I suppose it is possible. There is another request for that seam to be output as part of the API call to Boolean, and I will probably do that.
A challenge will be how to fit all of these options into the modifier panel without making it harder to use for typical use cases or confusing for newer users.
Yes, if more Boolean modes were added, I think instead of toggle-able buttons like there are now, a drop down menu would be better.
Speaking of new Boolean modes, would it be possible to add a knife mode to the modifier as well? There’s an option in edit mode (called ‘Intersect (knife)’), but it’s not available in the modifier. Here’s an image showing what I mean with a beveled cube:
I am strongly considering adding the Knife operations to the Boolean modifier. Another possibility would be to make a new “Knife” modifier, since the Knife operations don’t have much to do with the term “boolean”. What do people think?
It seems like a nice idea, but could you explain a little bit more what the functions of that modifier would be, and how it would work?
When I hear that, I think of houdini being able to pull out the knife tool when clicking on the modifier, allowing you to perform any number of freeform cuts on your mesh that all belong to the modifier. Is that even possible in blender?
I would love to have an improved knife tool for Blender. And a unique modifier would be even better.
In short, yes please:-)
I’d say split can be a boolean “mode”, while I’d like the knife to be in a separate modifier. Options can be to set weights to the cut geometry as: edgeloops, inside or outside
Knife modifier sounds great. Feels like the resulting output should be able to be used for shutlines or other swept/rail geometry (I don’t know the Blender terms, so I’m using CAD terminology).
yes, it is what I meant 
Hi Howard - just out of curiosity, are there any plans to make the “precise mode” in the boolean modifier as fast as the “fast mode”? so basically having only one mode instead of two. Or do you think this will be always a limitation of the new boolean modifier?
cheers
It will always be somewhat slower because of the arithmetic it uses. The question is: can I make it close enough that people wouldn’t mind? I don’t know. I haven’t finished trying to speed the new one up, so I may succeed in getting it “close enough”, but I can’t predict for sure whether this will happen.
At any rate, it seemed prudent to keep the old one around for at least a few releases, in case people were depending on the behavior of that old one, or in case bugs in the new one show up. After the new one has been thoroughly stress tested, and found to fill all the needs of users (including acceptable speed), I think it might be considered to retire the old one.
Would like to chime in here. I am an architect and often times we want to show expressive 3d section cuts of large projects and parts of towns. For this purpose, the old boolean suffices for the most part and (albeit also slower than i would like). Whilst the accuracy of the new boolean modifier will definitely come in handy many times, it is also important to consider the performance of the old boolean for larger projects. As such, I would suggest that it remains intact until the new boolean could perform as fast as the old system.
This might be interesting to you:
https://developer.blender.org/D8954
(A bisect modifier) Combined with modifiers on collections I think this would be a game-changer for your use-case. Same for boolean actually.
Hello @Howard_Trickey , I’d like to point your attention to this idea, which is as easy as useful:
what do you think?
Re bounding box as option for boolean operand.
My reaction is: this seems kind of niche, and one that isn’t too awful to just deal with by hand. It would be fairly easy to implement, but would it be worth adding more complication to the interface, documentation, etc? I don’t know … I’d be happy to hear other people’s opinion on this.
I can see it being useful for archviz (quickly adding windows, doors, staircases) but other than that I don’t see much use case for this.
You’re right i guess. Also, now with whole collections booleans…
This is useful, for some cases for a true section, but often times, we wish to be more expressive and cut out just a part of the model, still in 3d
This type of fairly simple geometry (lots of solidify modifiers) works pretty well with the fast booleans, but still takes quite a bit of time.
@Howard_Trickey I’m not sure if these are known issues (I haven’t followed the entire thread), but I’ve noticed that with not-perfect geometry the new exact boolean makes some mistakes where it fills the entire boolean object. I guess the geometry has been non-manifold in most of those cases, like the suzanne in this quick example:
Ticking the “self” option helps a little bit (the latter half of the video), but there’s still a few glitchy areas when moving the cube. I guess booleans are never completely perfect, but I was wondering if there’d be ways to at least minimize this “filling up” thing. The old boolean makes even more mistakes, but it usually doesn’t have this filling up problem. I have to do a lot of animated cut sections for product demos etc, and we get all kinds of dirty .stl geometry to work with, so this issue comes up quite often 
Again, sorry if these are known issues, and thanks for the awesome work on the new boolean!
