For those who have tested 2027...

[kdg]

2022 since MGS turdification in 2023.  In 3x or 3+1, even 3+2, there is nothing new that provides benefit, just more clicks and more bugs.  Every version, I try to use it, check out what's new, try to just deal with it, but they always break more stuff than they fix, and it always seems like the new stuff is released half-baked, and doesn't get debugged for a year or two, by which time they've added/broken more stuff that just disrupts workflow with no upside.

Same, 2022 FTW. I'm well aware of its bugs, & know how to work around them.
I'm not interested in beta testing the new releases.

[TSmcam]

I can't believe I am going to say this after all the years on the bleeding edge...

Am I the only one considering sticking with 2026?

I sense a disturbance in the force... LOL

[mowens]

... and it always seems like the new stuff is released half-baked, and doesn't get debugged for a year or two, by which time they've added/broken more stuff that just disrupts workflow with no upside.

Boy that sounds familiar.

[TSmcam]

Boy that sounds familiar.

When I was using it, I always joked that it seemed like they had a fixed number of features. And when they introduced a new one, they broke or took away an old one

[Jeff]

I'm still on 2025 as my daily driver, I skipped 2024, and am skipping 2026.
2026 offers nothing. It should have been a free release to current customers.
I have a funny feeling that maintenance will jump once 2027 hits, and it will NOT be ready to be released.
So I'm prepared to stay with 2025 until 2028.

[gcode]

My personal opinion is that the people calling the shots at Sandvik don't know the first thing about
the CAM industry.
They are looking at Fusion and thinking, "we need to be doing that", and it's off to the races.
At this point the most compelling reason to stay on maintenance is for post processor support.

I remember some years back before Sandvik bought CNC, their salesmen came around with a CAM app on a tablet.
It could do dynamic toolpaths on simple shapes and IIRR it could also do Prime lathe toolpaths.
These guys were really proud of it, though my initial reaction was "How cute, WFT am I going to do with that?"

They kept coming around with this thing for nearly a year.
One time they even brought one of the developers because I clearly did not understand the magnificence of their creation.

Now they own Mastercam and the best part of their website (tool assembly construction) is locked down
with a seriously overpriced subscription.

[MIL-TFP-41]

The milling team has brought some significant improvements to holemaking solids selection time lag for 2026 R2- hoping to hear some good feedback from users when the time comes.

So making it work in a timely manner like it did in 2025 is now an improvement? Sorry, not buying that. Sounds more like the team un-fucked it up. I've sent my sample file in. What would regenerate in seconds in 2025 is unusable in 2026.

[CADCAM396]

"un-fucked it up"
Can borrow that phrase for a day. I like it.

[mowens]

When I was using it, I always joked that it seemed like they had a fixed number of features. And when they introduced a new one, they broke or took away an old one

I have bitched about this before. My programming software worked fine until the original author retired. Then the people who took over for him decided to rewrite the entire software suite. It was originally written in Fortran. They decided to rewrite it in Python. The problem is, they didn't have a detailed understanding of the software before they rewrote it. Now, it works fine for test panels and simple surfaces. The problem is that, as the technology for this process improves, the surfaces we use are becoming increasingly complex. The software can't keep up. When I started with the original software it was at build 89 I believe. The last one I had before the rewrite was build 426. And it was pretty solid. Now, it's just one bug after another. They fix one and two more show up. There are times when half of my programming effort is just trying to work around the latest bug fix that caused other issues. Pretty frustrating. The software was originally developed in conjunction with one of the OEM machine builders. Now that company is exploring other programming software options.

[Dylan Gondyke]

So making it work in a timely manner like it did in 2025 is now an improvement? Sorry, not buying that. Sounds more like the team un-fucked it up. I've sent my sample file in. What would regenerate in seconds in 2025 is unusable in 2026.

No. The time lag on some CAD files was a regression and clearly unacceptable in 2026. The team has been working to resolve these things and improve the gen times versus 2026 as released. Thank you for sending files in and providing us with different test cases.

Seconds went to minutes if you had certain solid body properties, features, complexities, etc. 2026 R2 will bring it back to seconds.

[Dylan Gondyke]

You know we like it crunchy around these parts. Can you give us the 101 on the internal differences between driving a path from a surface, vs a solid? I know it to say surfaces are "pure math" (easier to manipulate and run calculations on, for some reason), but why are solids not? Are faces made of a bunch of smaller units in the backend, are we dealing with a tessellated bunch of triangles when a toolpath looks at a solid vs a user's eyeballs? Inquiring minds want to know.

It's probably conceptually simpler than you're thinking. When referencing surfaces or wireframe for something in a toolpath, you're referencing a single entity that doesn't have any complex association, history, or metadata. When referencing a solid, you have all of the benefits of a solid (History, or perhaps direct edits/push pull/etc, but you have to reasonably programmatically check against everything that might be going on with that solid if you are regenerating a path. Or if you changed the solid, the path has to figure out if you changed anything that would affect it. Certain features on solid bodies such as physically modeled threads don't add much "weight" (file size) to the solid, but are hideously expensive to check against if you have to peek at the solid and say "has anything I care about changed? What about that tab on the far side of part?" Etc.

When a user selects a face of a solid, they (rightly) expect a toolpath to be able to infer some information to do things like, take the vector of the hole, figure out the direction of a hole properly if it's a blind hole so you don't drill from the underside, not crash the tool/holder against other parts of the same solid if we can avoid it, etc. That means looking at information from that solid other than just the thing you directly selected. Optimizing what we look at so that we catch the stuff that matters, and ignore the stuff that doesn't, is the challenge here. Falling back to wireframe or surfaces is an easy one for power users because you know exactly how to drive the software, know exactly what you want, and don't want the potential performance hit of solids because the solid isn't providing intelligence in your case- YOU are by manually twiddling each knob and overseeing each entry/transition as a programmer.

[CADCAM396]

but but
this 6 (ty) year old just wants to push a button and it works.

LOL
thanks for hangin with us Dylan

[megatronprime]

When referencing a solid, you have all of the benefits of a solid (History, or perhaps direct edits/push pull/etc, but you have to reasonably programmatically check against everything that might be going on with that solid if you are regenerating a path.

Is the Solid History really used in the toolpath?
That would surprise me, because the toolpaths usually work even if u remove the history from the Solid.

[Jeff]

I have a solution, just change how Mastercam looks at solids.

I should be a software programmer, this stuff is easy!
 

[Dylan Gondyke]

Is the Solid History really used in the toolpath?
That would surprise me, because the toolpaths usually work even if u remove the history from the Solid.

It's one of the things that can trip the flag that says, "Hey something happened to this solid, you need to re-examine it and ensure that nothing you cared about for your toolpath calculation changed."