Software bundle 3.5beta1 released

[dc42]

On behalf of the Duet3d team I am pleased to announce this last-minute Christmas present to our users. Those of you brave enough to unwrap it can find the files for standalone systems at https://github.com/Duet3D/RepRapFirmware/releases/tag/3%2C5beta1 and the files for SBC systems on the unstable feed of the package server.

This is the first release since major changes to the motion planning system (to accommodate multiple asynchronous motion systems) so I am expecting bugs to turn up, despite the testing we have done on multiple printers covering the most popular kinematics.

Feel free to post your experiences in this thread; but please don't post long problem descriptions that require a response here. Instead, describe them in a new thread with [3.5beta1] in the title. Bear in mind that support from Duet3D will be limited until 3 January, so you may need to revert to release 3.4.5 to continue using your machine.

SBC users can find instructions for switching to the unstable package server and reverting if necessary at https://docs.duet3d.com/en/User_manual/Machine_configuration/DSF_RPi#switch-to-unstable-packages.

RRF release notes: https://github.com/Duet3D/RepRapFirmware/wiki/Changelog-RRF-3.x-Beta#reprapfirmware-350beta1
DWC release notes: https://github.com/Duet3D/DuetWebControl/wiki/Changelog-DWC-3.x-Beta#version-35-b1
DSF release notes: https://github.com/Duet3D/DuetSoftwareFramework/wiki/Changelog-DSF-3.x-Beta#version-35-b1

Happy holidays!

[CNCModeller]

@dc42 has any testing been done with polar machines?

If not I'll hang fire on mine until later in the new year when things are likely to be more stable.

Many thanks
Barry M

[T3P3Tony]

@CNCModeller no testing done on Polar machines so if yours is working properly with 3.4.x and you are prepared to take the plunge then you feedback will be welcome!

[dc42]

@CNCModeller no, we don't have any polar machines to test on. However we have made very few changes to the kinematics classes, and we have tested 3.5beta1 with other nonlinear kinematics such as Delta.

[Nightowl]

Just out of curiosity @dc42, did you manage to implement any improvements for Power Loss/Resume for CNC machines, please?

It was a while ago, but we discussed it here.

Thank you.

[mikeabuilder]

Sweet! Have a great holiday, Duet folks. And thanks for your solid work on RRF. Much appreciated.

[Exerqtor]

@dc42
Trying out some of the new stuff and i can't seem to get the fileexists function to work.

Just to try it out i added this at the end of homey.g:

if fileexists(homey.g)
  echo "fileexists works"

When i run the macro i get this error:
Error: in file macro line 65 column 15: meta command: unknown value 'homey.g'

[jay_s_uk]

@Exerqtor maybe try the full file path? I.e. 0:/sys/homey.g

[Exerqtor]

@jay_s_uk said in Software bundle 3.5beta1 released:

0:/sys/homey.g

That returns Error: in file macro line 65 column 16: meta command: expected ')' so I have no idea what I'm doing wrong😅

If I try /sys/homey.g I get this error X Y steppers at 100% of max current! Error: in file macro line 65 column 15: meta command: expected an expression.
And with sys/homey.g I get Error: in file macro line 65 column 15: meta command: unknown value 'sys'

[jay_s_uk]

@Exerqtor you may need to encase it in ` or "

[Sindarius]

@Exerqtor you probably need to quote the file name so it is a string

“0:/sys/homey.g”

Looks like @jay_s_uk beat me to it

[Exerqtor]

@jay_s_uk said in Software bundle 3.5beta1 released:

@Exerqtor you may need to encase it in ` or "

@Sindarius said in Software bundle 3.5beta1 released:

@Exerqtor you probably need to quote the file name so it is a string

“0:/sys/homey.g”

Looks like @jay_s_uk beat me to it

Yup, that sure worked!

Edit:

Don't need the full filepath either, since it's on 0:/sys/ it works with just the filename, and above that you only need /sys/foldername/ if it's located in 0:/ 🤘

[MaxGyver]

@dc42

I have noticed two minor things after updating:

• G10 without L parameter will cause my macros to get stuck

• M569.1 without an E parameter defined will trigger a warning/error

Thanks to the Team for all your hard work!

[chernogorsky]

Create a separate ticket. Movement issue + some minors

[dc42]

@MaxGyver said in Software bundle 3.5beta1 released:

@dc42

I have noticed two minor things after updating:

• G10 without L parameter will cause my macros to get stuck

Please start a separate thread about this. In particular, show the G10 command(s) you are using.

• M569.1 without an E parameter defined will trigger a warning/error

Please see the upgrade notes.

[dc42]

A bug in the 3.5beta 1 main board files has come to light. The bug causes a reset or other odd behaviour if the G32 command is used to perform true bed levelling on a machine with multiple Z motors.

I have now identified the cause and fixed this issue. New main board binary files containing the fix are available at https://www.dropbox.com/sh/28zqy8t165nwa2f/AACZvmVSGhWMSpjKX43MsSWSa?dl=0. A few other changes have been made in these builds, see https://github.com/Duet3D/RepRapFirmware/wiki/Changelog-RRF-3.x-Beta#reprapfirmware-350beta2-in-preparation for details.

Note, these main board binaries have received minimal testing; however the changes since version 3.5beta1 are small.

[A Former User]

This post is deleted!

[chernogorsky]

To anyone like me stacked with SBC and 3.5
my way of resotoring stable version (after changing /etc/apt. ... to stable)

sudo apt install duetcontrolserver/stable duetpimanagementplugin/stable duetpluginservice/stable duetsd/stable duetsoftwareframework/stable duettools/stable duetwebcontrol/stable duetwebserver/stable reprapfirmware/stable duetruntime/stable

[dc42]

@Nightowl said in Software bundle 3.5beta1 released:

Just out of curiosity @dc42, did you manage to implement any improvements for Power Loss/Resume for CNC machines, please?

It was a while ago, but we discussed it here.

Yes, commands to restore the spindle speed and direction are now included in resurrect.g.