Without precise details of what you are doing, it is impossible to judge.
It is well known that big gcode files will cause Axis to eat all memory
on small boards when doing a preview, but that is NOT a leak.
Whether that happens more quickly on Stretch, which may well be a
'heavier' system,
Can you please quantify the assertion that Stretch has a memory leak
with Axis.
I have been using Stretch since it was alpha status. Usually any memory
leak is with a program and will manifest irrespective of the distro.
On 8/11/2018 8:43 PM, Timothy March wrote:
I have a question for you
, 'schoone...@btinternet.com' via Machinekit wrote:
HI Charles
Yes Comp, plus headers, Makefile.inc and everything else in the -dev
package was moved into the main machinekit package some time ago.
There are only four machinekit-dev packages left in the repo and they
match the also deprecated wheezy a
HI Charles
Yes Comp, plus headers, Makefile.inc and everything else in the -dev
package was moved into the main machinekit package some time ago.
There are only four machinekit-dev packages left in the repo and they
match the also deprecated wheezy and raspian packages in the repo,
which
If you are building from sources please read the documentation
http://www.machinekit.io/docs/developing/machinekit-developing/
Sounds like you probably have not set the pathspec, which is covered
in this document.
On 08/08/18 11:43, mngr wrote:
On 08/08/18 03:43, Cameron McQuinn wrote:
..was able to successfully start the config several months ago,
but I had to set the project aside for a while. When I returned, I
upgraded the kernel from 4.9 to 4.14, and upgraded machinekit and
QtQuickVcp to the latest versions, but
Dear All,
Some while back I undertook work to split out machinekit into its HAL
stack and its CNC stack.
This work has reached maturity, to an extent that if the machinekit-hal
package is installed and then the machinekit-cnc package is installed,
a fully functioning machinekit is
On 27/05/18 11:26, Michael Brown wrote:
I dont know if you are aware that capitalization matters
the correct filenames are:
/lib/firmware/socfpga/DE10_Nano_FB_Cramps.3x24_cap.rbf
Hi
Yes, we are trying to get the HAL side of Machinekit as generic as
possible, there is a lot more could be done aside from CNC machine
control.
If you fork machinekit, clone to your computer and then create a
suitably named branch and put your driver in
On 18/05/18 22:14, fairbanks-john
wrote:
To day I decided that the cause of my problem is most
likely the version of gmoccapy available with machinekit. I
had modified my configuration files to make gmoccapy happy but
it
On 17/05/18 09:45, fairbanks-john
wrote:
Schooner I think that I pushed a wrong buttons while
posting and messed things up a bit. I got a new computer and
build a new configuration since I think the old machine may
have
On 16/05/18 16:39, Eric wrote:
I'm guessing that updating the kernel on the
Machinekit image isn't as trivial as installing an app? (can you
tell I'm new to Linux! :)
I don't think you have ever said what image you have, so no idea
On 15/05/18 08:29, fairbanks-john wrote:
I just made a fresh install from the git hub clone today and wondering
if I did something wrong since there are a lot of warnings and notes
in the installation log which I have attached..
The .hal file I have used before for testing fails to complete
On 14/05/18 19:10, Eric wrote:
BeagleBone wireless, including the wifi
chip: http://octavosystems.com/2016/09/27/the-beaglebone-black-wireless-and-beagleboard-compatibility/
Getting wifi up and running on an image I didn't create
On 14/05/18 12:39, Eric wrote:
Ahh, ok so that got me farther, but now I'm
getting a "no keyserver available" error...
Unless you detail what you did and exactly what the error was,
no-one will be able to help
If you
> Apr 27 11:13:30 D525MW dockerd[4167]: Error starting daemon:
Error initializing network controller: Error creating default…e
network
The print is truncated, there should be a more detailed error
explanation following a colon after the second ' Error'
eg.
I am running v18.03 on an amd64 Buster desktop and v17.03 on my amd64
Stretch server without problems.
As Charles says, it will not run on other than a POSIX kernel, had
problems in the past when I booted into the RT
kernel without thinking about it.
There may be something else in the wind
Have a look at rs274ngc_interp.hh:L233
The functions starting comp_get_current(...) appear
to get and set the values in the initial settings struct.
Initialised in rs274ngc_pre:L113 and struct detail in
interp_setup.cc / interp_internal.hh
I haven't looked
On 21/04/18 09:29, fairbanks-john
wrote:
I'm finally continuing my work on building a new
gui and switching this in progress work from linuxcnc to
machinekit. I'm using my ACRA milling machine as a test bed
which I do need to be able to
Please see also https://github.com/machinekit/machinekit-hal
This is a standalone Machinekit HAL stack without interpreter, motion,
NML etc etc.
At present the python vcp stuff is in the corresponding CNC stack at
https://github.com/machinekit/machinekit-cnc
If you need virtual panels etc
On 12/04/18 16:00,
j...@allwinedesigns.com wrote:
I definitely have to stick with Machinekit as I'm
running the machine with a Beaglebone. What is the nature of the
changes made to LinuxCNC that allowed this to work? Do you think
I
On 12/04/18 09:29,
'schoone...@btinternet.com' via Machinekit wrote:
This code snippet from my lathe pulley speed setting script shows
the process to convert and use case / esac to switch the input to
the correct call
Interesting concept.
In theory there should be nothing to prevent the combination of two
kinematics types
and a conditional test based upon whether a pin is set, to decide
which is used.
Implementation might be a little more complicated.
The data
Well done Michael,
Just tested it and works fine on a DE0-NANO too.
If you have no objections I will add them to the images at
http://deb.machinekit.io/uploads/de0-nano/
Notes for anyone looking to test.
Instructions are in the file
On 05/04/18 10:07, 'fishy' via
Machinekit wrote:
On Wednesday, 4 April 2018 16:41:46 UTC+1, Schooner wrote:
It is a community project, but thus far everyone (with minimal
On 28 Mar 2018, at 22:02, Rick Mann wrote:
I'm just starting on building a CNC router using MachineKit and a BBB. You'll
still need a board of some kind, if for no other reason than to isolate the BBB
from the rest of your system. I bought an OPTOCAPE ($165) to
Yes, Happy Birthday to us
What is your inbox like if you have 3 years of these saved to reply to
John !!!
On 04/04/18 00:37, John Morris wrote:
Happy Birthday #4 Machinekit!
While I've probably *contributed to* Machinekit *less* in the last
year than ever before, it turns out I've *used*
from the header to a buffered BOB and connect
to that.
You don't want to be connecting directly to a £90 board if you can help
it, too easy to fry something.
On 4 April 2018 at 18:51, 'schoone...@btinternet.com' via Machinekit
<machinekit@googlegroups.com> wrote:
On 04/04/18 08:48, Richar
It is floating point, otherwise it could only ever be accurate to
the nearest whole unit
On 30/03/18 08:59, John Dammeyer wrote:
I'll answer my own question. I was searching through
the HAL docs rather than the
The linuxcnc log only shows after the error, need to see what it was
that produced 'operation not permitted' from insmod
We don't know what you upgraded from exactly, but for quite some
time now the /usr/lib/linuxcnc/prubin/*.bin have been symlinked
through to
On 12/03/18 14:48, Maxime Lemonnier
wrote:
Hi MachineKit,
It's been almost 8 years since I used my pico systems
univpwm system (5 axis brushless servos). At the time I was in
the process of rewriting Axis, that I
On 03/03/18 16:17, Aurelien wrote:
Many thanks for reply
yes this is your code but modified by Mike Ross for 6 tool turret
For the Axis i have set as you says LINEAR and with trial and error
have found the good SCALE for turn 360° exactly
i don't understand "You simply home axis A"
It would be very helpful if you actually said what you had found and
provided a link to it.
From the comments in the code, I can tell that it is one I wrote
some time ago and posted here
http://www.machinekit.io/docs/developing/toolchangers/
amongst other
On 28/02/18 15:31,
'schoone...@btinternet.com' via Machinekit wrote:
On 28/02/18 12:11, Richard Thornton
wrote:
Hi,
Thanks for all the help.
So I
On 28/02/18 12:11, Richard Thornton
wrote:
Hi,
Thanks for all the help.
So I have the DE0-Nano-SoC, G540 and steppers
are there instructions on how to install machinekit on the
On 23/02/18 07:26,
'schoone...@btinternet.com' via Machinekit wrote:
On 23/02/2018 01:12, Aurelien wrote:
this message is sended by the driver i have found
the output text error in source code
On 23/02/2018 01:12, Aurelien wrote:
this message is sended by the driver i have found the output text
error in source code
xhc.hal = (xhc_hal_t *)hal_malloc(sizeof(xhc_hal_t));
if (xhc.hal == NULL) {
fprintf(stderr, "%s: ERROR: unable to allocate HAL shared memory\n",
modname);
exit(1);
Hopefully with this info, someone who knows about the BBB will be
able to see what is likely happening
A SWAG from me would be that the correct device tree file is not
being used, but what do I know
;-)
On 18/02/18 12:25, Marius Alksnys
wrote:
On 18/02/18 12:01, Marius Alksnys
wrote:
Line 28:
loadrt hal_bb_gpio
output_pins=816,822,823,824,825,826,914,923,925
input_pins=807,808,809,810,817,911,913
I know nothing about BBB specifically
On 18/02/18 11:37, Marius Alksnys wrote:
CRAMPS.hal:28: insmod failed, returned -1:
Whatever you are trying to load at line 28 of your hal file is what is
producing the error
That dangling 'DEALER' socket error sometimes shows up even in BB Black.
Don't worry about that.
Since we
Not a subject I know much about, this seems to give the best overview of
the kinematics.
http://linuxcnc.org/docs/ja/html/motion/5-axis-kinematics.html
I think the XYZAC example at the very end is nearest what you are
proposing, with A and C rotating in relation to Z
The G codes remain the
Ha, very good!
On 17/02/18 11:41, fritz wrote:
I first read that it had a name on the webcomic XKCD, but it appears
that this phenomena now has its own website:
http://xyproblem.info/
On 02/16/2018 11:21 AM, 'schoone...@btinternet.com' via Machinekit wrote:
Now we know what you are doing
Now we know what you are doing, it is easier to assist.
You really need to ask a question about your problem, not how to
effect the solution you have thought of.
The usual way to test backlash with a mill / lathe is to use a dial
gauge and move the bed a
I suggest you read the posts on the linuxcnc forum.
More words have been written on this than most other subjects.
For the most part, the people who ask, want to exceed the limits of
their stepper motors but escape the consequences.
On 15/02/18 07:37, Filip Kotouček
wrote:
Hello everyone,
I would like to ask if there is possibility to build
Machinekit to other Linux distribution than Debian. To be
specific we have platform based on (am335x)
First thing to check is probably that your dip switches are correctly set.
Seem to recall my DE0 board arrived with them in FPGA programming mode
positions
On 08/02/2018 05:09, mugginsac wrote:
Thanks Schooner and Michael,
I followed the instructions through the setenv default -a; setenv
On 07/02/18 07:03,
'schoone...@btinternet.com' via Machinekit wrote:
Read the gist that Michael did for the DE0
Specifically here
https://gist.github.com/mhaberler/89a813dc70688e35d8848e8e467a1337
where you will see exactly
Read the gist that Michael did for the DE0
There is a link to it in the blogs.
You have to set the NIC MAC address etc and then restart
On 06/02/2018 23:54, mugginsac wrote:
I finally got the USB console working under stretch.
I didn't have the dtb file that it was looking
48 matches
Mail list logo