#hazzy Logs

May 24 2020

03:03 AM Mufdvr0978[m]: hi, I am trying to reset my feed rate override, I am using a mpg to override the feed rate and it is working well but I want to be able to use a pin to reset the override to 1. I found a post that suggest that it should be done via the ui but I can not find the option on probe_basic, here is the post I found https://forum.linuxcnc.org/49-basic-configuration/28974-hardware-button-to-reset-spindle-feed-override
03:03 AM Mufdvr0978[m]: Any suggestions ? thanks!
03:09 AM Joco[m]: Lcvette: you about? probe basic question.
03:15 AM Joco[m]: if there is not hal pin on the UI then have you looked at the halui.feed-override.value pin? Resetting that to 1 will get your feed override back to 100% (i.e. 1). The GUI should be representing what that pin is saying.
03:17 AM Joco[m]: actulally - sorry. That is a out pin not an IN pin
03:20 AM Mufdvr0978[m]: I have tried via halscope to reset etc but it does not work, if I reset the mpg count the feed goes to zero, the others won't let me change them.
03:22 AM Joco[m]: just do it via the halui.feed-overrides.counts
03:23 AM Mufdvr0978[m]: I traced the UI button action to machine.feed-rate-reset (from memory) but I can't see how to action it
03:23 AM Joco[m]: looking
03:23 AM Joco[m]: I don't use probe basic so have to snoop about a bit
03:24 AM Mufdvr0978[m]: I think I tried but it won't let me change, I will back in front of the PC in 30 min to try again
03:24 AM Mufdvr0978[m]: Thanks!
03:29 AM Joco[m]: ok the UI is using a linuxcnc binding to do that reset
03:30 AM Mufdvr0978[m]: Is that something I can access from linuxcnc?
03:30 AM Joco[m]: looking. At the moment there is no pin on the UI. And I'm looking to see if there is something in HAL that you can link to
03:45 AM Joco[m]: I can't see a way to do it that is consistent and not complex. I think if you need some of these as UI pins that you ask for it as a feature.
03:57 AM Mufdvr0978[m]: thank you, I will check that out.
03:59 AM Mufdvr0978[m]: do I request a feature using the issues logging function?
04:08 AM Not-944e: [02probe_basic] 07lmclaren7000 opened issue 03#13: feed rate override reset - 13https://git.io/JfaHJ
04:29 AM Joco[m]: that seems a reasonable way to do it.
05:49 AM JT-Cave: morning
05:53 AM Joco[m]: Hi ya
06:34 AM Joco[m]: sleeps
09:52 AM hazzy-m: lmclaren: hello! Thanks for the issue.
09:52 AM hazzy-m: I think you will have to reset the override counts. This should be easy enough to do from the UI though
09:53 AM hazzy-m: lmclaren: when you press the override reset in PB, thoes the displayed override value change?
09:54 AM hazzy-m: If not then we can add an override reset HAL pin to PB, and you can connect that in your posgui to reset the override
11:06 AM Lcvette1: morning
11:07 AM Lcvette1: i used the feedrate slider and override reset all the time on mymachine, works great
11:07 AM Lcvette1: what is the issue?
11:08 AM Lcvette1: lmclaren: what are you not having work?
11:08 AM Lcvette1: im confused
11:08 AM Lcvette1: is it something you customized?
11:09 AM * Lcvette1 uploaded an image: Screenshot_2020-05-24_12-00-15.png (52KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/MEISiezqHfWooxItnBnmICaQ >
11:09 AM Lcvette1: it is an action button and it needs to be set to machine.feed-override.reset
11:10 AM Lcvette1: as seen above in the QTDesigner snapshot
01:45 PM Joco[m]: I think the “issue” is a feature request.
01:46 PM Joco[m]: I think he wants to be able to hook a physical button to that function.
01:49 PM Joco[m]: I tried doing some if the rest directly in hal via hal show on the halui counts. Maybe its a quirk but I could not get consistency. I could get into the situation where counts 0 was 0% and after some fiddling would be 100%.
01:52 PM Joco[m]: More broadly I had a look at the version of PB I have down and there are no hal pins exposed. Is there a plan to expose some navigation, functions to hal so as to support specialist physical buttons?
01:53 PM Lcvette1: you can use ain
01:53 PM Lcvette1: aout
01:53 PM Lcvette1: dinaout
01:53 PM Lcvette1: * din dout
01:54 PM Lcvette1: M68 E# Q#
01:54 PM Joco[m]: What is that?
01:54 PM Lcvette1: thats what my atc uses
01:54 PM Lcvette1: for sensor inputs and outputs
01:55 PM Joco[m]: Oh i see.
01:55 PM Joco[m]: Motion digital in and out pins
01:55 PM Lcvette1: yes
01:55 PM TurBoss: Hello
01:55 PM Lcvette1: turboss lives!
01:56 PM Lcvette1: \o/
01:56 PM TurBoss: \ø/
01:56 PM Joco[m]: Not sure that helps though? How would a user use that to link into functions on the UI ?
01:58 PM Lcvette1: use an mdibutton
01:58 PM Lcvette1: ?
01:58 PM Lcvette1: im not sure i undrstan the question fully
01:59 PM Lcvette1: you need to have an external physical button change the state or activate a gui button?
01:59 PM Lcvette1: is that correct?
02:02 PM Joco[m]: Perhaps a better question to ask is this “as an end user who wants to create specialist control box with a number of on screen buttons represented as physical buttons, pots or such like in the control box, how would you go about that?”
02:02 PM Joco[m]: From what I can see at the moment you need to start hacking PB. Which many people are not comfortable with.
02:03 PM Lcvette1: motion.analog−in−NN IN FLOAT
02:03 PM Lcvette1: These pins are used by M66 Enn wait-for-input mode.
02:03 PM Lcvette1: ?
02:03 PM Lcvette1: something like this?
02:03 PM Lcvette1: http://linuxcnc.org/docs/html/man/man9/motion.9.html
02:05 PM Joco[m]: Yes I understand these pins. I use them in my auto tool offset measuring macro. But i do not think they answer the question.
02:06 PM Lcvette1: TurBoss: ? can a physical pin input trigger a gui button?
02:07 PM Joco[m]: Yes
02:07 PM Joco[m]: If said gui button exposes a hal pin
02:08 PM Lcvette1: i vaguely remember a comversation about a similar thing back in 2018 when we were starting
02:08 PM Lcvette1: it was to do with an external MPG triggering on screen increment buttons when switched at the mpg physical pin
02:08 PM Lcvette1: * it was to do with an external MPG triggering on screen increment buttons when switched at the mpg physical pin input
02:09 PM Lcvette1: so that both the mpg and onscreen increments selection would remain in sync
02:09 PM Joco[m]: Gmoccapy esposes a number of pins to allow physical controls to linked up. Feed overide reset is one of them.
02:09 PM Lcvette1: i don't remember if there was a resolution to that or not
02:09 PM Lcvette1: but i would think it is similar
02:09 PM Lcvette1: in fact i don't think there was because i don't notice m,y incrmeent display changing with the mpg input
02:13 PM Joco[m]: It may be that some are easy to do and some may be so problematic that you dont do them
02:13 PM Lcvette1: i think at the time we were so early in that it was a long term thing to look at that we never came back too
02:13 PM Lcvette1: this was way back at the very beginning
02:13 PM Lcvette1: lol
02:13 PM Lcvette1: soo im not sure
02:13 PM Lcvette1: maybe hazzy did some stuff on it im not aware of
02:15 PM Joco[m]: Well there are a set of hal widgets that deliberately expose hal pins. I use some of them in my gui.
02:15 PM Joco[m]: There is also functionality to be able to add hal pins to any gui element if you want to gwt your hands dirty.
02:16 PM Lcvette1: yeah i don't know hal
02:16 PM Lcvette1: i have very lmited linuxcnc experience
02:17 PM Lcvette1: in the 2 years i have been "using" it, i have been priomarily working on the gui itself the layout, the probe routines etc
02:17 PM Joco[m]: Hal is quite powerful. It can also be an evil of poo
02:17 PM Lcvette1: i know enough to make basic things work but only if i have an example to follow
02:18 PM Lcvette1: i still do not understand the root dynamics of how it all works and when i read the docs it just makes me drift off
02:18 PM Lcvette1: its like they are for those who already have an understanding of programming
02:18 PM Lcvette1: so im lost in it almost instantly
02:18 PM Joco[m]: Once you get into it its not too bad. But its more like designing for an electrical circuit than software.
02:19 PM Lcvette1: right but the syntax of eveything
02:19 PM Lcvette1: which is everything
02:19 PM Lcvette1: if you don't know it then your eff'd
02:21 PM Joco[m]: The syntax is simple. However some of the “rules” are not well explained. Im still not confident i have it all figured out. But my joystick hal is getting pretty complicated. So i must have some clues.
02:25 PM Lcvette1: lol
02:31 PM Joco[m]: Getting back to the original question, when i was talking to person who came in asking about PB pins. I think they represented the group of people who see the gui as a black box or magic. They look for prebuilt plugs on the gui so they can plug in the bits they want. They are used to being able to do this with the likes of gmoccapy. At the moment PB is a shiny black box of gui gorgeousnesses. But no plugs.
02:31 PM Joco[m]: * Getting back to the original question, when i was talking to person who came in asking about PB pins. I think they represented the group of people who see the gui as a black box of magic. They look for prebuilt plugs on the gui so they can plug in the bits they want. They are used to being able to do this with the likes of gmoccapy. At the moment PB is a shiny black box of gui gorgeousnesses. But no plugs.
02:33 PM Lcvette1: if i had my druthers on the matter i would have evenrything configurable from inside the gui so that no one ever needed to look at a hal or ini file and it would all be automatically built from settings entered in the gui
02:34 PM Joco[m]: I think some things should not need gui plugs.
02:34 PM Joco[m]: And gmoccapy was probably providing a crutch
02:36 PM Joco[m]: But there might be some things where exposing pins makes sense and is significantly easier and more consistent to have the gui front. Maybe we start a list of those. I would be happy to help implement.
02:37 PM Lcvette1: i would like stuff like that to be in settings
02:38 PM Lcvette1: so you can just enter the physical input on the controller number and have a drop down box for all the potential input buttons in the gui when you click enter it does the code and adjusts the hal
02:38 PM Lcvette1: :d
02:38 PM Lcvette1: * :D
02:39 PM Joco[m]: And it will wipe your nose as well.
02:40 PM Lcvette1: pretty much
02:40 PM Lcvette1: lol
02:43 PM Joco[m]: It would be possible.
02:44 PM Joco[m]: To a point
02:45 PM Joco[m]: We will need a way to dynamically add hal pins to gui elements.
02:46 PM Joco[m]: Or a significant re-engineering of converting many buttons and such like to hal widget versions.
02:48 PM Joco[m]: TurBoss: when you have a chance, could you explain to me why you felt ini plugin was a “no go”. Assuming I understand the situation correctly.
02:50 PM * Joco[m] uploaded an image: ima_e845785.png (365KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/vKYkwcEYCkBGgVXTxkBOqbTT >
02:52 PM Joco[m]: Lcvette: had to do own head model in end. Too many differences appearing in the grizzly model. But the internal components are awesome.
02:54 PM Lcvette1: top view of that
02:56 PM Lcvette1: fyi those top bolt holes are not uniform
02:56 PM Joco[m]: Will get in a bit. I still need to adjust for where the belt will run
02:56 PM Lcvette1: better measure each one or your parts won't fit
02:56 PM Lcvette1: need to shrink your pulley sizes
02:57 PM Joco[m]: Yeah. Front holes are 98mm seperation back two at 100mm
02:57 PM Lcvette1: the belt will slap the walls of the stand at certasin rpm from natural frequency
02:57 PM Lcvette1: i mean they are all different
02:57 PM Lcvette1: there is not uniform patter
02:57 PM Lcvette1: widths locations are all different
02:57 PM Lcvette1: no clue why but they are
02:58 PM Joco[m]: Seem to be uniform patterns on bf20
02:58 PM Lcvette1: i had to scrap the first set i made
02:58 PM Lcvette1: mine seemed to be until they didn't fit and i remeasured
02:58 PM Lcvette1: each one
02:59 PM Joco[m]: My plan is to make paper templates and check.
02:59 PM Joco[m]: Had thought of possibility i would screw up measurements
03:00 PM * Lcvette1 uploaded an image: image.png (141KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/igeGiPXYSFjMQybrKixBZkvz >
03:01 PM Joco[m]: Have checked distances front to back about 3 times now and get same numbers. Same with distance between each hole set left to right. Still going to check with template.
03:01 PM Lcvette1: just put a bolt in the hole and measure from outside head wall to bolt
03:01 PM Lcvette1: those are the ones that bite you
03:04 PM Joco[m]: Fuck. Mag 6 earthquake!
03:04 PM Joco[m]: Sorry. House just rocking and rolling
03:05 PM TurBoss: <Joco[m] "TurBoss: when you have a chance,"> editvcp would require info from the ini wich is not abailable yet there
03:06 PM TurBoss: I mean there is no ini on designer
03:06 PM Joco[m]: Ok. So designer knows about status?
03:07 PM TurBoss: I mean no ini is passed to designer
03:07 PM Joco[m]: Phew quake was deep. So nothing going to be flattened.
03:07 PM TurBoss: wow a earthquake?¿
03:08 PM Joco[m]: Yeah we get em. That had us all rushing to the door ways.
03:08 PM TurBoss: ini pluging would need to know about all posible sections and options
03:09 PM Joco[m]: Or use yaml to define what to pull across.
03:10 PM TurBoss: https://hub.pathpilot.com/login?next=%2F
03:12 PM TurBoss: could work
03:12 PM Joco[m]: Ok - will ponder things a bit. Need to think through what makes sense.
03:13 PM Joco[m]: Also we should sure we know what the use cases are for it. Ie what is the upside?
03:13 PM TurBoss: you asked about the pathpilot emulator the other day
03:13 PM Joco[m]: Cheers
03:53 PM TurBoss: hazzy: hello!, do you know how can I access widgets from the VCP main file?
03:53 PM * TurBoss sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/AekwcxPiFNHrkxPczmkPulzp >
03:55 PM Joco[m]: Think I have an example for that.
03:55 PM TurBoss: great!
03:58 PM Joco[m]: https://github.com/joco-nz/bf20_mill/blob/master/bf20_mill/mainwindow.py
03:59 PM Joco[m]: self.widget-name.method
03:59 PM TurBoss: Amazing
03:59 PM TurBoss: thanks
04:05 PM * TurBoss sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/caKQHCnrECtvNDLEpsyiaoaO >
04:06 PM TurBoss: and should be a better way to get the ini file
04:09 PM * TurBoss sent a long message: < https://matrix.org/_matrix/media/r0/download/matrix.org/ncFcyJuUCHswhLSLFjNoHMki >
04:13 PM TurBoss: Joco: Thank You! works like a charm!!
04:14 PM TurBoss: Lcvette: Joco did it the stack_widget based on ini options
04:14 PM TurBoss: \o/
04:14 PM Joco[m]: all comes down to what you are comfortable working in
04:15 PM Joco[m]: if you are ok with python then TurBoss approach is nice.
04:15 PM Joco[m]: If like Lcvette you are not comfortable with python having access to the data to be used in rules and/or the designer is what is needed
04:16 PM Joco[m]: There is a lot of ini sourced data in the status plugin. Because linuxcnc reads it then exposes that via the linuxcnc python interface
04:19 PM Joco[m]: TurBoss - on the axis front check what is supplied by status plugin. I think you could be making work for yourseld
04:19 PM Joco[m]: * TurBoss - on the axis front check what is supplied by status plugin. I think you could be making work for yourself
04:20 PM TurBoss: how?
04:20 PM TurBoss: :|
04:20 PM TurBoss: I don't get it
04:20 PM Joco[m]: status:axis_mask gives you a bit mask that allows you to asses exactly what axis are in place.
04:21 PM TurBoss: OH!
04:21 PM Joco[m]: you don't need to read ini file for that info
04:21 PM TurBoss: cool!
04:22 PM Joco[m]: http://www.qtpyvcp.com/components/status_items.html#axis-mask
04:23 PM Joco[m]: you know the pattern to access the status plugin data from python?
04:24 PM TurBoss: yup
04:24 PM Joco[m]: also with that info, and depending on what you are doing in the gui you could use a rule to drive what page in a stacked widget is shown
04:25 PM Joco[m]: depends on how clever your gui display is trying to be
04:25 PM Joco[m]: status:axes gives you an integer for the number of axis. e.g. xyz == 3, xyza=4, xyzb == 4, xyzab == 5
04:26 PM Joco[m]: not too sure how it would treat a gantry setup where you can have axis double up
04:27 PM TurBoss: its shown in binary representation 😆 XYZ 111
04:27 PM TurBoss: = 7
04:27 PM TurBoss: :P
04:27 PM TurBoss: XYZAB = 31
04:27 PM Joco[m]: that is axis_mask
04:27 PM Joco[m]: axes is just a count
04:27 PM TurBoss: oh
04:27 PM TurBoss: ok
04:28 PM Joco[m]: so if you are keeping it simple like in PB you can just get status:axes as the channel. then have a logic expression of ch[0]-3
04:28 PM Joco[m]: cose you are only dealing with xyz, xyza, xyzab
04:28 PM Joco[m]: page indexes: 0, 1, 2 respectively
04:29 PM TurBoss: yup
04:29 PM TurBoss: great!
04:35 PM TurBoss: works but there is something else driving the stcak widget
04:35 PM TurBoss: I'll head to other thing
04:36 PM TurBoss: Joco: do you need something?
04:36 PM TurBoss: :P
04:39 PM Joco[m]: me? What another project?
04:39 PM Joco[m]: work day now. yah. Already had first meeting of the day
05:28 PM roguish[m]: Lcvette: Hey, Mr. Macro King...... could you write me a probing macro? to map a bore ID. start in the center, go out to the surface, back to the center, rotate a set amount, and go out and back radially, and repeat all the way around.....
05:29 PM Joco[m]: on mill or lathe?
05:30 PM roguish[m]: variables: diameter, depth, number of points around.....
05:30 PM roguish[m]: I can help with some of the math. just don't have a clue about macros....
05:30 PM Joco[m]: Lcvette is the man on macros
05:30 PM Joco[m]: what outputs are you wanting?
05:31 PM Joco[m]: a point cloud?
05:31 PM Joco[m]: or just a more accurate centre?
05:31 PM roguish[m]: yes, point cloud. x,y coordinates. read into excel and see the circle
05:32 PM Joco[m]: if doing point cloud would you want to go back to the cenre all the time? Seems like a lot of wasted movement.
05:33 PM roguish[m]: maybe not. depends on the macro writer skill.
05:33 PM Joco[m]: bags not. lol
06:07 PM Joco[m]: TurBoss: Lcvette gents - have we started to tackle conversational milling i nPB yet?
06:07 PM Joco[m]: it's the next thing I am starting to think about for my wee BF20 gui.
06:08 PM TurBoss: mill_tpuch _v6 has some if i remember correctly
06:08 PM TurBoss: * mill_touch _v6 has some if i remember correctly
06:09 PM Joco[m]: is that the starting point ? Or is PB looking to do a fresh take on things?
06:10 PM TurBoss: no idea
06:20 PM Joco[m]: looking at things it reads like there is UI in place but I don't think anything behind it yet
06:34 PM Joco[m]: some good thinking in there.
06:34 PM Joco[m]: on the data points to be trapped in UI
09:40 PM Lcvette1: Joco: for pb i wanted to do a more graphical approach
09:44 PM Lcvette1: seemed more inline with the flow of the UI
09:44 PM Lcvette1: but its gonna be a big project i think
09:44 PM Lcvette1: i would like to complete lathe first
09:49 PM Joco[m]: Yup. I’m only starting to think about conversational stuff. Many cose some simple things I would like to build for my GUI.
10:10 PM Lcvette1: get lathe caugfht up to where mill is and then do conversational for both of them in tandem
10:10 PM Lcvette1: think it will be easier that way
10:10 PM Lcvette1: and faster
11:26 PM Not-944e: [02probe_basic] 07lmclaren7000 commented on issue 03#13: feed rate override reset - 13https://git.io/JfVn1
11:27 PM Joco[m]: Lcvette: let me know if I can help with PB. If you need some specific python glue for something I might be useful.
11:27 PM Joco[m]: Thinking on your lathe focus.
11:28 PM Lcvette1: i definitely will!
11:29 PM Joco[m]: just thinking that if you are able to spread the load out a bit then no one becomes s single point bottle neck
11:31 PM Lcvette1: for sure
11:34 PM Joco[m]: guess I need to fork a copy of PB so you can manage pull requests
11:34 PM Lcvette1: you should definitely ahve a copy of PB
11:34 PM Lcvette1: \o/
11:35 PM Lcvette1: and a widecreen
11:35 PM Lcvette1: :d
11:35 PM Lcvette1: * :D
11:35 PM Lcvette1: * and a widescreen
11:38 PM Joco[m]: got it
11:38 PM Joco[m]: oops - need to fork. Idiot
11:39 PM Not-944e: [02probe_basic] 07joco-nz forked the repository - 13https://git.io/JfVcv