On Thu, Oct 12, 2017 at 5:51 PM, John Dammeyer <jo...@autoartisans.com> wrote:
> Click into the field and hit F1 and get a full paragraph explanation with > examples. Think how long EMC or Linux CNC has been around and figure it might be around thatch longer. OK so it is not 2025 Problem: There is no F1 key. Tim Cook was right and anyone not using AR glases is irrelevant. Users expect to just "one finger point" to a value that is floating in space and say "5" and it changes to a 5. OK so I might be wrong. Point is that if the software is to have a long life, you want to de-couple the interface the user sees from the implementation. If the data is stored in XML the user should never see XML unless he goes digging. No matter what you do, it will look silly and archaic at some point in the future I like parameter servers. They provide a high level of isolation and would allow some one later to implement to silly AR glasses interface without need to change even one line of code in the machine tool app. In the end. I think you have to decide who the target users are and what you are building for them. Currently Machinket seem to be a product targeted at developers who want to build a machine controller. Those users are happy to edit all kinds of files (C++ included) But if you want to build a machine controller and the target user is some one who knows very little about computer exalt maybe how to send emails and work a web browser then you This is what the people at Tormach did. They used one product, LinuxCNC to build another product, Pathpilot. This is the best way to go, to have several layers of abstraction > > When you click CANCEL on the form, all the parameters (which are copies of > the original) are discarded. If you click SAVE, the parameters are verified > and copied into the original values and a call is done to update the XML > file. > > If you don't want to use the form, then load the XML editor and the file and > tab down and change the values directly. Not sure why you'd want to but > it's possible. Especially if some sort of code upgrade doesn't update the > user interface. But a code upgrade of parameters in something as > fundamental as an ini file should generate updated forms. > > This isn't rocket science since I've been writing software like that for > users for more than 25 years. Probably longer but I don't want to do the > math. > > The copout for Linux has always been that in depth user manuals etc. aren't > needed because you can just read the source code. Of course often that > source code is commented like this: > > localInt = 5; // Put 5 into localInt. > > Duh! > > The Xylotex.ini file is just another of the many examples of incomplete or > simply incorrect information that requires far more than what is required > for a system designed for users. > > I apologize. I've once again allowed myself to get pulled into this world > and started complaining. I'm probably bringing my work frustrations to the > forum because I've got a junior Mechanical Engineer In Training who's been > able to convince management that he's perfectly capable of setting up a > complete source code control system for a micro processor not supported by > Microsoft Visual Studio and once done he can check out files, make changes > and recompile. Apparently 10 years ago in school he took a couple of > software courses. The questions I get are mind boggling. > > John > > > >> -----Original Message----- >> From: Chris Albertson [mailto:albertson.ch...@gmail.com] >> Sent: October-12-17 4:31 PM >> To: Enhanced Machine Controller (EMC) >> Subject: Re: [Emc-users] Path Pilot on LinuxCNC -- Was MachineKit on the >> BeagleBone Black >> >> On Thu, Oct 12, 2017 at 3:17 PM, John Dammeyer <jo...@autoartisans.com> >> wrote: >> >> > > I would much prefer to edit a well thought out and documented config > file >> > > than use pretty much any Grunt n' Click interface. >> > > People who come from Win and OSX are not used to that, so it seems > crude >> > to >> > >> > # This gives a maxvel of 12.5/1 = 12.5 ips >> > # >> > # these are in nanoseconds >> > DIRSETUP = 1000 >> > DIRHOLD = 20000 >> > STEPLEN = 500 >> > STEPSPACE = 4000 >> > >> >> >> >> Agree. Yes the above is a perfect example of the worst possible >> user interface design I can't see any place there the four parameter > names >> are described. And there is zero enforcement of any rules or even basic >> sanity checking. >> >> At the very least this could have been implemented with an HTML form >> where clicking on the parameter name brings up a pop-up wit the > parameter's >> definition >> and clicking on the value allows either direct entry or a calculator. >> Then a validation is run to see that all values are valid and >> self-consistant. >> >> The text editor allows the user to do things like >> DIRSETUP = 1w4r >> Which is just about the poorest user interface design possible as it fails >> to >> catch even basic syntactical errors until AFTER the user has canceled out >> of the >> form. Seriously, who ould design an interface that tell you about > simple >> typo only AFTER closing out the screen where you enter the data? >> A reasonable design would at least catch an obvious error when the user >> clicks "submit" or "save" A better design would not allow bad data to be >> entered. >> >> Curent bast practice is to inform the use as soon as it is logically >> possible. >> >> Point is you do NOT have to give up flexibility for correct and modern >> design >> you can have both and all it takes is a simple HTML form and some >> javascript. >> Of it you don't like that Java/Swing or whatever. >> >> A correctly design interface would allow exactly the same amount of >> flexibility >> but would also provide information about what neds to be entered, have > links >> to a reference text and then validate the user's entry. >> >> >> >> >> -- >> >> Chris Albertson >> Redondo Beach, California >> > ---------------------------------------------------------------------------- > -- >> Check out the vibrant tech community on one of the world's most >> engaging tech sites, Slashdot.org! http://sdm.link/slashdot >> _______________________________________________ >> Emc-users mailing list >> Emc-users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/emc-users > > > ------------------------------------------------------------------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users -- Chris Albertson Redondo Beach, California ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
