Transition from Tormach to bigger machine

[ger21]

I haven't been following the Mach forum lately because I got tired of not understanding what they were talking about with back end code and programming, but the impression I got was exactly that. Mach 4 IS NOT for hobby use; Mach 4 lite or whatever it's actually called is. They are pushing it for robotics and big machines to handle a dozen or more "axis" at a time.

Actually, the main version of Mach4 is called Mach4 Hobby, and is targeted at the same people that use Mach3. There is also an industrial version, as well as Mach4 lite, which is only available bundled with machines.

I still don't know if Mach4 will be in full release in the next year. I think they need to spend a lot of effort in making it easy enough to setup and use for their typical customer. This includes building in a lot of the functions that are included in Mach3.

[WOTDesigns]

They set the bar pretty high for themselves so I really am looking forward to seeing their final release of mach4

Brian
WOT Designs

[daniellyall]

they have a lot of functions to add yet or get them of the second development team squire

[LeeWay]

Absolutely! I have always thought Mach 3 was on the cutting edge simply for ease of use and replacement controllers. PC's are cheap, reliable and available. As is information about mach controllers, especially with a few forums. I see a lot of guys here ask about specific NC controllers and get crickets. Ask a similar question about Mach 3 and the board comes alive sometimes.

[SCzEngrgGroup]

They set the bar pretty high for themselves so I really am looking forward to seeing their final release of mach4

Brian
WOT Designs

Mach4 is a total re-write, with only about 1% of the original Mach3 source code being carried over. Last time I tried it (a few months ago) it was still not to what I would consider "alpha", and many features were still not implemented. On a program of that complexity, I would expect at least 1-2 years to get from the current state to a stable, production-quality state (I spent many years in the software business....). Then there's gathering a critical mass of third-party support, especially motion controllers and peripheral plug-ins, ALL of which will require a complete re-write as well, which NO existing code carrying over.

And almost nothing users know about Mach3 will carry over either, as Mach4 screen-sets, macros, etc. are all completely different as well, and and will not work in Mach4. Even the macro language used is different (Lua instead of Cypress Basic).

When it's done, it will no doubt be a major improvement, especially in terms of reliability and maintainability. But it's still a looong way from being done.

Regards,
Ray L.

[WOTDesigns]

So glad I make aluminum parts instead of computer programs!

Brian
WOT Designs

[daniellyall]

what's there now is good for all basic stuff. its usable but there's only 4 plugins last time I looked, but its got a good laser controler

[WOTDesigns]

Hmm... Sure would be nice if Epilog upgrades their ridiculous 1995 vending machine style interface.



Brian
WOT Designs

[Steve Seebold]

There's all this talk about the Mach4 operating system. I still use Mach3 on my machine and it does everything I need it to do. "IF IT AIN'T BROKE DON'T FIX IT."

[LeeWay]

I use Mach 3 on every machine I own. It works and for the most part, trouble free in that application.
One version of Mach 3 should suit all applications. It doesn't. Some versions are just better suited to specific needs. They should all be generic. They aren't. I will continue to use Mach 3 after 4 rolls out. I will, however buy 4 for my mill.
By that time, I expect to have a PDB and ATC on it and the extra stabilized Mach would simply be a better choice over the long haul.

It is a machine control and not an operating system BTW.

[SCzEngrgGroup]

Just because Mach3 works well for you, doesn't mean it works well for everyone. Anecdotal evidence is not real evidence. That's exactly the problem - for some users, it simply does random, unpredictable things. That's the very basis of Tormach decision to use a specific version that they have test enough to be deemed "stable enough" for their products.

Just yesterday, I went to use one of my machines - the only one that still runs Mach3 - and found it had somehow managed to spontaneously change configuration so it powered up in metric mode, and absolute IJK mode - two settings I've never once used on ANY machine I've ever run. Needless to say, my code did not run well like that. That is exactly the kind of random flakiness I experienced for YEARS with Mach3, and the reason I abandoned it several years ago.

Since ditching Mach3, I've not had one, single unexplained "glitch" on any of my machines. Mach3 works fine for many people, and if you're one of those, you should count your blessings, but don't think that means anyone who has problems is doing something wrong. It works very poorly for some others (I was always "lucky" enough to be one of those chosen few....), and there is an entire spectrum in-between the two extremes, from those who see mostly good behavior with occasional hiccups, to those who see many problems with occasional brief periods of good behavior.

Mach4 will, hopefully, shift that spectrum profoundly towards better behavior for everyone, based on having an actual defined system architecture, rather than being a huge patchwork quilt of often badly designed modules held together with chewing gum and bailing wire. It almost can't help but be a quantum improvement over Mach3.

Regards,
Ray L.

[sansbury]

On the Tormach, I can manually zero all three axes using an old school spinning edge finder in less time than it takes on those high end machines. With a Haimer probe, it would be even faster (30 seconds or so).

My probing routine takes well under 30 seconds and has zero chance of error like misplacing a decimal point keying in an offset.

[LeeWay]

Probing is lot like sex. Everyone has their own way. Please be aware that one false move can mess you up