On Thu, Mar 15, 2012 at 10:10 PM, Roger Holmquist <ro...@abcnc.se> wrote: > One observation is that trajectory planning seem to be incomplete.
In what way? I think the whole LinuxCNC-community could benefit from automated interpreter+trajectory-control tests. These would be G-code programs that are run, and the machine position from HAL would be recorded each servo cycle. Analysis of these position trace-files is better grounds for saying this or that about trajectory-control than "I have a feeling"-type of statements. Also as there seems to be some alternative trajectory-control work in the Arais Robot Technology github fork of LinuxCNC, and possibly some old segmentqueue work somewhere (Taig?BDI?). If there was a set of (complete?) G-code+trajectory tests available then one could evaluate the correctness and "goodness" of these alternative trajectory-controllers. > There is also a need to find reliable HW for machine-interfacing. The wiki has a list of hardware options for I/O: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?LinuxCNC_Supported_Hardware AW ------------------------------------------------------------------------------ This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
