Subject:
|
Re: Legacy Files on the Parts Tracker
|
Newsgroups:
|
lugnet.cad.dev.org.ldraw
|
Date:
|
Tue, 30 Sep 2003 07:12:25 GMT
|
Viewed:
|
473 times
|
| |
|
|
On Fri, 26 Sep 2003, Franklin W. Cain wrote:
> Even if we stopped uploading new files, even if the two Admins (Steve
> and Chris) devoted their every waking moment to reviewing the existing
> files, it will still take (by my estimate) the better part of a *year*
> to get all the files we've currently got out the door.
>
> I respectfully suggest that something might not be right with this
> situation. (Not necessarily all-the-way-*broken*, I'll be willing to
> concede, but definitely something-ain't-right)...
>
> Unfortunately, I don't really have any reasonable suggestions on how to
> fix this problem. I sincerely hope that *someone* can think of a
> solution that's acceptable to everyone...
I can think if a few things that might help throughput (I'm sure that most
of these have been discussed before):
1) Create a new review attribute "compared to physical part" (which could
apply to an admin or regular review), which certifies that the reviewer
has the part and has actually measured it, made sure that the pattern
matches reality, whatever. Then require that at least or two reviews have
this attribute for the part to be certified. This a) encourages people
who don't have the part to still go ahead and check for gaps, overlaps,
wrong BFC, etc., secure in the knowledge that their approval will not let
a technically proficient but *wrong* part in the library, and b)
encourages people who *do* have the part to review it (their reviews
become more valuable), and to not sweat the technical details if that
would mean fewer reviews (because there are others whose contribution *is*
to sweat the technical details).
2) Create new votes: "will review" and "will not review". (The "will not
review" could be for whatever undisclosed reason---wrote the part, don't
have the part, don't care about the part, author stole the reviewer's girl
in the seventh grade. . .) This would most obviously benefit the
admins---if one admin won't review a part, the other knows he must. It
would really help all reviewers prioritize their reviewing activities.
Note that there is no reason to report names for these votes; a running
tally of "M Admins and N non-Admins have said they will review this part,
X Admins and Y non-Admins have said they will not review this part" (and
the ability to sort by these numbers) would suffice.
3) Create a bit for every part on the Tracker, set at submission time:
(a) Feel free to make fixes if you find problems
(b) Please notify the author if you find problems
with (b) changing to (a) if a part has a hold vote stand for more than 30
days. This would provide a formal convention which would allow many
small problems to be quickly fixed, while also allowing authors to retain
control of their parts throughout the certification process, if desired.
It would also end the problem of orphan parts. (There are many orphans
that have been on the Tracker for a long time.)
--
TWS Garrison
http://www.math.purdue.edu/~tgarriso/
Remove capital letters in address for direct reply.
|
|
Message has 3 Replies:
 | | Re: Legacy Files on the Parts Tracker
|
| (...) I like this, sort of. While I thinks it's a good idea, the overall effect would be to lower throughput since not everyone has an expansive LEGO collection. Isn't that part of the reason why some started using virtual LEGO in the first place? (...) (21 years ago, 1-Oct-03, to lugnet.cad.dev.org.ldraw)
| | | Re: Legacy Files on the Parts Tracker
|
| (...) I agree here. I own a bunch of "rare" parts and could compare the ldraw parts to the real thing. But I dont know anything about the technical stuff and cant say on those. Also, I could provide comments on part names, KEYWORDS etc (since I know (...) (21 years ago, 4-Oct-03, to lugnet.cad.dev.org.ldraw)
| | | Re: Legacy Files on the Parts Tracker
|
| Going back a ways for this one... (...) I considered going a similar route when I constructed the PT: offer reviewers a list of checkpoints that they should consider when posting a review. I abandoned this thinking because the 'rules' for processing (...) (21 years ago, 3-Nov-03, to lugnet.cad.dev.org.ldraw)
|
6 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
