oct 26 23:50:21 <dtzWill> but so talk to me about how this input model works
Oct 26 23:50:37 <dtzWill> and i'll help you implement it, or identify where your existing implementation differs from your ideas
Oct 26 23:50:51 <elpollodiablo1> ok
Oct 26 23:52:00 <elpollodiablo1> so 1 finger down is left click. double taping one finger is double left click, one finger down and drag is left click and drag.
Oct 26 23:52:10 <elpollodiablo1> how it is working now
Oct 26 23:52:19 <dtzWill> (i'll be in/out with other things, but I *will* ready any/all your write and respond for sure, just something's come up that needs at least some of my attention)
Oct 26 23:52:23 * dtzWill listens
Oct 26 23:52:50 <elpollodiablo1> that is fine, I will keep typing away
Oct 26 23:53:35 <elpollodiablo1> so right click is with one finger aready down putting a second finger down
Oct 26 23:53:48 <dtzWill> elpollodiablo1: it might help if we modeled things in terms of "mouse button 0 down" regarding what we send the x server
Oct 26 23:54:18 <dtzWill> since that modeling avoids the issue i see with what you just said: there's no way to do a right-click without left-clicking
Oct 26 23:54:47 <elpollodiablo1> ok
Oct 26 23:54:49 <dtzWill> sinec as you described it, one finger down is left mouse down, and one finger up is left mouse up, and 2nd finger down is right mouse down
Oct 26 23:55:11 <dtzWill> so there's no good way to have 2nd finger down without going through the "left mouse down" state, which i'd imagine is the problem you're seeing
Oct 26 23:55:28 <elpollodiablo1> yes that is the problem I am seeing
Oct 26 23:55:59 <dtzWill> so the task is to first fix it conceptually before worrying about arrays and bitwise whatcha doodads :)
Oct 26 23:58:45 <elpollodiablo1> lol
Oct 26 23:59:29 <dtzWill> how does scummvm do it? do they require right-clicks to generate left-clicks first?
Oct 26 23:59:52 <elpollodiablo1> is that backwards?
Oct 27 00:00:25 <dtzWill> it's poorly said, probably :)
Oct 27 00:00:38 <dtzWill> do they let you right-click without left-clicking first?
Oct 27 00:00:42 <elpollodiablo1> they virtually unclick the left click when the right click in enabled
Oct 27 00:01:54 <dtzWill> i'm not sure i understand what you're saying.
Oct 27 00:02:13 <dtzWill> but regardless, it seems like it'd be nice to be able to right-click without having to left-click first :)
Oct 27 00:02:42 <elpollodiablo1> actually it seems like the have a proccess mouse events function that they call when they need it
Oct 27 00:02:53 <elpollodiablo1> yeah I agree
Oct 27 00:03:11 <dtzWill> what seems to make sense, if a bit messy technically, is something like:
Oct 27 00:03:38 <dtzWill> single-finger tap--left 'click'. this means the x server does nothing until the user lifts the finger quickly, in which case it generates a left mouse down and left mouse up quickly to simluate a 'click'
Oct 27 00:03:53 <dtzWill> two finger tap --same thing for right-click
Oct 27 00:04:11 <elpollodiablo1> ok
Oct 27 00:04:36 <dtzWill> if a finger is down for > "X" amount of time or moves more than "Y" pixels, left mouse down, and enter "drag" mode internally
Oct 27 00:04:52 <dtzWill> while in drag mode, if you lift that finger it generates a left mouse up, and go back to default state
Oct 27 00:05:05 <dtzWill> there are all the edges cases like, what if you press a second finger while in drag mode
Oct 27 00:05:16 <dtzWill> and thinking a bit about whether all of that is actually usable or not
Oct 27 00:05:29 <dtzWill> (as for second finger in 'drag' mode, i'd say just ignore it)
Oct 27 00:05:46 <dtzWill> double-clcik is then just tapping twice, and "falls out" as a result of the single-click description
Oct 27 00:05:55 <elpollodiablo1> well my idea was to have a two finger down and up swipe to being the vkb up and down
Oct 27 00:06:02 <dtzWill> i'm not mandating this is the case, but it is an example of a model that offhand sounds like it would work and meet the goals mentioned above
Oct 27 00:06:31 <elpollodiablo1> I understand
Oct 27 00:06:35 <dtzWill> okay so if you have two fingers and don't release (see the "tap" metrics defined above re: X and Y), consider it a vkb toggle
Oct 27 00:06:46 <dtzWill> that capture what you were getting at?
Oct 27 00:07:09 <elpollodiablo1> yeeah
Oct 27 00:10:47 <dtzWill> well then that sounds good to me, but does get a bit messy to implement
Oct 27 00:11:06 <elpollodiablo1> I have a few things to try
Oct 27 00:11:38 <elpollodiablo1> so I will mess around with those and see how they turn out then go from there
Oct 27 00:37:16 <dtzWill> okay. let me know how it goes :)
Oct 27 00:37:22 <dtzWill> and of course, good luck! ^.^
Oct 27 00:44:43 <elpollodiablo1> thanks, what are your thoughts on a press and hold for right click?
**** ENDING LOGGING AT Thu Oct 27 02:59:58 2011