Hmmm... I can see ways to cache extra information (like entry and exit velocities) to get it close/exact, but it would likely require changes to the blending function. I will not have time for that any time soon.
On a related note, is there any way we can instrument the code to generate some statistics on how far it varies given various normal and extreme test cases? A few measurements will trump all the speculation (in the absense of mathematical proofs). On Sep 11 2015 10:16 AM, Robert Ellenberg wrote: > Some information is saved in the queue structure, but the exact path > taken > by a parabolic blend (as implemented in linuxcnc) depends on the > initial > conditions, so you wouldn't really get the same path played back in > reverse > unless you hit it at exactly the same speed. > > Rob > On Sep 9, 2015 11:21 PM, "EBo" <[email protected]> wrote: > >> Is it possible that the motions (including blending) are cached? >> Then >> it would be a lot easier to roll them back over the exact same >> trajectory. Just a thought... >> >> On Sep 9 2015 12:30 PM, sam sokolik wrote: >> > And to show more of robs awesomeness - Jeff said I should try >> > reversing >> > through a tool change.. >> > >> > It stops reversing once it backs up to the tool change location... >> > >> > sam >> > >> > On 9/8/2015 7:51 PM, Robert Ellenberg wrote: >> >> Hi Kenneth, >> >> >> >> There's no fundamental limitation here, it's just a matter of how >> >> big we >> >> make the queue structure. It's a circular buffer, so it could be >> >> 1000 >> >> segments longer if we could afford the space. I just made it 100 >> >> segments >> >> because it seemed reasonable. For context, the stock queue is >> 2000 >> >> segments >> >> (forward only, of course). One reason to limit the size is that >> if I >> >> ever >> >> get tangent blends working in reverse, it will be more expensive >> to >> >> optimize a lot of segments. >> >> >> >> Seb, l'll rebase onto master before I push it to the main repo. >> >> >> >> Rob >> >> >> >> On Tue, Sep 8, 2015 at 12:25 PM, Jon Elson >> <[email protected]> >> >> wrote: >> >> >> >>> On 09/08/2015 07:43 AM, Kenneth Lerman wrote: >> >>>> Why is there a 100 step limit? If it is because that's way more >> >>>> than >> >>> anyone >> >>>> would ever need, that's fine. >> >>>> >> >>>> Memory is cheap. At 1000 bytes per step, storing 10 thousand >> steps >> >>>> is >> >>> only >> >>>> 10 meg. That's not much in a machine with a gigabyte or more. >> >>>> >> >>>> >> >>> it is not "steps" as in stepper motor steps. it is move >> >>> segments, which is either blocks of G-code for linear moves, >> >>> or maybe interpolated segments of arcs for non-linear >> >>> moves. That covers a lot more territory! >> >>> >> >>> Jon >> >>> >> >>> >> >>> >> >> >> >> >> >> ------------------------------------------------------------------------------ >> >>> _______________________________________________ >> >>> Emc-developers mailing list >> >>> [email protected] >> >>> https://lists.sourceforge.net/lists/listinfo/emc-developers >> >>> >> >> -- >> >> Kenneth Lerman >> >> 55 Main Street >> >> Newtown, CT 06470 >> >> >> >> >> >> ------------------------------------------------------------------------------ >> >> _______________________________________________ >> >> Emc-developers mailing list >> >> [email protected] >> >> https://lists.sourceforge.net/lists/listinfo/emc-developers >> >> >> >> >> >> ------------------------------------------------------------------------------ >> >> Monitor Your Dynamic Infrastructure at Any Scale With Datadog! >> >> Get real-time metrics from all of your servers, apps and tools >> >> in one place. >> >> SourceForge users - Click here to start your Free Trial of >> Datadog >> >> now! >> >> http://pubads.g.doubleclick.net/gampad/clk?id=241902991&iu=/4140 >> >> _______________________________________________ >> >> Emc-developers mailing list >> >> [email protected] >> >> https://lists.sourceforge.net/lists/listinfo/emc-developers >> >> >> >> >> > >> > >> > >> > >> >> ------------------------------------------------------------------------------ >> > Monitor Your Dynamic Infrastructure at Any Scale With Datadog! >> > Get real-time metrics from all of your servers, apps and tools >> > in one place. >> > SourceForge users - Click here to start your Free Trial of Datadog >> > now! >> > http://pubads.g.doubleclick.net/gampad/clk?id=241902991&iu=/4140 >> > _______________________________________________ >> > Emc-developers mailing list >> > [email protected] >> > https://lists.sourceforge.net/lists/listinfo/emc-developers >> >> >> >> >> ------------------------------------------------------------------------------ >> Monitor Your Dynamic Infrastructure at Any Scale With Datadog! >> Get real-time metrics from all of your servers, apps and tools >> in one place. >> SourceForge users - Click here to start your Free Trial of Datadog >> now! >> http://pubads.g.doubleclick.net/gampad/clk?id=241902991&iu=/4140 >> _______________________________________________ >> Emc-developers mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/emc-developers >> > > ------------------------------------------------------------------------------ > _______________________________________________ > Emc-developers mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-developers ------------------------------------------------------------------------------ _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
