Pneumatic valve naming

[Philippe Hurbain]

The 42023 Mercedes Benz Arocs contains a variant of 47223 valve, with modified (stepped) output tubes.

According to Brickset, this new variant kept the same design number, but 47223a, 47223b and 47223c are used for various positions of switch. How should I name the new variant then?

[Michael Heidemann]

This is really a difficult question and I have currently no good solution.

But this leads to a decision that we should make now. Never naming physical states just with ..a, ...b, etc.
For these different physical states we should use in the future something like 4567_p1.dat.
If we use such an underscore and then something more descriptive like position 1 it will hopefully prevent us for the future from such a naming chaos that we are now faced.

Maybe we need to introduce something similar for the different variants of a part. We know there are many. Often we do not care about small changes. If we introduce an ending "-v" followed by a number for the variants we are free for many other options.
The number without the "-v" is then a "-v0".

Just two cents from me.

[Philippe Hurbain]

Looks like a good solution if everybody agrees...

The number without the "-v" is then a "-v0".

I'd rather favor no "-v" = first version then implies "-v1".
So here we would have 47223-v2_p1, 47223-v2_p2, 47223-v2_p3. Or rather 47223-v2-p1, 47223-v2-p2, 47223-v2-p3, I tend to prefer '-' to '_' as later one often melts into web link underscore...

[Michael Heidemann]

Your changes to my first shot seems reasonable.
So "-" for both usages is very ok.
And for the numeration, we always starts with 1 on counting and not 0 (except in applications :-) )

I hope Chris likes our ideas also.

Currently I do not see that DATHeader can help

cu
Mike

[Philippe Hurbain]

Maybe "p" is not the best since it could be confused with patterned parts. OTOH the dash in front makes a difference...

[Michael Heidemann]

You are right. Maybe we can use something more general. The only thing I come up now would be "state" and that starts with a "s". Not really better than "p".
So we should then maybe use "_p" in all the cases where we create position variants.

[Steffen]

Hi, sorry, but I disagree here:
The existing "c01" suffix is exactly for "shortcut", i.e. for things assembled together into a given position.

[Philippe Hurbain]

That's not what we did before (a, b etc...). This looks a very good idea to me.

[Michael Heidemann]

That is mostly true, but mainly in the time we did not know the correct number of the assembly. Now we know the numbers.

[Magnus Forsberg]

I have to agree with Steffen here. The c01, c02, .. suffix is the best way to go here.

This thread started with the naming of the valve, and that valve, and three older pneu cylinders, are the only affected parts.
I think they should be renumbered instead of introducing a third way of numbering parts.

Renumber these official files, and create Move to-files.
99798a => 99798c01
99798b => 99798c02

74981a => 74981c01
74981b => 74981c02

74982a => 74982c01
74982b => 74982c02

Old parts, at the Part Tracker
47223 => 47223a
47223a => 47223ac01
47223b => 47723ac02
47223c => 47723ac03

New parts, at the Part Tracker
19475-p1 => 19475c01
19475-p2 => 19475c02

19476-p1 => 19476c01
19476-p2 => 19476c02

19478-p1 => 19478c01
19478-p2 => 19478c02

47223-v2 => 47723b
47223-v2-p1 => 47223bc01
47223-v2-p2 => 47723bc02
47223-v2-p3 => 47723bc03

And this part is still missing from the PT.
It is only present, as a mock-up, in Philo's mpd-file of the MB truck

19482 Technic Pneumatic Pump 1 x 1 6L with Stepped Outlet (link to Bricklink)
19482c01
19482c02


All other pneu-parts are allready using the c01-suffix for different positions:
4694c01 - c03, Valve
47224c01 - c02, Cylinder
127c01 - c02, Cylinder
2793c01 - c02, Cylinder
335c01 - c02, Cylinder
2797c01 - c02, Pump

[Michael Heidemann]

Thanks for this very detailed information. I think you are right here. We should go the way you discribes.

[Philippe Hurbain]

Thanks for the summary - It at least alows me to notice that I forgot to submit 19482!

Actually I still think that p1-p2.... suffix is a good thing but currently not implemented correctly in all cases.

IMHO the c0x should be reserved for the various assemblies of elements (eg. with different colors). Different position of moveable parts are still the same part with same coloring, so they should get the same c0x suffix, and a -px suffix to indicate the various positions. And finally we don't need the c0x at all when the design number of the complete assembly is known.

Then we would have:

Renumber these official files, and create Move to-files.
99798a => 99798-p1
99798b => 99798-p2

74981a => 74981-p1
74981b => 74981-p2

74982a => 74982-p1
74982b => 74982-p2

Old parts, at the Part Tracker
47223 => 47223a
47223a => 47223a-p1
47223b => 47723a-p2
47223c => 47723a-p3

New parts, at the Part Tracker
19475-p1 => OK
19475-p2 => OK

19476-p1 => OK
19476-p2 => OK

19478-p1 => OK
19478-p2 => OK

47223-v2 => 47723b
47223-v2-p1 => 47223b-p1
47223-v2-p2 => 47723b-p2
47223-v2-p3 => 47723b-p3

And this part is still missing from the PT.
It is only present, as a mock-up, in Philo's mpd-file of the MB truck

19482 Technic Pneumatic Pump 1 x 1 6L with Stepped Outlet (link to Bricklink)
19482-p1
19482-p2

All other pneu-parts are using the c01-suffix for different positions: Renumber these official files, and create Move to-files.
4694c01 - c03, => 4694c01-p1 - 4694c01-p3
47224c01 - c02 => 47224c01-p1 - 47224c01-p2
127c01 - c02 => 127c01-p1 - 127c01-p2
2793c01 - c02 => 2793c01-p1 - 2793c01-p2
335c01 - c02 => 2793c01-p1 - 2793c01-p2
2797c01 - c02 => 2793c01-p1 - 2793c01-p2

[Chris Dee]

IMHO the c0x should be reserved for the various assemblies of elements (eg. with different colors). Different position of moveable parts are still the same part with same coloring, so they should get the same c0x suffix, and a -px suffix to indicate the various positions. And finally we don't need the c0x at all when the design number of the complete assembly is known.

I agree, and have always been a little concerned about the mixed usage of c0x suffixes. I would prefer a different letter to 'p' to avoid potential confusion with patterned assemblies.

For example, if we wanted a different positioned version of 52258p01c01, then 52258p01c01-p1 might be confusing.
[url=http://www.ldraw.org/cgi-bin/ptdetail.cgi?f=parts/52258p01c01.dat][/url]
Would 'a', for arrangement, work?

So, for example:

Renumber these official files, and create Move to-files.
99798a => 99798-a1
99798b => 99798-a2

74981a => 74981-a1
74981b => 74981-a2

74982a => 74982-a1
74982b => 74982-a2

Old parts, at the Part Tracker
47223 => 47223a
47223a => 47223a-a1
47223b => 47723a-a2
47223c => 47723a-a3


...

[Philippe Hurbain]

I agree, and have always been a little concerned about the mixed usage of c0x suffixes. I would prefer a different letter to 'p' to avoid potential confusion with patterned assemblies.

Indeed, p is no good because of "pattern" meaning. But I am not so fan either of 'a' (for assembly) since it is a suffix often used for parts variants. What about 'm' for "move"?

[Chris Dee]

Indeed, p is no good because of "pattern" meaning. But I am not so fan either of 'a' (for assembly) since it is a suffix often used for parts variants. What about 'm' for "move"?

'f' for form
'm' for move
'u' for usage

all work equally well for me.

[Philippe Hurbain]

I like 'f' for form!

[Chris Dee]

OK - I'll go with 'f'.