Sorry to mention that yes it is SP3 and today when i turned it on it did some auto updates.
How do i know if the Net framework updates are correct?
Go to control panel and then add and remove programs. It will tell you in there what net framework versions are installed. Net Framework 4 should be the latest.
@SungeunJeon - Good point, Actually UGS has been very solid for me right up till the time when it was not. But I was trying to run a very long job that is probably outside of any normal testing anyone would perform. So I should just notify Will Winder as you suggest and thank him for releasing his very useful software out into the world at no fee.
1 Like
downloaded current version of Net Framework and now different error message didnât jot it down too late and I get up early for work, something about not finding something to do with Net Framework.
UGS 1.0.8 was released 5 months ago, and 1.0.7 over a year ago. There are many bugfixes and performance improvements in the nightly build that resolve all the issues I see mentioned in this thread. GRBL performance improvements in the past year exposed some issues in UGS, all the issues I know of have been resolved in the latest nightly build. If you have any doubt at all you can (and should) always use the âCheck Modeâ option prior to sending a job. This will cause UGS to behave exactly the same as it would during a true file send.
UGS has a large community of users that report issues and those issues are generally fixed in short order - the majority of the open github issues are feature requests.
1 Like
Do you have any suggestions I can try to get 1.0.8 to run? 1.0.7 runs with no problem on my WIndows 8.1 machine, but 1.0.9 give me 5 Java errors when I run the batch file to start it.
I have tried removing Java and reinstalling the latest version of Java, but I still get the same errors.
If you make a ticket on the UGS github page and post your stack trace I can take a look. You could also try the 2.0 build which may already have a fix.
Thanks I will try 2.0 first
thanks will, i will give 2.0 a try and see how it goes. i like the ease of use UGS has i have just been unhappy (read:terrified) of the unpredictablility
@WillWinder - Version 2.0 loaded fine, not sure what the issue was with 1.0.8. Thanks for the suggestion, I was reluctant to load the 2.0 version not knowing how stable it was.
If you donât mind I have two questions
-
Does the number of line of gcode make any difference to how UGS function, for example if my gocde file contains 400,000 lines of gcode does that effect how UGS process the file. Is UGS loading the file into memory before it begins sending it to the grbl? I am just trying to understand the relationship (if any) to how the size of the file will effect the functioning of UGS.
-
Does UGS have any way to zero out the machine positions for X,Y,Z? I have been having to exit UGS and then restart it to zero those machine coords, but there must be a better method. The reset X,Y,Z will zero the Work Position but has do effect on machine position.
Thanks for your help.
Thanks for the feedback @BadWolf, @AllenMassey. My #1 goal when creating UGS 3 years ago was that it be as easy to use as possible - getting a single file that executes on Windows/Linux/Mac was a big hurdle early on.
@AllenMassey Yes, the current version of UGS loads the entire file into memory so there is a point where the Java Virtual Machine runs out of memory and things become unstable. If you leave the visualizer closed for large files that will help, but coming up with a real solution is definitely on my TODO list.
Work coordinates are another area that needs more work. My machine doesnât have homing switches so these sorts of features havenât gotten the attention they deserve. There are reset coordinate buttons, do they not work properly for you?
1 Like
Maybe I am missing something, the only reset buttons I see only reset the work coords. The machine coords stay at whatever value they have.
Allen, sorry for squeezing between. UGS have nothing to do if your GRBL runs out of Buffer. If there is no back-up Buffer, program stops. Try this as you see on the picture. I just took a shot of my Screen. You can increase Virtual Memory like I did.
Iâd like to challenge Sonny on his comments that ChiliPeppr is not 100% dependable. What data are you basing that on?
Both of the efforts for open source gcode sending software are to be commended. Chilipepper and UGS are amazing gifts that have been given to the public, and I for one am very proud of the guys that devote the time and effort and resources to release these products.
It is understandable that the respective âcampsâ, if you will, are protective of their âbabysâ, as well they should be. I would feel the same if it were my software development being discussed.
I have enjoyed this thread, and hope that it will help spur collaborative discussion that results in positive results for the end user.
My X-carve is scheduled to be delivered tomorrow at my house. I will probably have it assembled and start playing with it this weekend. Im sure I will have some feedback, both good and bad, withing the next few days.
Anyway, good job fellas, thanks for doing what you do.
My comment on Chilipeppr not being 100% dependable is based on my thorough stress testing of Chilipeppr, along with a whole host of other Grbl streamers. If you try to stream an aggressive and complex program (~5-10ms per g-code block), the Chilipeppr in the browser becomes completely non-responsive or slows to a crawl. On occasion, it causes the browser to crash itself. Even on a Macbook Pro only a few years old. (in both Chrome and Safari).
Having no control of the machine when this happens is a dangerous proposition, especially when machines get large, like a CNCâd Bridgeport knee-mill. Even though you may have the Serial Port JSON Server (SPJS) still going and streaming, you still have to reload the browser and reconnect or wait until it becomes responsive again to control your machine.
For a machine control, the top priority should always be safety, everything else is secondary.
In addition, SPJS itself seems to be fairly robust from my tests, but has a hard time with the upper end of streaming performance when the Go garbage collection kicks in. You can hear audible clicks during a high-performance job when this happens. This can lead to lost steps or a motor stalling.
John, youâve done an awesome job with Chilipeppr. Donât take these comments as a slight, but take them as constructive criticism. Iâm certain that you and your developers can take steps and alleviate these issues quickly. Until then, I can still only recommend Chilipeppr for small, scale hobby-level jobs that are relatively simple and slow.
@AlanDavis - I am confused, how does increasing virtual memory on the computer help the GRBL shield have more buffer. I was under the impression that the Arduino was running the GRBL process and the Arduino does not share any resources with the main computer to my knowledge.
You can solve the problem of your browser becoming unresponsive by disabling the 3D viewer if your computer canât keep up with it. ChiliPeppr uses a modern feature of the browser called requestAnimationFrame() where the OS gives a callback when there is ample CPU available. On Mac that call is not super reliable. Its useful though so the computer decides how many frames per second can be run vs you as the user having to toggle this, but I did give a manual override if you see performance issues.
All software has the possibility of not being readily available, so there are no greater risks in ChiliPeppr than anything else. Your computer could go to sleep (which many folks have run into), your OS could be crashed, etc. Thatâs why pretty much all folks have their e-stop next to their machine, not in their software.
SPJS is compiled and quite close to C code. That gives it the best chance of being the most performant software out there for talking to the CNC controller. It does have itâs own garbage collection, but the recent Go compiler has huge improvements to it. So, in my tests, SPJS is as fast as anything else out there.
nevermind moved the file to my desktop and it opened right up
I meant Arduino. When computer sending information, Arduino have limited amount of memory to hold. If computer sees that starts stocking buffer with that information and send it bit by bit. To hold this information generated by the UGS, windows needs more virtual space to hold it. Is that make sense.
1 Like