or is it just "by eye"
I hope les get's ahold of some litterature..
or well name's of papers containing the algo we use
actually it might be better working on whatever paul had so far, but this one is complete and should be working :)
did paul commit it to CVS?
hey storms over for a while
ah, very cool
heavy precip...even knocked out sat tv
I'm image googling 'celtic' for some ideas =)
I have some celtic spiral 3-d borders I drew for signs
I'll give you some when you are ready
graphics format or gcode?
you can see them on the web site
g code if you want
you saw em right?
I saw some on your site, I was looking to get a library goin and maybe sketch out a few of mine own as well.
I did mine in autocad 2000
Ah, I have Corel Draw then can export to whatever.
I was looking for other "borders" as well. like scrollwork, etc
mine are all designed for a 90 degree v tool
vines and the like
oh, I forgot about the cutter =)
I've always admired scollwork of various kinds, jewlery, sings, sketches, etc
when does your machine come back?
especially on guns and gunstocks
suppoe to be Aug 3rd
I SHOULD be building the vacuum table, but it's WAY too hot the last few weeks
I am all plumbed but don't use it
turkey calls are too small for vac holding
you dont route THRU the material either, do you?
yes I do
really? too small?
details break vac...so the planks are cam clamped
then you have bridges in the cutouts?
did you make the cam clamps?
worker just whacks it with a hammer and they pop out
yes made the clamps
"Worker WACK A PART"
a splintered edge is left
wack a mole is more fun thoguh =)
but that goes to a machine that takes it right off
you have any pics of those cam clamps?
I just need ideas
I don't think I have pics
but not hard to devise
ok, no biggy. I'll google =)
hang on jon has some emails concerning the tp problem
let me read them
I am going to write a synopsis of what we did today and post it
Ask Jon to run some plots at various speeds and see if the blips are identical.
increasing the resolution as needed.
I couldn't get the resolution of plots he had at all, but thats due to operator ignorance
ok sent a mail
so far we have conflicting results
need to find a pattern
Anders sees the same glitch at any speed
Jon does not
Ok, are the blips EXACTLY the same?
but has not graphed low speeds
anders blips were the same at any speed
we need to figure that out
When I saw those blips, they slopped down gradually, then ramped up at almost 50% the rate
in use it very clearly gets less bad at low hobby speeds
but this is just numbers. Just math. No reason to get worse at higher speeds as long as the machine can do it.
This, and only this, is what keeps emc from being a high quality high performance controller
I'm thinking it's hitting a condition, reducing down, then a "oh shit" facotr and ramping back up as fast as it can.
after finding that sometimes the blip faces the opposite direction, I'm convinced that this is just a bug, and not necessarily a fundamental problem in the planner or algorithm.
could be cris
results seem conflicting so far
anonimasu put in some parentheses that had no effect, and reported that the blip reversed. So he's seeing it both directions "at random" too.
I saw the initial blip on yours reversed
In the valley, the blip is positive, on the peaks it's negative. But the ramp down is consistant across the timeline.
I know I saw it both directions on different runs of the same program
Jymmm: no, that's not always the case
les: the initial jog to get to the beginning point in the program was probably different.
cradek ok, maybe now the pos/neg part, but what about the gradual rampdown, then the 'oh shit' factor?
Jymmm: I never got enough detail to see the shape like JE did
I remember watching the blip with Fred standing there a couple years ago.
I wish I could reproduce his plot
cradek see the red? it's the cmdVel, real heavy on the 'oh shit' part
les: this will be easiest to find (for me at least) in the colinear case
He muttered about the tp just being a temporary quick plug in with plans of putting in a proper one
les: I think talking about arcs is a red herring
arc is not the issue
it's diring a transition that something is fubared
people just see it more with round cuts because the machined bump is more visible
hey les, i've been reading the list
nasty tp huh
I might be able to go do a couple tests for you tomorrow
we really really have to fix this matt
my buddy has an emc knee mill
and it seems that with those plots you're closer than ever
We tried for a better planner like segmentqueue but it was not to be
I keep feeling like tp/tc isn't fundementally bad, but that it's a simple problem somewhere
But I think now if the original cubic sub interpolated trapezoidal actually worked we would have a pretty good system
what should i run on a knee mill?
what could we do to find the original paper/book that was used to write it?
chris, are we running all the same thing?
I really suspect I could fix it in an afternoon if I had that.
les You know the math, Can you think of someplace/somewhere where someone that tried implementing it an area that's known for common mistakes?
les: I'm using the cradek_stable branch of emc1
I wrote that down somewhere - fred proctor & les both used it in school
Perhaps we should change to co linear blend moves?
matt it is not arc specific
the knee mill has some random cvs version, but i can change it if required
right - line blending too...
does Matt, or anyone for that matter, really need a machine. I thought the errors in the commanded velocity could be seen on a simulated machine.
i've never plotted anyting like jon did - we have a doc on that?
well net to plot axis velocity vs time
mshaver: it's just on the tkemc menu, very easy
I guess it is all on thegui
ok, i think i've seen it as a menu choice
Ever feel liek you're having one of these kinda days? http://www.guru-international.com/gray_monk/Photos/2004-02-18--Prison_Escape.jpg
how will we fix it?
sure got quiet... I guess I'll print out tc.c & tp.c to see if it makes any sense to me
right now i've got to go over to another guy's shop & look at his yaskawa mx1 control - comm problems
* Jymmm lights the tumbleweed rolling by on fire and grabs the marshmellows!
mshaver is now known as mshaver_away
any of the glitch busters still around this evening?
a series of linear moves starting at x0 to x0.25 to x0.75 to x1.75 to x3.75
makes a very interesting plot.... at f60, the glitch after the first segment is about 75% decreasing as longer segments are blended.
I dont have the plot on a web site....
get er done!
[04:37:20] <LawrenceG> http://members.shaw.ca/cncstuff/glitch.png
the top and bottom red lines should be flat if the linear moves are blended properly
whats going on Jymmm?
Nada, just reading manual after manual, etc
Ya know, I'd like to know how Jon created his graph. It's really detailed.
Pimp MY Ride!
his was from a real machine with feedback
a 'real' machine?
I am running sim on a laptop
tuning of axis may somehow contribute..... his positions seem to track very well,,,, it would be nice to know what program segment he ran to capture that... in my screen capture, I includes the program editor window
I have no clue, just have to wait till he pops in I guess
This is a great show
blending seems to be looking too far back in time when it tries to match velocities
You saw Jon's right? http://jelinux.pico-systems.com/images/circleglitch.png
see ya later... wife just got home....
Phydbleep must have died in New Orleans
Nope.. Been running my ass off for a while.. Remodeling the house, fixing the car, building a 60Hz mechanical oscillator for a fluid dymanics experiment...
Would you beleive a high precision vibrator?
I could comment, but I better not =)
I am going to make the tp open a file and write out all velocity velocity values.
and move my routine around until I see where it miscalcs.
tp joy :D
* anonimasu is toying around with tc.c
but well, no luck yet
I have no idea what a TC is ;)
going to write a routine so I can check what velocity gets calculted..
Discriminate-based trajectory planning
and what is a TC
I know the concept
that's the trouble
we have no papers
it's determinates accel..
but TP says: trajectory planning based on TC elements
seen the plot?
[08:01:14] <anonimasu> http://jelinux.pico-systems.com/images/circleglitch.png
I found a variable where I could tweak it down
* alex_joni yawns
but it boils down to a accel miscalc somwhere earlier.. I think
I can do that in photoshop
it should be a bit nicer..
running with that
but the problem seens to be that the velocities gets calced all wrong..
I'll add debug stuff so we can see it happen..
need to go to a customer
[08:19:10] <anonimasu> http://www.bojn.net/~an0n/tweak2.jpg
Thoose images are with a slight sourcecode mod.
anoniamsu is now known as anonimasu_
stop talking to my logger :P
you'll distract him
or well I havent gone to the customer yet
same here ;)
so I am all free to play for a bit ;)
whatcha playin with?
I am going to write a log routine this afternoon
so I can save every accel calc as it's running..
or well move it around so I can get the _steps_ of the calculation
I should be working with some UI design but I dont have the laptop set up yet since I am going to go to a customer..
how's things going for you?
_Poincare is now known as Poincare
alex_joni is now known as alex_joni_away
Time: 06:22:15 -0500 GMT, Windows XP Professional, Service Pack 2 (5.1 - 2600), AMD XP 2800+ running at (1-AMD , 2124MHz, 512KB (38% Load)), DDR400 RAM Usage: 572/1024MB (55.86%), C: 37.77gb of 75.42gb free, D: 0gb of 0.36gb free, N: 0.38gb of 372.62gb free, Current Uptime: 1wk 4days 15hrs 22mins 5secs, Record Uptime: 3wks 12hrs 48mins 50secs, 3 Samsung 19" flat monitors (1 Trinitron, 2 CRT).
alex_joni_away is now known as alex_joni
yo an0n, are you there?
wot news ?
it's too damn hot
can't think straight
Paul, if you read the lists you will see we have really
done some discovery on tp/tc misblending
[10:32] <anonimasu> http://www.bojn.net/~an0n/tweak2.jpg
[10:32] <anonimasu> ;)
[10:32] <anonimasu> Thoose images are with a slight sourcecode mod.
oh I just posted a little bit of the algo on the user list
tp is indeed screwed - But we knew this from the spiral tests.
yes, but this set of observations might shed some insight
especially with graphs like this:
[13:41:51] <les> http://jelinux.pico-systems.com/images/circleglitch.png
had a scope on the DAC outputs.... See a spike in velocity at each segment.
that is on simple blend of long arcs.
The dv/dt appears to be consistant with accel.
yes I think it happens on all blends
yup - And the best bit.....
regardless of G61/G61.1/G64, the spikes remain the same.
yes....at high speeds anyway
I can only see a dif between g61 and g64 at very low speeds
At first I thought this was a G64 mode flag problem
I think we have proved that it is not
it's just a horrific math error
Now we need to see if it is related to velocity and or accel clamping
because we know velocity adaptation is screwed
I am beginning to think that if the cubic subinterpolated trap planner just worked properly we would have a fairly good control
Full cubic or quintic splining would be much better....
ugh... You really don't want to see the plot with a lower accel.
But as is at much above sherline hobby speeds the tp is (sorry to say)...garbage.
Gets worse with lower MAX_ ACCELERATION?
Get that Anders?
I think he's not around.. or?
So...implies accel clamping problem?
increasing accel, get a small dip in velocity in one quadrant
If accel is set very high will it go away? ( check only in simulation)
and a huge increased spike in another quadrant.
that is a clue.
A clamp is missing ?
I got some spare clamps ;)
what size do you need :D
a big one I think...
accel clamping is not the root cause in my opinion....
but it is a term that uses max accel?
I put up part of the trap algo on a list message just to show that it is simple math. I copied it right from my book.
Why Why is tp/tc such a jumble?
why is the whole code like that ;)
ok.. going home
The exact values of velocity & accel are irrelevant - I do not think it is a clamping calc that is at the root of the problem.
That the traces look the same regardless of blend mode would suggest it is an error in the blend algo.
I can see a difference at some speeds between g61 and g64
but often the glitch in g64 is enough to be an exact stop...or look like it
I can see and measure differences in path between G61 & G64, but the velocity discontinuity remains
path is ok
but the tangiential velocity thing trashes chip load and causes burning, bumps, and divots
a cnc machine that can't control feedrate is of little use even if it can follow a path
Agree that it is not a g61/g64 flag mis set?
Suggestion: Ring Matt and/or Ray, give them some ear bashing.
Was preparing to call Fred....but is that gonna help?
If you are calling Fred, the problem is the point at where decel at the end of a segment is flagged.
the decel is being flagged some 200 servo cycles later than it should be....
and the TC queue just happens to be 210 deep....
I would need to log the data in detail, but initial measurements seam to suggest 200 cycles.
but isn't tc queue at traj rate rather than servo?
actually I would think it is asyncronous...just a list of positions and times...transformed gcode really
I had been running a 1:1 servo:traj
so just to understand...decel is flagged in what should be cruise...causing a premature blend?
nope - It is delaying the blend
I would summise that decel is being flagged well in to the decel phase.
the math is so simple for that...could it be a scaling thing?
There is one point in particular in tc where decel flag is suppressed
the comments there do not make sense to me
You need to review the commit logs from revision 1.1 onwards.
way back machine huh
Sourceforge have the logs.
this problem appears old...very old
I will do that if I can in doze
the cvs logs are all html, so even IE can read them.
it is so hard to understand Fred's comments....
things like "combined velocity of this move and the next"
WTH doe that mean?
wash and brush up time
back in a while
Jacky^ is now known as Jacky^cnc
Eeeeesh, my stomache is talking smack already!
Jacky^cnc is now known as Jacky^
I bouhgt a wood table 20x300 cm, thickness 3 cm, cut to a thickness of 1/2 and it has been bent :\
how did you cut it?
i didnt cut it, I have made to cut from the carpenter
it has taken the shape of an arc
Dinner?! it's only 1030
here 19:30 ..
it is not! =)
after cutted, the wood begin to move.. bah :(
if it was a table, why cut it?
because i did not found a table of 1 cm
just 3 cm
1cm is NOT a table - too thin
i was thinking 1 cm for some relief was good...
also simple to machining..
well you said it was cut in half, can you plane it?
from one of 3 cm he has gained 2 of 1 cm
it has said that the thickness of the blade consumed 5 milimeter
then smoothing ..
I learned a new thing :)
when bought wood tables better not to divide them why they are folded
and going to the lumber store wouldn't have been easier?
Can't find it here ..
there is some big store
comes down ot understanding the material one is working with - and being smarter than that material
RonB you lie!
any material that is split - will curve a bit
wood is far more sensitive thatn asy ... titanium
but - don't blame the material
blame RonB instead!
also, 20 cm width is too much ..
fine - I'm used to dealing with Ass_oles
Jacky^ isn't an asshole
RonB a lil frustrated today by chance?
not at all - just finished another 6 hours of downloading and got another bad MD5 sum
ok, so a LOT frustrated.
why should i be frustrateD? - I think this is the 14th time
have your tried bittorrent?
oh it's 19:42 here
anonimasu WAY past your bedtime
Jymmm: I need to play with the tp a bit..
*grabs the hammer*
anonimasu grab teh debugger instead
Jymmm: why dont you debug it?
anonimasu I'm not the one looking for the casue of the problem, you are!
anonimasu and do you REALLY want me submitting to cvs? I dont' think so
Jymmm: somone else can code a fix, if you know where the error is ;)
anonimasu yeah right, I can't even get the plot to work
err logging I mean
what does it do?
nothing at all (really), click the START button and it does nothing.
Besides, I don't do CVS
actually I dont do anything.
except grab hammers
fuck... do I grab a NAS box for $70 w/o HDD or not?!
do you need a NAS box?
It be very nice, but a ext usb/fw would do me more good I think
I wish this 2nd screen would be good..
* anonimasu sighs
my mill-screen is doubling as dualhead for the normal comp
or well good for it's purpose
i hate rebates
but for being coded on 20 hours a day i'd bleed from my eyes
thats meep in Finnish huh?
M I C... K E Y M O U S Eeeeeeeeeeeeeeeeeeeeeeeee
if you dont understand it already dont bother trying to decipher it..
it's my feeling about the tp
my honest one.
* anonimasu is having a hammer
the digging through the tp is not going well?
just started for today
adding some debug routines to write the newVel to a file..
so I can follow the calculations..
if you start off be refactoring any routine longer than a page, you'd be in with a chance
well, mot of a chnace then
as there's no desc of how the planner works..
if its not obvious from the code, and its not documented, I'd call that bad coding style
robin_sz: have you had a look at it?
sadly, it has now come to bite us in the arse
some while ago ...
but you know my opions on that already
yeah they help as much as the advice to reduce it down..
welll ... the advice to reduce it down is good advice
if I knew how I'd rewrite it..
you dont need to know how it works to refactor it
rafctoring doenst chnage the way anything works, it just moves stuff about and makes it easier to read
being able to read it may then help you understand how it works
nope, as there's no desc on how it does what it's supposed to do :D
or what it's supposed to do..
but to NOT understand how the thing works, then to try and rewrite one section is probably how it got to be how it is now in the first place.
who said anything about rewriting anything?
did he, oh must have been before i got here
I'll write something to catch the output
so I can move around my tpLOG(x);
robin_sz : [11:01:21] <anonimasu> if I knew how I'd rewrite it..
[11:01:28] <robin_sz> eh?
[11:01:41] <robin_sz> you dont need to know how it works to refactor it
and watch what happens during the calculations..
robin_sz PAY ATTENTION! =)
Jymmm: are you aware of the difference between re-writing and refactoring?
robin_sz: I played a bit by fiddling with the code.. here's how it ended
robin_sz anonimasu> if I knew how I'd --> rewrite <-- it..
[18:51:50] <anonimasu> http://www.bojn.net/~an0n/tweak2.jpg
not that I know if it's any better in reality..
but it's less of a blip ;)
Jymmm: answer the question
anonimasu, yeah, that look lots better
robin_sz : <robin_sz> who said anything about rewriting anything?
[11:03:34] <Jymmm> anonimasu did
[11:03:45] <robin_sz> did he, oh must have been before i got here
anonimasu, esp on the upper end of the curve .. gone
Jymmm: are you going to answer the question or just keep pasting?
robin_sz: upper end?
the bottom is between the 2 circles..
where the blending occurs
robin_sz : theres no point to your question
anonimasu, right, it was top and bottom yesterday IIRC
Jymmm: are you or are you not aware of the difference between refactoring and re-writing?
robin_sz I really dont care.
but it's better then http://jelinux.pico-systems.com/images/circleglitch.png
fsck yeah ;)
les: can you forward that message with the tp stuff to me?
anonimasu why is it so fast between 0 and 1?
no clue, it's just the scale I guess..
ahh, yeah of course thats velocity
I wonder if that modification does any better when actually cutting stuff..
so thats accel
have you plotted dV/dt?
is it finite during the blip?
in 20 sec ;)
and d2V/dt2 would be cute too
dV/dt is accel
dt is time..
d2v/dt2 is jerk
I cant plot accel
I suspect accel is stupidly high at that point and jerk may well be unbounded
wait a sec
the trajectory points are all right :) atleast
err yeah I can plot accel..
this is matchcad or something?
looked a bit windows to me :)
acceleration magnitude goes ^
off the scale?
want a screenshot?
good morning all === http://members.shaw.ca/cncstuff/glitch.png
=== for linear move in x glitches
although that's with my mod robin_sz..
I should revert my tp.c to a orginal one ;)
and re-do the plot
anonimasu imagine what that would do for steppers?
has anybody seen my missing steps?
* anonimasu imagines
I dont know about lost steps..
limiting accel to the motors is key to avoiding lost steps
setting max accel down makes glitches worse....
yes much worse according to paul
can you re-mail me that mail you wrote with calcs
I am working on a debug func to the tp..
so I can print stuff to a file and move it around ,)
so we can follow the calc.. what it does where..
les: did you have a look at my modification of it?
it is only part of it....but will forward
was bending and lasering some aerospace stuff today
without my stuff.
the magnitude goes to 300 when you start motion
we have to hold 0.2mm tolerances on all bends
and 0.1mm on all holes
fair enough, but this is just a cable support .. some of the holes are just so you can get a screwdriver in to bolt the thing to the ceiling
robin_sz: look at this ;)
in use, someone just cableties some wires to it ... theres no need for such precision
* robin_sz looks
are you getting treatment for that?
www.bojn.net/~an0n/sc.jpg & www.bojn.net/~an0n/sc1.jpg
sc is the first with the mod i made.. just for testing
sc1 is the other one with the stock tc..
I had to make it a less high res in the plot to get it to skip the initial points..
sc1 is actually better is it not?
since they had a amplitude of 300..
look at the * accel for y
the blip was bigger for v, but hte accels where constrained
perfection im afraid is the only answer
it should do neither..
but it's a bit better ;)
it needs to be perfect eventually ...
* anonimasu sighs
I figure that needs "deep understanding"
ofcourse it should.
but as there is no living person that has any reference about it it's kind of hard.
well, scrap it and start again :)
sometimes its a better thing to do
like with dogs
sometimes its better to just start over with a new one
put the old one to sleep
and get a cat instead :)
les: that math is useful
les: with the whole page it would be even better
I am working on it now
the whole thing
drawing charts in autocad now
I'll have the logging done in maybe 5 minutes
please name the variables..
It will be right from the book
can't just scan the book...copyrights
I need caffeine
hm, I have logging in a sec ;)
hm, wont work that easily.. need to put stuff in a buffer and grab it later..
* anonimasu stuffs crap into memory
les : The book wouldn't be over 75 years old would it?
Italy arrests 'fourth UK bomber'
thanks to scotland yard
anyone here ?
say, has any of you ever seen hand/vise brakes like this before ? top level: http://www.dacmachines.be/productplooibankhobbyned.htm
hi dpy , never seen..
those look handy
anyone here have experience with logging & plotting
mshaver is now known as mshaver_away
mshaver_away: I have been using it, but it recently (last ten minutes) stopped working for me!
oh I bet I know why
I changed TP_STRUCT but only rebuilt some of the system
eh, shit happens
anyone has tried rubber varnish to paint wood ?
what's the best way to obtain a smooth finiture ?
I am working on the tp math page illustration
way too much work
did you find the reference used for this implementation, or something else?
I have to use my (and Fred's) old college textbook
25 years old
my edition is anyway
do you think that's what FMP used in 1997?
He used a "discriminate" version with help from delta tau
I guess I don't know what that means
It seems to be a computationally efficient form I guess
but harder to understand
I can only do what is in the book TP chapter
efficient is good
clever + undocumented is bad
funny this 25 year old text has quintic interpolation
Senior engineering/first year grad course in robotics
there's some stuff in here to limit d/dt(accel), but it's disabled
if I/we get the basic system fixed, maybe we can try that (again?)
well in general it is going to have unlimited jerk
well, there's code that limits it
Fix the basic system and it will be a pretty good planner
not state of the art, but not garbage either
Quintic is the lowest order that has complete control over jerk mafnitude
I think the bogons were introduced in version 1.3
I tried to get the commitlog on a doze box but could not
[21:58:47] <cradek> http://cvs.sourceforge.net/viewcvs.py/emc/emc/src/emcmot/tp.c?r1=1.2&r2=1.3
eh? it should work fine
if not, get firefox, like you should anyway
oh the diffs..yeah that I can get
matt and i did have a long phone talk this morning
He is willing to help too
I can only help with math.
who is WPS?
william shackford...he works for fred
Jacky^ rubber what?
les: I think he broke it :-)
I think he means contact cement jymmm
cradek: yeah could be
Wait a sec.... Fred works at NIST and doens't have access to any additional docs?
he did some key changes there
Fred wrote emc (mostly)
Jymmm: is an ancient way to paint wood, from french I think ..
Jacky^ rubber varish?
the french polish.
Jymmm: yeah, in italian is Gomma lacca ..
hmmm, never heard of such a thing
oh might be dammar varnish
it is passed to pad
no.. translate Lacca
french polish for sure
google...much information on it
also alcool is used ..
we call it "spiriting off"
pad is used with alcohol
main resin is dewaxed shellac
do you it's good for paint 3D relief ?
# Rubber varnish : it is an other alternative to the cracking varnish, always in sale in flakes in the storees of Fine Arts, and goes melted in the alcool.
It is often considered the finest wood finish in the word
english --> http://translate.google.com/translate?hl=en&sl=it&u=http://www.artekjara.it/manuale/decorazione/craquelet.shtml&prev=/search%3Fq%3DGomma%2Blacca%26num%3D100%26hl%3Den%26lr%3D%26safe%3Doff%26sa%3DG
it is slow and difficult to do
Italian --> http://www.artekjara.it/manuale/decorazione/craquelet.shtml
the pad is lubricated with oil
it is laborious, bu should be nice
at the end..
Jymmm: yeah, i've seen
but never used first ..
the craquelet is something else
look on here: the jewelry box near the bottom of the page is an example of french polish
let see ..
[22:12:55] <les> http://www.lmwatts.com/gallery.html
seem very nice :D
I will buy it tomorrow ..
the hard thing will be to learn how to get a good job ..
Jymmm: gommalacca: looks like this: http://www.il-legno.it/restauro/gommalacca.html#
Yeah... practice till you throw out your elbow! lol
on a 3d relief i don't know how easy is to pass it :(
I will try
$99 and a $30 rebate, I'm tempted --> http://www.adstech.com/products/NAS-806-EF/intro/NAS_806_intro.asp?pid=NAS-806-EF
I'm using videolan to share TV or video on lan/internet
oh you got it working huh?
I also got speed dsl connection now :)
are you streming content?
I tried with my girlfriend, she's in naples now
oh home porn!
better use encryption for that =)
audio work fine, video lost some frame
but she have 640 dsl connection, not fast as the mine
ah.. motion is another funny application :)
[22:33:13] <Jacky^> http://www.lavrsen.dk/twiki/bin/view/Motion/WebHome
les : You know PostScript by chance?
heh, dont we all =)
not just med's
Sorry, I dont know anyone named Gueeedo
but i know guido... (cor)
7 axis beburring tool??
wanna ROCK & ROLL and get $$$$$ for it??
assuming you meant DEburring
plastic bead blasting sounds easier
NOT shotpeening... yet... more $$$ for all of us..
Jymmm? how's the machine now?
no have to remove kelllering
A-L-P-H-A let me check UPS tracking...
A-L-P-H-A I won't get it back till about Aug 3rd
dynfyle belts.. on an air qualified pressurre plate.. (nmi)
for what A
im lookin for backup landing zones....and machine desighn for these guys is $$$ from heaven
it'll be a retro on a robotic arm ...
dmess? :) need help? I need $$$ :)
any takers?? long weekend at the Marawanna Lodge to do the R & D iff all looks akording to plan ,, (which is flexible... on all parts)
for what A
i am NOT a parrot
dmess want a cracker?
how many $$$
cheeseburger, cheeseburger, chips, pepsi, no coke
dmess, I don't know... i just need work... I'm thinking of working even at ikea.
white stff ifn they got
A-L-P-H-A you're too young to know what the line means.
1. pays $11/hr (piddly compared to my $50K a year). 2. The girls there are hot, and my age.
ROCK & ROLL
fly or die
im hopin to get milege this weekend
waiting on a call about a new parachute.... wishfull thinkin'
but if ya got it it adds to the fuzzy feeelin'
rule 1. never get out of a plane if the wings are still attached, chute or no chute
robin_sz's Rule 2. Never experiance ALL that life has to offer.
oh .. you'd be suprised what I;ve done ;)
and experiencing ALL that life has to offer isnt always such a great plan.
for example ... being gang-raped by a Brazillian baseball team is probably not something to experience if you can help it
blech about half done with this TP web page
had about enough
on the wiki?
have not published it yet
just put it in the wiki will ya?
will need a link on the wiki yeah
I hope it is helpful....it's a real pain
any good reason not to put the content in the wiki?
well it's html
can I do that?
text and tables?
pictures and formatted text
formatting can be a bit odd, but yeah, you can mostly do that .. just put the page up and see if someone elese gets the urge to copy it over maybe
nice to keep it all in one place if we can ...
I make all my engineering book reports web pages these days
LOL @ robin_sz, yeah right
wiki is like a stack of punch card after a hurricane
I am trying to make the math and description simplified and easy to read
that would be a change for emc stuff huh?
almost unheard of
well we will see how we do
pretty simple mathreally
I wont be doing more work for today on the tp
to be fair the interp is nicely doc'd
I am at a friends place
going to slack
the interp is great
the tp is garbage
the rest ... well .. jons HAL stuff is nice
I'll dedicate tomorrow afternoon to the tp..
I will bang away at this page
but I need to eat food and drink water (well 95% water) or i'll die
s/water/ice cold beer/
gonna go eat out now