Streamlining the exchange of models that use unofficial parts

[Jim DeVona]

A common issue with sharing LDraw models seems to be unofficial parts usage. To recap the problem, since unofficial parts are subject to changes in origin, orientation, filename, etc., models built with unofficial parts may appear incorrectly or incomplete to users with different (or no) versions of that part. The recommended solution (see recent examples) is to include needed unofficial parts as MPD submodels, so the version used will always be available to others viewing the file. This is a good solution, but I don't think it is an obvious solution to the larger community of casual users. Plus, if you use certain unofficial parts frequently (or have downloaded all of them), I find it easy to forget that they are even unofficial - and there may be prerequisite unofficial subparts or primitives needed as well.

So:

1. Are there any editors or utilities that currently support saving models with unofficial parts (and dependencies) included as sub models? For example, a save dialog option or insert menu command to "Include unofficial parts as sub models", or a single-purpose utility to do the same for existing files. I suppose unofficial parts would be identified by looking for "Unofficial_" in the !LDRAW_ORG header line (or any non-compliance with the official library header specs, for that matter), or also, obviously, if they are housed in the "Unofficial/" subdirectory of the LDraw library.
   
2. Imagine if the Part Tracker kept a version history of the actual part files (like source code version control), not just an event log. (Does it?) Conceivably, this could then be used to retrieve the unofficial parts used to create models found in the wild - perhaps even programmatically, based on file creation date. I can imagine a number of cases where this strategy might break down, but maintaining a full version history seems like it might be useful anyway. (Moving forward, it could probably even be implemented as a third party mirror that monitors the main part tracker's activity log and commits copies of each new change...)
   

Thoughts and comments?

[Orion Pobursky]

This has really only became a problem as the major viewers/editors started added support for reading the Unofficial directory and then only if a user has consciously installed the unofficial parts package. This mean that a model author knows what he/she is getting into and should know to distribute the model appropriately. In my case, I simple overlooked it. My point is that I think leave things the way they are is best. Otherwise we risk "officializing" the unoffical parts.

[Philippe Hurbain]

Honestly Orion, I think you're too optimistic on average LDraw user awareness of unofficial parts issues. I have almost never found MPD's of models including unofficial parts... If only PT did kept complete history of files and name changes!

[Tim Gould]

This is why I think it's important to get the positioning and angle right quickly. Ideally I'd like a mid-stage where you have these approved even with a crappy design.

A part that looks wrong can be used and then fixed later, one that has the wrong orientation (eg. the early version of the 1x1x2/3 slope) can cause big problems.

[Philippe Hurbain]

Agreed. Remains the problem of name changes!

[Tim Gould]

Number changes are worse. For existing models a name change is unimportant. Or perhaps that's what you meant?

Luckily most new parts have readily accessible numbers.

[Steffen]

to add my 0.02 EUR to this discussion, I'd like to share my thoughts:

From my perspective, the problem is the backlog of parts on the PT.
The normal way this tracker should work is like a workbench.
Just a handful of files would sit on it, they would be cooperatively worked on,
then released. The time window while they are unofficial would be minimal.
People would not be forced to use so many unofficial parts if there were not so many of them,
sitting for so long on the PT.
So instead of being a pile of long unfinished parts which sit on the PT for years now (literally),
we should strive to finish lots of them and bring the PT more to a workbench character
instead of being an archive.This means that we'd have to continue bringing the backlog down
more and more. The last official parts updates were big steps in that direction. They brought
out lots and lots of now finished parts, giving users of our library plenty of beautifully crafted parts.

[Philippe Hurbain]

> Number changes are worse. For existing models a
> name change is unimportant. Or perhaps that's what
> you meant?
Yes, I meant file name, not description.
>
> Luckily most new parts have readily accessible
> numbers.
Not always immediately, unfortunately.

[Philippe Hurbain]

> So instead of being a pile of long unfinished
> parts which sit on the PT for years now
> (literally),
> we should strive to finish lots of them and bring
> the PT more to a workbench character
> instead of being an archive.
I'm afraid that despite the hard work of everybody involved, this remains a pious hope...

[Chris Dee]

I couldn't agree more. For example many of us worked hard on the 4.5V and 12V train track, yet we still lack sufficient reviewers to get these things released.

[Tim Gould]

Sadly I cannot review any of those lovely parts because I don't have them.

I'd like to see the policies altered slightly so that the PT Admin _can_ release a part with _no holds_ and at least one certify if it has been on the PT for more than two years. With the caveat that it be brought to the attention of reviewers first.

I think this would be valid since ultimately PT Admin is responsible for positioning, orientation and naming.

I know I repeat myself but I think that the Parts Library can suffer from excessive attention to non-functional detail at the expense of speed.

Tim

[Tim Gould]

This may help alleviate the problem somewhat

[Jim DeVona]

I'm glad to see this topic has generated some discussion.

> I know I repeat myself but I think that the Parts
> Library can suffer from excessive attention to
> non-functional detail at the expense of speed.

I agree with Tim that getting the basics right (filename, origin, orientation) are more important, at least as an initial priority, than cosmetic detail or optimal subpart/primitive usage. As a parts user, having a rough-looking placeholder that I can use with confidence is more valuable than a good-looking prototype that may move around or later disappear for other users.

[Steffen]

for reviewing a part, you don't have to possess it physically.

you can also check for just formal correctness.

[Jim DeVona]

Ah, cool. That answers a question I just posted in the other thread Tim started about parts in urgent need of review.