Subject:
|
Re: IMPORTANT - [ldraw.org] Official Model Repository
|
Newsgroups:
|
lugnet.cad.dev.org.ldraw
|
Date:
|
Sat, 12 Jun 1999 22:05:26 GMT
|
Viewed:
|
3118 times
|
| |
|
|
In lugnet.cad.dev.org.ldraw, tim@zacktron.com (Tim Courtney) writes:
> [...]
> > > I think that the way something like the Mos Espa set should be
> > > done is
> > > like this(I don't have the set so this is just an example)
> > > S7171-99.dat The complete set
> > > S7171a.dat Anakin's Podracer
> > > S7171a1.dat A's P cockpit section
> > > S7171a2.dat A's P engine front
> > > S7171a3.dat A's P engine back
> > > S7171a4.dat A's P Pit Droid
> > > S7171a5.dat A's P Pit
> > > etc. for the rest
> >
> > Under that scheme, how would you place the three models from set #1974? How
> > about the two models from #8205? How about the two different #8857's? Or
> > the two different #6083's?
> >
> > See how that approach quickly crumbles?
> >
> > If the year is used to make a unique identifier for a set, then the sub-
> > models had better be tacked on in addition to the year, otherwise you ruin
> > the namespace by making it ambiguous again if you drop the year for sub-
> > models. If you use a letter instead of the year to make a unique identifier
> > for a set, then you run into problems with the ordering because it's not
> > always known which one came first.
>
> Give an example of tacking submodels on in addition to the year. I am not
> clear what you mean by that.
First, tacking on submodels in addition to the year:
S8857-80.dat 8857 Motorcycles
S8857-80-a.dat First model
S8857-80-b.dat Second model
S8857-80-c.dat Third model
S8857-93.dat 8857 Street Chopper
S8857-93-a.dat First model
S8857-93-b.dat Second model
That's a clean, orthogonal, workable, forward-thinking namespace.
And now, -not- tacking on submodels in addition to the year (as proposed
three articles up the tree):
S8857-80.dat 8857 Motorcycles
S8857a.dat First model
S8857b.dat Second model
S8857c.dat Third model
S8857-93.dat 8857 Street Chopper
S8857d.dat First model
S8857e.dat Second model
Where that approach crumbles is that the models of the '93 8857 get un-
naturally labeled as 'd' & 'e' rather than naturally named 'a' & 'b' inside
a new namespace. That's not clean at all.
Bram's approach is much cleaner.
> Also, with the 8.3 system it is not possible
> to come up with a unique identifier for EVERY set.
As Charlie Babbit's brother would say, "8.3 sucks." :-)
> [...]
> The only way of accomplishing this unique identifier (and even this won't
> do it for all) is to drop the S in the beginning. But that will make a lot
> of part authors unhappy. It will probably also make me unhappy because all
> the sets won't fall under S in my LDAO file manager anymore. We need to
> accept the miniscule limitations of this 8.3 system which we still use and
> also accept the fact that our big friendly TLG will find a way to make a
> set that goes outside of these rules.
There's gotta be a clean way to do it with subdirectories and still be 8.3,
as Bram was getting at.
--Todd
|
|
Message is in Reply To:
94 Messages in This Thread:
 
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
