 | | | | | | |
| |
|
In lugnet.cad.dat.parts, Matthew J. Chiles <mattchiles@gorge.net> wrote:
> You know, this piece typifies the problem with the process and why no
> new parts get published, at least from my view.
>
> The burden of detail required for approval is too onerous. In this
> particular part the ice cream is fine either way - as the part
> actually has it in real production, or as it "should be". The part
> certainly should NOT be held for this reason.
I agree that it shouldn't be held for this reason. As for the original
question, I think that they should be modeled in the way that it appears they
were "intended" to be if and only if at least one of the various copies of the
part that show up in the real world matches the assumed "intended" look. If
they all look the same in real life, and they seem to be funny, then that's
tough. If, on the other hand, there's wide variation in the alignment of the
pattern from one part to the next, it seems reasonable to model the pattern in
the way that looks the best. So for this part in particular, if there are
copies that aren't misaligned, then I think it should be modeled that way, but I
DON'T think that's grounds for a hold vote. Add a No-Vote with your concerns,
sure, but not a Hold. If all copies seem misaligned, then perhaps it's not
actually misaligned, but looks just like it was intended to look. For this part
in particular, I looked at the DAT file carefully, and I'm not convinced that it
wasn't intended to look just the way it does.
> For the sake of L-Draw continuing as a useful platform parts need to
> be pushed through and approved. Many parts approved long ago don't
> meet todays strict standards, and they are slowly being updated as
> authors revisit them. But for all the time since they are created
> they have been in the system and they are useful. To continue to keep
> the L-Draw system useful, parts need to get in the system so people
> can use them as they are "decently good".
I agree, with a few caveats. First of all, I think there should be a new
official header comment that basically says, "ok, but could use some work".
Secondly, reviewers need to be very careful about the orientation and
positioning of parts. There are plenty of official parts that don't have good
origins or orientations, and they CANNOT be fixed now, because they're
official.
> If not all quads are co-planer, so what, that isn't going to impact
> anything. Invisible gaps between planes, or unimportant undersides
> not detailed 100% correct should not be reason to hold up parts. My
> personal pet peeve is BFCing - I can see the value, but can't see
> holding a part up for it, especially with the high speed, high memory
> computers we use today where a few added back-faces are not going to
> make an impact.
The only reason a part should be held due to BFC is if the part is
BFC-certified, but the BFC code is wrong. If there are errors in the BFC-ing of
a part, it definitely needs to be held, because otherwise the flipped polygons
will be invisible on BFC-compliant renderers.
The part reviewing documentation specifically states the BFC-certification isn't
required, though. If I authored a part (mind you, I don't), and somebody put a
hold on it because it wasn't BFC-certified, I'd send an email to them politely
asking them to remove the hold and explaining that BFC certification wasn't a
requirement. If that failed to produce a result, I'd send an email to the
admins asking THEM to remove the hold.
--Travis
| | | | | | | | | | | | | | | | |
| |
|
In lugnet.cad.dat.parts, Travis Cobbs wrote:
> In lugnet.cad.dat.parts, Matthew J. Chiles <mattchiles@gorge.net> wrote:
> > You know, this piece typifies the problem with the process and why no
> > new parts get published, at least from my view.
> >
> > The burden of detail required for approval is too onerous. In this
> > particular part the ice cream is fine either way - as the part
> > actually has it in real production, or as it "should be". The part
> > certainly should NOT be held for this reason.
>
> I agree that it shouldn't be held for this reason. As for the original
> question, I think that they should be modeled in the way that it appears they
> were "intended" to be if and only if at least one of the various copies of the
> part that show up in the real world matches the assumed "intended" look. If
> they all look the same in real life, and they seem to be funny, then that's
> tough.
Yeah, tell me about it! Me and a friend made a mock-up of the sign for the
Datsville post office:
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=192884
The low-res picture shows just a little of the flaw, but the letter 'S' is just
so wrong. We decided to let it reach below the other three letters, but not as
much as at the actual LEGO part.
When I made my first 3005-letters, I made them a compromise between accuracy and
visibility. It was more important that the letters were readable in as small
scale as possible than that they were true to the originals. Do you find this
philosophy shocking? ;) That was in the days when LDraw was meant to produce
readable instructions. Today, I don't know what's become of LDraw. A playground
for perfectionists, where nothing or very little passes, maybe?
> I agree, with a few caveats. First of all, I think there should be a new
> official header comment that basically says, "ok, but could use some work".
Why not keep to the tradition of ending the description with "(needs work)"?
Eg. "Minifig Flipper (needs work)". IMHO, there are already far too many useless
official header comment and META statement to keep track on.
> Secondly, reviewers need to be very careful about the orientation and
> positioning of parts. There are plenty of official parts that don't have good
> origins or orientations, and they CANNOT be fixed now, because they're
> official.
I fully agree. This is a very important issue when revising parts.
> The only reason a part should be held due to BFC is if the part is
> BFC-certified, but the BFC code is wrong. If there are errors in the BFC-ing of
> a part, it definitely needs to be held, because otherwise the flipped polygons
> will be invisible on BFC-compliant renderers.
True.
/Tore
| | | | | | | | | | | | | | | | | | | | | | |
| |
|
On Tue, 16 Jan 2007 11:06:31 GMT, you wrote:
> When I made my first 3005-letters, I made them a compromise between accuracy
> and
> visibility. It was more important that the letters were readable in as small
> scale as possible than that they were true to the originals. Do you find this
> philosophy shocking? ;) That was in the days when LDraw was meant to produce
> readable instructions. Today, I don't know what's become of LDraw. A playground
> for perfectionists, where nothing or very little passes, maybe?
"A playground for perfectionists"
Excellent summary of what I am trying to say. Most of us are not
perfectionists even if we would like to be, and we don't have time to
be perfectionists. But we do want useable parts.
There is a step below perfection that is "good enough". If L-Draw
continues to strive for perfection at the expense of all else, it will
be perfect at doing what it does, but what it does will be very little
- too little to be useful.
-Matt :)
-----------------------------------------------------
www.auctionbrick.com - username mchiles
Matt Chiles
1006 Horseshoe Bend Rd
Centerville, WA 98613 USA
Phone: 509-773-5724
| | | | | | | | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.cad.dat.parts, Matthew J. Chiles <mattchiles@gorge.net> wrote:
>
> On Tue, 16 Jan 2007 11:06:31 GMT, you wrote:
>
> > When I made my first 3005-letters, I made them a compromise between accuracy
> > and
> > visibility. It was more important that the letters were readable in as small
> > scale as possible than that they were true to the originals. Do you find this
> > philosophy shocking? ;) That was in the days when LDraw was meant to produce
> > readable instructions. Today, I don't know what's become of LDraw. A playground
> > for perfectionists, where nothing or very little passes, maybe?
>
> "A playground for perfectionists"
>
> Excellent summary of what I am trying to say. Most of us are not
> perfectionists even if we would like to be, and we don't have time to
> be perfectionists. But we do want useable parts.
>
> There is a step below perfection that is "good enough". If L-Draw
> continues to strive for perfection at the expense of all else, it will
> be perfect at doing what it does, but what it does will be very little
> - too little to be useful.
You're right MAtt,
"Only the best is good enough" but "la surqualité est de la non-qualité"
(overquality is non-quality) and "le mieux est l'ennemi du bien" (better is
enemy of good).
Didier
| | | | | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.cad.dat.parts, Tore Eriksson wrote:
> In lugnet.cad.dat.parts, Travis Cobbs wrote:
> > I agree, with a few caveats. First of all, I think there should be a new
> > official header comment that basically says, "ok, but could use some work".
>
> Why not keep to the tradition of ending the description with "(needs work)"?
> Eg. "Minifig Flipper (needs work)". IMHO, there are already far too many useless
> official header comment and META statement to keep track on.
I did a quick scan of the official parts on my hard drive. The most recent one
with "(needs work)" in the part title is 30375s01 (Minifig Mechanical Torso
without Chest/Rib Surface (Needs Work)), and it's from the 2002-05 update.
There are two more recent files, but they both have a 0 NEEDS WORK comment
immediately following the header.
I couldn't find any reference to the phrase "needs work" in the part tracker
reference pages. A Google search of ldraw.org didn't match "needs work" in any
of the reference pages of the whole site. My suggestion would be to add this as
an officially sanctioned method of allowing parts to be approved that aren't
"perfect".
--Travis
| | | | | | | | | | | | | | | | | | | | | | | | |
| |
|
On Tue, 16 Jan 2007 21:45:40 GMT, you wrote:
> In lugnet.cad.dat.parts, Tore Eriksson wrote:
> > In lugnet.cad.dat.parts, Travis Cobbs wrote:
> > > I agree, with a few caveats. First of all, I think there should be a new
> > > official header comment that basically says, "ok, but could use some work".
> >
> > Why not keep to the tradition of ending the description with "(needs work)"?
> > Eg. "Minifig Flipper (needs work)". IMHO, there are already far too many
> > useless
> > official header comment and META statement to keep track on.
>
> I did a quick scan of the official parts on my hard drive. The most recent one
> with "(needs work)" in the part title is 30375s01 (Minifig Mechanical Torso
> without Chest/Rib Surface (Needs Work)), and it's from the 2002-05 update.
> There are two more recent files, but they both have a 0 NEEDS WORK comment
> immediately following the header.
>
> I couldn't find any reference to the phrase "needs work" in the part tracker
> reference pages. A Google search of ldraw.org didn't match "needs work" in any
> of the reference pages of the whole site. My suggestion would be to add this
> as
> an officially sanctioned method of allowing parts to be approved that aren't
> "perfect".
An excellent idea. If my vote counts for anything, I say yes.
Under the Needs Work comment authors or reviews could then note
specific items that need to be done such as "Needs BFC" or "Underside
needs more detail" or "needs more primitives in painted area".
This proposal could move a lot of 85% parts that are very useable into
official status so people can easily use them.
As long as a part is rotating on the best point and is shaped properly
with all obvious features, planes and lines it should be adequate.
-Matt :)
-----------------------------------------------------
www.auctionbrick.com - username mchiles
Matt Chiles
1006 Horseshoe Bend Rd
Centerville, WA 98613 USA
Phone: 509-773-5724
| | | | | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.cad.dat.parts, Travis Cobbs wrote:
> In lugnet.cad.dat.parts, Tore Eriksson wrote:
> > In lugnet.cad.dat.parts, Travis Cobbs wrote:
> > > I agree, with a few caveats. First of all, I think there should be a new
> > > official header comment that basically says, "ok, but could use some work".
> >
> > Why not keep to the tradition of ending the description with "(needs work)"?
> > Eg. "Minifig Flipper (needs work)". IMHO, there are already far too many useless
> > official header comment and META statement to keep track on.
>
> I did a quick scan of the official parts on my hard drive. The most recent one
> with "(needs work)" in the part title is 30375s01 (Minifig Mechanical Torso
> without Chest/Rib Surface (Needs Work)), and it's from the 2002-05 update.
> There are two more recent files, but they both have a 0 NEEDS WORK comment
> immediately following the header.
Have you counted the ones at the Tracker?
http://www.ldraw.org/cgi-bin/ptlist.cgi?sort=file
There are currently 30 parts or shortcuts with (nedds work) in description
line.
/Tore
| | | | | | | | | | | | | | | | | | |
| |
|
On Tue, 16 Jan 2007 05:39:17 GMT, you wrote:
> > If not all quads are co-planer, so what, that isn't going to impact
> > anything. Invisible gaps between planes, or unimportant undersides
> > not detailed 100% correct should not be reason to hold up parts. My
> > personal pet peeve is BFCing - I can see the value, but can't see
> > holding a part up for it, especially with the high speed, high memory
> > computers we use today where a few added back-faces are not going to
> > make an impact.
>
> The only reason a part should be held due to BFC is if the part is
> BFC-certified, but the BFC code is wrong. If there are errors in the BFC-ing
> of
> a part, it definitely needs to be held, because otherwise the flipped polygons
> will be invisible on BFC-compliant renderers.
>
> The part reviewing documentation specifically states the BFC-certification
> isn't
> required, though. If I authored a part (mind you, I don't), and somebody put a
> hold on it because it wasn't BFC-certified, I'd send an email to them politely
> asking them to remove the hold and explaining that BFC certification wasn't a
> requirement. If that failed to produce a result, I'd send an email to the
> admins asking THEM to remove the hold.
That is good advice because I have had more than one part I authored
get held ONLY because the reviewer thought it should be BFCed when no
claim was made by the part that it should be BFCed.
BFCing may be a good idea, and if someone wants to go to the extra
effort to do so them more power to them. But I agree that parts
should not be held only for not being BFCed, and I am glad to see that
my thoughts on this are not out of line.
-Matt :)
-----------------------------------------------------
www.auctionbrick.com - username mchiles
Matt Chiles
1006 Horseshoe Bend Rd
Centerville, WA 98613 USA
Phone: 509-773-5724
| | | | | | |
