| |
(Note that I’m sending followups to the lugnet.cad.dev.org.ldraw newsgroup
alone, because that seems to be the appropriate group, but I think the audience
of the lugnet.cad.dat.parts group should see this.)
In the process of testing my next major LDView version, I have discovered that
at least one (and probably more than one) official part is both BFC-certified
and contains bow-tie quads. All things considered, this isn’t too surprising,
but I would like to suggest that something be done to attempt to prevent this in
the future.
According to the L3P -check messages page
here, any “Bad vertex
sequence” warnings in a BFC-certified file should result in a hold. The reason
for this is that there’s no “right” way to reorder the vertices, and half of the
possible ways will end up with the quad facing one direction while the other
half will have the quad facing the other direction. (For reference, I found the
problem in the first place because LDView sometimes reorders in a different
order from L3Lab.)
Unfortunately, this “Bow-tie in a BFC-certified file” check isn’t enforced
during the automated submission check. Any chance of getting that added? I
know the tracker already has a fair list of planned future features, and I
understand the realities of volunteer software development, so I’ll understand
if it doesn’t get done for a while.
Unfortunately, it’s very easy to overlook this problem when reviewing a part.
Additionally, the part review FAQ
here states that bow-tie
quads will show up in blue when mytest6 is enabled. This doesn’t appear to be
the case. Presumably anyone who has read that section assumes that the lack of
blue quads indicates a lack of bow-tie quads. That FAQ question should really
be updated.
On a side note, I’ve added code to LDView that will disable BFC support in any
file that contains bow-tie quads (as well as generating an error). So it’s not
like the problem can’t be worked around, but it does go against the official
review guidelines. (Yes, I could have made it so only those particular quads
didn’t get BFC’d, but that would have been a lot more work, and I didn’t feel it
was worth it.) So the the first official LDView with BFC support won’t be
hampered too much by this problem.
--Travis
|
|
Message has 1 Reply:
 | | Re: Bow-tie quads in BFC-certified parts
|
| (...) It's funny how quickly I can forget the details of a complex task. For instance, I totally forgot about the NOCLIP BFC command (which can disable clipping for specific sub-sections of a file). Mind you, I implemented it, and believe that it (...) (20 years ago, 10-Sep-04, to lugnet.cad.dev.org.ldraw, FTX)
|
4 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
