Hi Rich,
Wish I was a lathe user, sometimes. 

Believe me you do not

Brian always tells me to nag to get him to work on Turn but so far I have resisted the urge as I want him to spend time on Rev4 and fixing up Turn in Rev3 will just distract him from doing so. Now that the new guy has started maybe I will

Will be sending you a PM in a few mins.
Hood

Tweakie:

  Thx, that narrows it down , also makes it applicaion, not driver so at least I know I didnt make that bug on my end
over time. . Ill discuss it with Brian when he returns and we'll see whats up.

Have fun,
Art

Thanks Art, your assistance is much appreciated.

Tweakie.

Hello guys,
I would like to say that I am sorry that I am not able to make it so the software can be everything to everyone as much as I would like it to be. The trouble that we have is that Mach3 (originally Mach1) was designed to be a simple CNC app (for routers and mills) that had little in the way of customization. Then Mach3 grew into a CNC App with some customization with scripts and so on.. GREAT! That makes it so you can do anything! (also you can screw up anything). Okay Now we have added Plugins so you can have external devices and add IO. What a joy! we can add and have about anything we like on the machine but this came at a very big price. The base code in Mach3 is not ideal for what we have today and working with it is like trying to add on to a house of cards. ANY card out of place it will all fall. You have no idea the pain that I go through to work on the code in Mach3 and how much I fight to NOT change it.. I have been trying to add features in a way that is not going to mess up the basic application. But with everything being global pointers I can't stop anything from changing data in Mach3. So there is the sob story.. what are we doing about!

I am working on Mach3 all the time looking for issues. I spent months working on Rays issue and found some of them (both of us to blame). At the end of the day he will end up with his own little interface for the KFlop and I am happy that he is happy. But I lost months of dev time and I wouldn't change what I did. The fact that he didn't get what he wanted was not a lack of effort. I offered more then one time to fly him out with his PC work work at the office here and to get it working as he would like. I also offered to fly out to work on his machine. I don't know what more I can do other then to build up machine and send it to him. If that is not good support then I just don't know what to do and I give up. I work with so many people to fix there issues (90% is special interests) that I do very little dev work. to fix this we have brought on a full time programmer and support guy. This programmer works everyday on Mach4 and he is a programmer of over 25 years. He and I talk everyday making the API and then he makes the code to match the API. This is what needed to be done to make Mach4 be the most stable app that it can be. We have also made a decision that we are ripping out ALL the custom code for users with "special functions". THC, Foam cutting, Tangential,  Laser etc.. are getting ripped out of the software. We spend more time trying to get everything working for everyone and we can't do this anymore. Now having said that we are still going to have THC, Foam cutting, Tangential,  Laser etc.. but it will be moved off into modules that you can add.  This is making the core code to Mach4 much more fun to work on (it is simple). Also we are not going to have global pointers anymore (this is because we are making it so we can change the code and it will not effect the plugins) Plugins will no loger be able to play with Vars in Mach4. They will have to go through an API call so we can be sure that we are not getting any data overwrites. Also the P POrt is not going to be part of Mach4. the P Port will be a plugin that you can get or not.. The future is not going to be 25pins  LOL. I know this is not going to fix the issues that you have now but I would like you to see just a small glimpse of what I have to do everyday...

So what is this rant... Yes there are errors and I work everyday trying to fix them when I know about them. And we know what the base issue requires a complete rewrite that I can't do because I am supporting Mach3. To get you what you need in the future we have a programmer working full time and to get the current code doing what you need I am working full time. Contact me if you have an issue and I will work with you..


This may be more then you wanted to know but it is what it is.

Thanks
Brian

Hi Brian,

I have sent you a PM.

Tweakie.

The trouble that we have is that Mach3 (originally Mach1) was designed to be a simple CNC app (for routers and mills) ...
The base code in Mach3 is not ideal for what we have today and working with it is like trying to add on to a house of cards. ANY card out of place it will all fall. You have no idea the pain that I go through to work on the code in Mach3 and how much I fight to NOT change it..

Actually, some of us do have an idea of the pain.
And yes, the complete rewrite of the kernal is the only solution. Some of us KNOW this. (Painfully.)

So, two things: this sort of communication is vital, and some faint idea of the schedule would be very nice. Are we talking a week (hoho), a month (hoho) a year, or what? NO promises, just a hint to help your customers with their planning.

Cheers

Brian, please make the Tempest trajectory planner an option right from the start. Even unfinished, as it is right now, it can be a big improvement over Mach3's current CV mode.

Just to backup what Brian said, his support has been truly outstanding, and not just for me, but for many others as well.  He put a huge  amount of time and effort into working on the issues I raised, and quite a few were fixed along the way.  But, for the reasons he described, fixing one bug sometimes uncovered, or created, a different one.  That is why it is so important to get Mach4 finished - Mach3 will NEVER be as stable as it needs to be, because of its legacy.  This is the way of software - it always needs a major re-write after a certain period of time, and Mach3s time came some time ago.  It was ALWAYS a real pleasure working with Brian, and Greg, and I will really miss that.  I have no doubt when Mach4 finally arrives, it will be a HUGE improvement over Mach3!  And when it's ready, I'll be eager to give it a test-drive!

Regards,
Ray L.

Just want to ad my (usually controversial) two cents worth to this topic. I don't have a dog in this fight and I am not a Mach guru by any means, I do have a lot of experience in machine design, manufacturing processes and programming. I have recently taken my first foray into some basic 'bare metal' programming of Mach using Modbus (both serial and TCP), brains and screen mods. I have a need that Mach does not fulfil, so as I have done previously, I extract the data from Mach and create my own solution.

My observation is this; Mach is extensible and open, yet when someone has a specialty application, instead of using that open architecture to tap into the data that is available and using it to create their own application, they lay their special needs on the shoulders of the developer.

While I have to say that I am also distressed by the general instability of Mach, I can see now where some of it comes from; namely catering to the 'lobbyist's' instead of bulletproofing the existing code for those who dare to try use it in a production environment, or those who are foolish enough to invest a lot of time in developing around it, and let the 'special needs' group do their own customizing.

As an example of how pervasive the 'save us' attitude is, I made two suggestions for solutions to the issues raised by the OP.  Not a single comment on those ideas. Not even to blast me and say they suck and won't work? Nothing. No other ideas were put forward, yet the complaining continued unabated that Mach was not solving every problem.

Certainly there are going to be exceptions to this rule, but so long as the data is retrievable from withing Mach, people should use that capability to create their own special applications and not pressure the developer to refocus resources on the few to the detriment of the many.

My observation is this; Mach is extensible and open, yet when someone has a specialty application, instead of using that open architecture to tap into the data that is available and using it to create their own application, they lay their special needs on the shoulders of the developer.
While I have to say that I am also distressed by the general instability of Mach, I can see now where some of it comes from; namely catering to the 'lobbyist's' instead of bulletproofing the existing code for those who dare to try use it in a production environment,

Harsh words, and yet some truth maybe.
However, more debugging of the core and less 'feature extension' might prove commercially profitable.

Cheers