Subject:
|
Re: DAT voting page up
|
Newsgroups:
|
lugnet.cad.dev
|
Date:
|
Mon, 1 Mar 1999 14:57:59 GMT
|
Viewed:
|
2118 times
|
| |
|
|
On Fri, 26 Feb 1999 21:58:59 GMT, "Duane Hess" <DNJHESS@ZDNETMAIL.COM>
wrote:
> Since I am not an LDraw part developer (although I did try my hand at one
> part), I will add my $0.02 to the fracas and be done with it. This outlook
> is based upon my knowledge of engineering documentation systems in the real
> world. With that said, my opinion is as follows:
>
> Steve Bliss writes:
> > It's not an update. It's an A/B version thing--Tile 1x1 Old Style, Tile
> > 1x1 New Style.
>
>
> I agree. In the real world parts change all of the time. It is better to have
> a new part number for the new piece, if both pieces are currently in use.
> That avoids having to track what revision is used where. However, since we
> only have one number to work with, I would suggest using 3070a for the
> old part (without finger lip) and 3070b for the new part (with finger lip).
> Part 3062 is a perfect example of this method put to practical use.
In this case, all the tile elements probably have new, 5-digit part
numbers. It would be appropriate to use these numbers for new versions
of the tile.
In general, if TLG modifies a part but retains the original part number,
appending a letter to retain the original part and differentiate the
part-files in LDraw is appropriate.
> This may get confusing where sub-parts of an assembly are already numbered
> as such. Take for example Hinge Plate 3149. It's sub-parts are broken up
> into 3149a, 3149b, and 3149c. I would suggest discontinuing this practice.
> Would LDraw accept a part number such as 3149-1, 3149-2, etc? If so, I would
> recommend that this method be used. That creates a distinction between
> a sub-part and a part revision. It would also allow for 3149-1a if that part
> were to ever be changed by TLG.
TLG assigns part numbers to all sub-parts. Going forward, the plan for
LDraw's part library is to use actual part numbers (when possible) for
all sub-parts. In this particular example, 3149 is the number for the
base subpart, not the entire hinge. (my personal opinion: I'd rather
leave 3149 alone, because the 3149 is useful to people who look at parts
to find their LDraw number).
> As for colors, I do not believe that it is productive to assign individual
> part numbers for different colored parts (specifically the dingy in
> question). My reasoning behind this, is that the LDraw operator has control
> over the color when it is inserted. If he wanted a purple dingy, he could
> have one. LDraw is more about fit, form and function and less about spectral
> appearance which brings me to another topic being discussed here.
I agree. Also, we don't know what colors TLG may use for parts in the
future. If a part is hard-coded, and a new color-version is released,
then a change to the part-library is required (whether that change is
removing the hard-coded color or releasing another hard-colored part
doesn't matter--it's still a required change).
> I feel that all parts created for LDraw should resemble their physical
> counterpart as close as is reasonably possible. That means that they
> should be created with all visible surfaces in their proper places. The part
> author should not settle for "close enough, needs work." How is the part
> author to know what creative way the part my be used in real life? If
> shortcuts are taken, the part may not be able to be used in the same
> orientation in LDraw. TLG has tolerances for the amount of dimensional
> variance allowed before a part is rejected. I see no reason why the L-Cad
> listserv shouldn't adopt at least a 1 LDU tolerance on all parts submitted.
Generally, the accepted tolerance is between 0LDU and 1LDU. This is
because fractional errors are very noticeable, when sections of a part
don't match up.
Steve
|
|
Message is in Reply To:
 | | Re: DAT voting page up
|
| Since I am not an LDraw part developer (although I did try my hand at one part), I will add my $0.02 to the fracas and be done with it. This outlook is based upon my knowledge of engineering documentation systems in the real world. With that said, (...) (25 years ago, 26-Feb-99, to lugnet.cad.dev)
|
97 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
