Joincopy is acting weird in the latest SR:
Version 7 SR8
(7.8.21196.5001, 2021-07-15)
It’s occasionally acting like Join, overriding LockChildren=Yes and joining history enabled lines/curves into a polyline/polycrv. I’m having a hard time creating exact duplication steps. It seems to happen at random.
Also, you still cannot specify the curve seam and direction of Joincopy crvs. This tool will be invaluable for working with patterns once the seam can be controlled.
Hi @EricM,
JoinCopy is an undocumented test command. I’m not sure it’s something you should be using in production…
– Dale
Don’t I know it. MergeCrv is documented, but it has more issues than Joincopy. I’ve been begging for someone to look at either of them, but to no avail.
They both came from my feature request that somehow became two youtracks:
https://mcneel.myjetbrains.com/youtrack/issue/RH-45817
https://mcneel.myjetbrains.com/youtrack/issue/RH-56306
It’s been a while but If I remember correctly I think the “randomness” of the output stems from whether or not the curves were picked and then the command is ran vs running the command and then selecting the curves. I seem to recall pre-picking and post-picking producing different results.
Can you elaborate more on the issues you’re having with MergeCrv over JoinCopy?
No, it’s random. It’s not like _CurveThroughPt which does rely on the order of point selection. _CurveThroughPt’s seam is chosen at random (not first or last point selected), but the seam will not migrate or switch directions.
Complete info with demonstration video showing the issue is here:
Exactly why MergeCrv < Joincopy can be found here:
So your primary issue is with closed curves changing seam point location and the direction of the resulting curve getting flipped.
Your issue with mergcrv over joincopy is the with the resulting shape of the curve being changed.
So it sounds like you like joincopy due to the resulting shape but need it to retain some fixed curve seam and direction.
Is that a fairly correct assessment or am I missing more? There were a lot of different conversations and extra info in the links and videos.
Yes. Controlling the seam location and direction is critical. Downstream workflows use ! -Loft Pause Natural Enter Type=Loose Enter.
Yes. MergeCrv is inferior because you can’t create/control the transitions. Joincopy lets you create/control your own Fillet/Chamfer/BlendCrv.
I’ve tried both, but have been using Joincopy exclusively and many times per day since December 2019.
Also, I just realized _CurveThroughPt’s seam/dir isn’t random. I was able to reverse engineer it and understand it, but juniors won’t.
Misaligning seams is something I commonly see new people struggle with, so I started a separate thread to request it as a feature:
Hi @EricM,
Since the projected release target of this feature is 8.x, I’ve moved the conversation to the Serengeti category.
– Dale
Please don’t @dale.
This was a V5/6 feature request.
The issue was raised multiple times during the V7 WIP.
It got worse with SR8.
This is a major boost to productivity, especially with History changes in SR8.
FWIW, I just posted a video to my 5k subs showing the awesome things we can do with the History changes in SR8, and a solid 20 mins ends up being about stupid Joincopy.
It’s definitely worse in SR8. I’ve developed an acute sense of when the seam would flip. I tried to demo it at the 23:00 mark and it wouldn’t. Then at the 37:00 mark I’m caught off guard by a flipping for no reason. So frustrating.