Hi Psygorn,

re:
> Second, is ACIS library used only for fillets?

Yes at the moment only for fillets, but once a few more kinks are worked out I expect to use it in some other areas as well like chamfer, shell and offset, and face editing.

> (sometimes when I use ACIS to do the fillets I get torn faces but MOI says the
> geometry is solid & when I do the fillet with Solids++ I don't get that torn face) it
> is not a big issue at all, I just wanted to ask you about it out of curiosity. :-)

Yes if it looks like the thing Pilou shows and it goes away if you save and reload the .3dm file then it's probably that same display mesher bug.

It seems that ACIS tends to create trimming boundaries that sometimes run across some kind of bug in the display mesher.

I haven't quite gotten that figured out yet,

For the time being when you run across it, just try saving it to a .3dm file and reload it back in again.

- Michael

Hi Michael,

Thank you for your reply :-)

> Yes if it looks like the thing Pilou shows and it goes away if you save and reload the .3dm file then it's probably that same display mesher bug.

I did export my mesh as a .stp file and looked it up inside 123Desig there was not any issue whatsoever :-)

Best,
Psygorn

hi michael,

not sure if you can answer this or not. however, i was curious from older posts of yours. you had previously mentioned you tried either parasolid or acis (not sure which, maybe both). then you tried kubotek kosmos. you said you couldn't switch to parasolid/acis due to tight tolerance issues or something like that. also, kosmos wasn't developed enough. so i was curious what changed to where you could use acis now. also, how did you pick between acis or parasolid. i've seen interesting things about others decision making process. so i was curious about yours. in any event, i'm really glad you are making this change. my biggest headache with rhino is the terrible geometry kernel and all it's failures, quirks, etc... it's an endless source of frustration and takes a lot of my time. i have to pound on it forever, to beat it into submission, and it's always a stressful process. even then, i don't trust it's fillet engine at all. it seems to limit the radius values to much smaller than realistically could be achieved.

anthony

Hi anthony, I had previously looked at Parasolid, Kubotek KCM, and C3D.

The difficulty I had with Parasolid was that it was very strict about what geometry it would accept. It would refuse to create a closed surface if it was something like 1 degree off from being smooth. This was a problem because of the particular way I wanted to integrate it which is in some gradual stages instead of everything changing over all at once.

Also the other problem with it is that it's expensive and very "enterprise" big business oriented.

KCM and C3D were offering much friendlier terms for small businesses so those could be a better fit. I waited on KCM for a pretty long time but it seems to have stalled out. Next, just as I was about to dig into C3D the Russia/Ukraine war broke out and C3D is developed by a Russian company. It's pretty much impossible to do business with a Russian company now so that one was off the table too.

That left ACIS as the last viable one to try. I had not previously tried it because it is also pretty big business oriented although not to the same extent as Parasolid.

In my initial tests I found that I was able to round trip geometry into and back out of it again very well so I decided to move forward using it.

- Michael

awesome. thanks for that info. i've been really curious to know. yeah parasolid and acis do give off a big business vibe for sure. that's why i stay away from their software as well. one interesting thing is parasolid and acis both started from the same group of guys from the uk. so it's rather hard to know what exactly are the pros and cons of each. agree with you on the russian kernel as well. even before the war that sort of thing is a concern for me. i imagine they have the same feelings, so that's why they have their own kernel.

i hope the acis kernel works out. this really makes moi stand out from rhino a lot more now. one huge disappointment with rhino is they keep adding stuff i have no interest in and keep leaving stuff broke or working poorly. the development time is also way too slow. if they would at least sell features in packages it would help a lot. so a lower price for just nurbs, more for subd, more for rendering, etc... they just seem happy to make one huge blob.