Also, to remove doubt, these GUI tears could occur even after a reboot, a log out snd log back in, whether the temp folder was junky or manually clean.
The common theme seems to be Ctrl+drag of objects, whether they are curves or solids. OS, graphics, laptop manufacturers, and session time or file size, old or new, no apparent bearing. Not at-will reproducible. I can go weeks without it happening, then, out of the blue, it strikes.
Occasionally, I get a sixth sense that I will happen within minutes, and 40% of the time it does. Sometimes, I'll CAD on for 6 hours without an incremental save, with no adverse events.
The saving grace of the GUI tear/collapse is that CTRL+S remains available. The catch is one MUST REMEMBER to get to the file explorer and make a pre-save copy before trying to dave or abandon the file. Why, you never know ehen your file is corrupted or by what, snd don't want to save a corrupted one over a good one.
It would be nice of instead ofusers having to remember to immediately save a new session file (fresh or revisited) as a new file before getting far along in edits, the app instead makes that an option, or at least aloends the parent file adjacent to the base file. Live archiving could help, but if bad files are archived or if innreal time a backup isn't made before a file is accessed, well, some will say that's a user problem. But, in the early windows days, some devs did do that on behalf of users. Then, MS laid down some edict on user expectations, I feel. I believe some devs donstill peovide that functionality for high-risk/high stakes app operations.
NOW...
Example of saving bad files over good files.
I in 2024 or late 2023 had several days and 20+ corrupted files across 3 to 5 different ship files I was hopping around in. One with geometry I wanted to use in several others was tainted, and because importing was the only way to get content with associativity and layer mames retained, I unknowingly imported corruption.
It was devastating and deeply deflating to every single file for hours of backtracking turning out "Unable to open file at location...(some cryptic numerical address)". Ultimately, I'd later be bitten again by the same taint, and it out me on a long session of hunting through thousands of files manually yo see just how far back the damage extended. It was a lot for some, nil for other files. Fortunately, I have a neurotic habit of doing File SaveAs to multiple names just frequently enough that a losss is not a total loss, especially if it is in the recent few minutes letting me recreate something.
Strangely, the root cause files that *could* be opened behaved fine until the contents of them were imported into and destroyed other files. The bad-donor files do remain usable, but I have to remember not to use their actual SharkCAD file type as...
It took a few months of restrospection and a chance hunch to try doing exports as Concepts Unlimited formatting of the.
The tainted files then seem to be stripped of some piggybacking mischief maker.
Nicely, the CU apparently doesn't strip attributes from even SFC 12.
I had trepidation saying this because I'm creating the eisk that the CU file time could be removed or without documentation be changed.
Regards,
ZLC
.....
Clarification: crawled through dozens of folders that contained 1,000s of files, and manually opened and closed dozens to up to 100, maybe as many as 200+. I didn't open and close 1,000s. But, it sure felt like it.
Edited by user Thursday, July 23, 2026 4:57:35 PM(UTC)
| Reason: See very last para under .....