#robotics Logs

Jan 12 2019

06:15 AM Tom_itx is now known as Tom_L
09:00 AM rue_mohr: hahahah
09:01 AM rue_mohr: my code that tests where two line segments touch and returns NULL if they dont is faster than my code for testing IF they touch
09:01 AM rue_mohr: Intersection locate: 10M/1.936 sec
09:01 AM rue_mohr: Intersection test: 10M/2.011 sec
09:01 AM rue_mohr: hahahahahah
09:05 AM polprog: lol
09:08 AM rue_mohr: it converts them both to general form, resolves the intersection, then checks if that point is on both of the lines...
09:09 AM rue_mohr: which is faster than checking the dot and cross products, then seeing if the points of one are on the other
09:09 AM rue_mohr: I wonder how they fair if the lines are on each other
09:10 AM rue_mohr: ah, the faster one throws it out if they are parallel
09:38 AM rue_mohr: printf("%s\n",geom2dIsLineTouchingLine(l2d(pt(0,0),pt(1,1)), l2d(pt(0,1),pt(1,2)))!=0? "Collision <-- bad":"No collision <-- good" );
09:38 AM rue_mohr: my code is failing on that test
09:41 AM polprog: My eyes are failing on that code
09:42 AM rue_mohr: pt returns a point, l2d returns a 2d line, the function determines the string the conditional returns
09:43 AM rue_mohr: printf("%s\n", foo()==value?"true":"false");
09:44 AM rue_mohr: #define colinearIsBetween(P,L) (vectDotPoint(L.p1, P, L.p2)>=0?1:0)
09:44 AM rue_mohr: hmmm
09:45 AM rue_mohr: so there is a bug in my alg.
09:46 AM rue_mohr: that happens at 180 degrees
09:46 AM rue_mohr: the bug might be in the way I'm looking at the problem tho
09:46 AM rue_mohr: two lines can have 3 states
09:47 AM rue_mohr: not touching, crossing, or overlapping
09:47 AM rue_mohr: no...
09:47 AM rue_mohr: non-intersecting, intersecting, or overlapping
09:48 AM rue_mohr: :S the sun is comming up, thats not good
10:16 AM SpeedEvil: rue_mohr: better than it not coming up.
10:17 AM rue_mohr: ------- 24 colinearIsBetween ------
10:17 AM rue_mohr: line (0,0)-(1,1): point(-0.000, -0.000) False
10:17 AM rue_mohr: line (0,0)-(1,1): point(0.100, 0.100) True
10:17 AM rue_mohr: line (0,0)-(1,1): point(1.000, 1.000) True
10:17 AM rue_mohr: line (0,0)-(1,1): point(1.100, 1.100) False
10:17 AM rue_mohr: iiiinteresting....
10:19 AM rue_mohr: for (i = 0; i < 30; i++, p.x+=0.1, p.y+=0.1) {
10:19 AM rue_mohr: printf("line (0,0)-(1,1): point(%s) %s\n",point2d2s(s, 40,p), stringTrueFalse(colinearIsBetween(p, l)));
10:19 AM rue_mohr: }
10:20 AM rue_mohr: so, colinearIdBetween, has a bug of some kind
10:20 AM rue_mohr: #define colinearIsBetween(P,L) (vectDotPoint(L.p1, P, L.p2)>=0?1:0)
10:20 AM rue_mohr: #define vectDotPoint(A,B,P) ( (P.x - B.x) * (B.x - A.x) + (P.y - B.y) * (B.y - A.y) )
10:25 AM rue_mohr: -0 != 0
10:25 AM rue_mohr: wtf
12:39 PM veverak: pche
12:39 PM veverak: lower-cast language
12:40 PM veverak: needs these inferior language macros!
12:40 PM rue_mohr: na, I mostly use macros to ditch stack io for simple junk
12:41 PM rue_mohr: push push push call... return pop pop pop
12:41 PM rue_mohr: dont know why everyone hates macros
12:41 PM veverak: hard to debug if something is wrong
12:42 PM veverak: and there are better alternatives in C++ for example
12:43 PM rue_mohr: what about C?
12:43 PM veverak: there are none
12:43 PM veverak: (hence the language sucks!)
12:43 PM veverak: anyway
12:43 PM veverak: </trolling>
12:43 PM veverak: :)
12:44 PM rue_mohr: no language sucks, just the people who cant code in it
12:45 PM rue_mohr: except java, java sucks.
12:45 PM rue_mohr: :)
12:46 PM veverak: agreed on java sucks
12:46 PM veverak: but, I don't agree that "no language sucks"
12:46 PM veverak: uncapabale languages sucks
12:47 PM veverak: anyway that is longer debate and I have to go afk to attent to my girlfriend :)
12:47 PM veverak: detach!
01:39 PM Mikeee: so I took one of the in-person 32 hour fanuc robot courses and damn was it a game changer
01:54 PM deshipu: what changed exactly, in practical terms?
01:55 PM deshipu: apart from taking 4 days out of your life?
02:00 PM Mikeee: Basically I can confidently use system variables now
02:09 PM deshipu: of what system?
02:10 PM Mikeee: well not system specific, just handling tool
02:10 PM Mikeee: software specific
02:10 PM deshipu: I still have absolutely zero idea what you are talking about
02:11 PM Mikeee: that's fine
02:12 PM deshipu: if you say so
02:33 PM veverak: no idea here too
04:36 PM deshipu: as long as he's happy with it
05:13 PM rue_mohr: Mikeee, cool!
05:14 PM rue_mohr: Mikeee, was it all programming and toolpaths?
05:46 PM polprog: You know guyd
05:46 PM polprog: guys*
05:46 PM polprog: System variables
05:47 PM polprog: Like... the variables... in the system!
06:20 PM Mikeee: rue_bed it was supposed to be but about half of everyone was already pretty well versed in it so we did higher level stuff like system variables, which is pretty awesome because there's data there you'd otherwise have to derive in your own program or use sensors for
06:23 PM Mikeee: plus it's possible to run assembly on the fanuc controllers, which means I can create my own functions to call whenever I want, and never have to worry about the time it takes to get the results
06:41 PM rue_mohr: huh
09:14 PM rue_shop1: 8-| the induction heater circuit just killed a 100V capacitor (the resonance cap)
09:16 PM rue_shop1: er no, no, that was the irfz44 rupturing
09:16 PM rue_shop1: which, I thknk happened first
09:16 PM rue_shop1: ..
09:18 PM rue_shop1: but it did oscillate
09:22 PM rue_shop1: and its a 200V cap
09:22 PM rue_shop1: on the scope, I saw the osc amp. go over...
09:22 PM rue_shop1: 40V p-p
09:23 PM rue_shop1: which might explain the fet failure, the irfz44 is only a .... 60V? fet
09:24 PM rue_shop1: yea, it might have been over what it could take
09:24 PM rue_shop1: but... also, the drive signal is just a diode off the coil, and thats not rated for more than 20V
09:42 PM rue_shop1: ok I dont know what to work on
09:43 PM rue_shop1: the town is working on seizing the boat my ma lives on, cause she wont pay the property taxes on it
09:43 PM rue_shop1: cant concentrate any more with that going on
09:43 PM rue_shop1: maybe if I can work out a homing system for a little robot
09:44 PM rue_shop1: somnething to focus on
10:36 PM rue_shop1: anyone alive?
10:36 PM rue_shop1: this sucks, I need someone to bounce ideas off of
10:36 PM rue_shop1: I'm still here ya know?
10:36 PM rue_shop1: its just not hte same, sorry
10:36 PM rue_shop1: hmm
10:37 PM rue_shop1: is there nobody else out there?
10:37 PM rue_shop1: no, just us
10:37 PM rue_shop1: what about the fpga people?
10:37 PM rue_shop1: they think your insane
10:37 PM rue_shop1: damnit.
10:37 PM rue_shop1: kinda ironic
10:37 PM rue_shop1: shutup
10:37 PM rue_shop1: ok
10:37 PM rue_shop1: no, wait, thats not actually what I want
10:38 PM rue_shop1: meh
10:38 PM rue_shop1: ok, work with me, we have a charging station
10:38 PM rue_shop1: yes, yo do
10:38 PM rue_shop1: damnit, I forgot, there IS a charging station
10:38 PM rue_shop1: yea
10:39 PM rue_shop1: we need to make a new ramp
10:39 PM rue_shop1: yea, time eh?
10:55 PM Tom_L: you haven't done any fpga stuff
10:56 PM Mikeee: FPGAs are the goddamn future
10:56 PM Mikeee: if only I had the time to learn literally anything about fpga
10:56 PM Tom_L: i did a little a long time ago
10:56 PM rue_shop1: hahah
10:56 PM rue_shop1: they been around forever
10:57 PM Tom_L: my cnc uses them
10:57 PM rue_shop1: but there is a HUGE group on twitter
10:57 PM Tom_L: pal
11:31 PM rue_shop1: veverak,
11:31 PM rue_shop1: ?
11:31 PM rue_shop1: oh wrong kicad version, arg
11:37 PM zhanx: encoders to read wheel postion
11:37 PM rue_shop1: yes! I have those too!
11:37 PM zhanx: oh the scroll was 9 hours behind
11:37 PM rue_shop1: but just tick counters, no direction
11:37 PM rue_shop1: I use it for tracking/correcting veer
11:39 PM rue_shop1: the remplacement motor on the one side didn't match as well as I'd liked it to have
11:39 PM rue_shop1: "smoking progress"