On Fri, Feb 14, 2020 at 2:56 PM Chris Albertson <albertson.ch...@gmail.com> wrote:
> On Thu, Feb 13, 2020 at 11:43 AM Gene Heskett <ghesk...@shentel.net> > wrote: > > No. You should not need to know anything. The fact that you do need to is > a sign that the technology is immature. Back when cars were a new and > immature technology drivers had to know what a "spark advance" was and also > understand the effect of the mixture controls and how to use a clutch and > shift gears. In 20 years a small child will be able to operate a car > by sitting in the back seat and saying "Take me the gramma's house" > > All tech works that way. At first, it is usable by experts then it gets > easy. Hobby level CNC has a LONG way to go before it gets easy. > > As I pointed out before a good example to follow is "cleanflight". It > lets anyone with minimal tech background configure the software inside a > multi-copter drone and download it onto the chip. They do not need to > even know the system uses an RTOS. http://cleanflight.com It is all > open > source and is easy to use > > There is no good technical reason a CNC machine can't connect to a normal > desktop computer or phone over WiFi and "just work". > > -- > > Chris Albertson > Redondo Beach, California > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users It seems to me the conversation should be about creating a "printer driver" for the machine tool instead of burdening the development of LinuxCNC with the attempt to integrate functions beyond controlling the machine tool. Almost any CAM system I have used will do that already if you have a restricted set of parameters to generate the symbolic code to send to the machine ie. printer machine tool defined paper source and size defined raw stock source and size defined product (2d printed page) defined product (something fitting the available tools and raw stock) etc. I have had sets of product with defined parameters and a program in my CAM system to take the input (a very restricted geometry set) and output the code to run the machine. This was done 30 years ago and was very successful. My customer would send me geometry for his product. I would then input the geometry into my system in much the same manner as creating a flyer to print. This would take at most 5 minutes from receipt to final machine code. I would then send it to the machine tool. Load the raw stock (analogous to loading the paper size into the printer). Load the cutter (analogous to loading the ink pens/cartridges). The only thing I didn't do was have a grid on the machine tool table to allow me to press cycle start without setting an origin point for the program. This whole process could have been automated to be exactly like sending a print job to the printer. That level of automation was deemed by me to be not worth the time it would take. When each and every step of your process is repeatable it is easy (but not always financially realistic) to automate in a complete fashion. 3D printing is much the same fashion as 2D printing. The code generation driver can be (but not always is) embedded into the 3D printer controller. I remember being told NC programmers would soon be a thing of the past as IGES and programming system would soon be integrated to allow manufacturing almost without human interference. This was in the late 70's/early 80's. Heh not quite there yet but almost. :) StepNC started with one of their goals to automate symbolic code generation for controlling machine tools without human interference. If I remember correct this was in the early/middle 90's. We are almost there but StepNC seems to have a slightly different goal. The original laptop computer (the abacus) was able to compute much faster than the human mind. I wonder if people thought then it would only be a matter of time until man had to not think anymore. The manufacturing process remains a moving target for the the CAM systems. The removal of human interference is the goal of many projects not only in manufacturing. Supposedly a worthy goal but this goal comes with the necessary reduction in complexity of the project. The reduction of complexity is a worthy goal but it remains a very fast moving target. I also remember the propaganda statement "NC/CNC will allow the use of monkeys off the street successfully running the machine tools making parts." This hasn't worked out very well either. The insistence machine tools should be as easy to operate as a printer has been a success for a few decades now, when all the necessary parameters are defined and repeatable. Symbolic code generation (of which g code is a subset) would be a worthy focus. I don't know what is sent to a printer to cause the printer to output a printed page. It seems driver development is what allows printers to evolve in function and capability. Printers have long had the capability to take generic (although defined) instruction sets and generate the necessary commands to print a page. It is possible on many printer drivers and printers to choose whether to process the printer instructions in the computer prior to being sent to the printer or having the printer internally process the printer instructions. I think it is eminently achievable but it will not be done any time soon with the current tools. If unlimited funds was the only requirement to achieve 'Take me to gramma's house' it would have already been done. There are just some things with parameters and variables beyond computer capability. Without restricted parameters and variables (such as rails for a train/trolley/elevator) and vastly changed infrastructure 'Take me to gramma's with perfect safety' without human interference is impossible. Even the speed limit on our highways is a compromise of safety. The chosen authority is given a set of parameters (acceptable deaths per mile) and tasked to assign speed limits to meet the set of parameters. If a lack of deaths on the highway was the goal the speed limit would be 5 or 10 miles per hour but in certain instances that would be too fast. Pretty soon, we would have to remove all cars and roads because they cause too many deaths. Is this what is known as a conundrum? The parameters and variables in manufacturing with machine tools is one of the conundrums. It used to be a lathe was a lathe did turning and a mill did milling THEN someone decided to put them together. THEN someone decided multiaxis is the way to go. THEN new tooling is developed. Waterjets and plasma cutters and turret punch presses are much like printers except someone put 5 axis capability on waterjets and bending capability on turret punch presses. WHEN OH WHEN do the changes stop? In my view hopefully never. Hobby level CNC control is just as capable and easy to run as "commercial level" CNC control in a production environment. Hobby level CNC control is MUCH easier to implement on a machine tool than commercial level CNC. You just don't see many Hobby level attempts to install commercial level CNC and you NEVER hear of that person attempting that install complain about the complexity of the software and install. You WILL see that person spend a lot of money for assistance and probably a lot of time for training in how to interpret the instructions and handle the hardware. It already seems a lot easier to install the Hobby level LinuxCNC. The Hobby level help response time is certainly MUCH quicker and unquestionably MUCH cheaper. If you are short of time and capability by all means have someone install a commercial CNC control. You will probably have the machine running sooner than you would a first Hobby level LinuxCNC control. A second commercial CNC control would install in much the same time frame but a Hobby level LinuxCNC install would be much quicker than the first Hobby level LinuxCNC install. The third Hobby level LinuxCNC control install would likely rival the commercial CNC control install in the elapsed time of the install. The Hobby level installer would have an intimate understanding of the machine and control that would lead to quicker/cheaper diagnosis and correction of a problem. No waiting for the commercial technician to arrive. Operation of commercial level controls is always a learning experience. They try to stay consistent but new technology always inflicts changes in the HMI to make it another learning experience. I have an IFM key on the bottom of my keyboard. When the problem gets overwhelming I turn my keyboard over and push it. Problem solved. :) Just my 2 cents Stuart -- Addressee is the intended audience. If you are not the addressee then my consent is not given for you to read this email furthermore it is my wish you would close this without saving or reading, and cease and desist from saving or opening my private correspondence. Thank you for honoring my wish. _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
