PTC has sent back a Repaired file.
They were not able to determine a root cause... not without having everything that lead up to the issue (trail files, exact configs, historic files etc).
So in the end, I have no idea how this can be avoided in the future but I have a working file now (without the Ambiguos sketch)
UPDATES! I had been asked to dig deeper at work, so I have...
I was able to find exactly what Rev/Ver the corruption occured. Much to my anoyance, I was not able to do this before because I had gone backwards through Rev's and Ver's which made it so the corruption was always there! This time I started at the bottom working my way up... and suddenly the corruption wasn't there?! Anyway - I found it... still no idea why it is corrupted - it just is.
However! because I have a working version, I could now walk through to where the original issue (and title of this thread) occurs! "Section regeneration failed. Ambiguous sketch." It is exactly what I did not expect - Caused by a Draft?
There is a draft on a circular surface that is parted both sides (think cylider parted both side of it's middle)... When I part one direction, no issue. When I part the other direction, no issue. When I part one half, no issue. When I part the other half, no issue. But when I split it both sides - issue. Even when I do it in two seperate drafts, whichever comes second causes the Ambiguous sketch - so it's not just one side, it's a combination of the two sides.
Well - it works - it just gives the Ambiguous sketch message... but why?
I would let PTC answer that one. Thank you for putting in the effort and reportinig it back. This can be very useful to someone else some day. Sections have always been a problem in Pro|E for the silliest things.
Please submit the file again with the updated information. This really needs to be looked at for future releases.
Error message ‘Ambiguous sketch’ issue is corrected in (1) M230 datecode of Creo Elements/Pro 5.0 and in (2) M110 datecode of Creo Parametric 2.0
These two releases will tentatively be available for customers by first week of April and first week of May respectively. I tested this issue in these two datecodes (M230 and M110 are available internally for Tech Support for testing purposes) and found that the error message does not show up (C11263145)
The issue that was crashing ProE 5.0 (and recently Creo 2.0) was/is an unrelated corruption that can not be corrected by a user and must be done at the coding level by PTC. They could not explain how/why it happened... so no idea how to aviod it from happening again. But it seems unlikely.
So, we're at datecode M230 of WF5/creo now? Wow, and to think we just struggled to "upgrade" to M180.....?
Thanks for the update, Daniel.
Cross-sections have always been buggy as I stated before.
Maybe now they actually fixed it.
AD, It's not a cross-section failure.
The software is blaming a sketched section that is used to create a feature, though it is likely that it is a failed pointer that is the actual culprit. It shows up as a symptom following a draft feature. In creating the draft feature PTC is probably not getting the correct reference to evaluate and the failure mode is to generate a message that something is wrong with the referenced item.