color codes for identifiers

May 16, 2026 Last reply: 2 months ago 24 Replies

You are stuck with "combinations" if you can't infer an order from the encoding. Simon's scheme effectively imposes an order as you "know" all of the colors can potentially be present -- or not. So, you can effectively order their presence or absence in your interpretation.

I suggested an implied ordering by using SILVER + colors vs. GOLD + colors -- in the hope that the marks would reliably be reported in the order encountered (even if that is the reverse of what you expect).

E.g., someone reporting the color bands on a resistor as GOLD BROWN BROWN VIOLET RED would still be understood as RED VIOLET BROWN BROWN GOLD.

Being able to use P(n,m) instead of C(n,m) greatly increases the number of different "encodings" available for any n & m.

[Sorry, forgot to get back to this...]

I think you always have to assume one of the colors that you adopt may be, effectively, unusable for a given application.

Consider marking a set of BLACK objects (the barrel connectors that I mentioned). You cant use any code that contains BLACK.

If, however, the set of objects includes some variety in "background colors", I *think* (but haven't formally proven this to myself) that you can keep all of the colors in play by choosing to assign that otherwise invalid color to the markings applied to at least one object that does NOT have it as a "background color".

This guarantee can only be made if you have the COMPLETE set of objects available at the time you make the initial assignment (think: use every encoding that references WHITE as none of the initial set of objects have a white background -- then encounter one that DOES at a later date!)

Moral: hard to come up with a universally applicable scheme (but, no reason it HAS to be universal as long as the number of available "unique codes" is sufficient to cover the set of objects of interest!)

As stated elsewhere, there is no significance to the color set or specific colors. BLACK-RED is just "BLACK-RED" and bears no interpretation relative to any other combination of markings.

I.e., BLACK-RED isn't bigger/smaller/lighter/heavier/cheaper/costlier than BLACK-anythingelse or anythingelse-RED, etc. There's no "ordering" involved so no need for a "valuation".

The sole point is to uniquely identify something as per the initial definition of that identifier by whatever party created the initial binding.

I once saw a system such as you describe used for labeling large wooden shipping crates. They were being shipped from a factory in Sweden to India. Every face of the crate was painted with three parallel diagonal lines of different colours. The purpose was to tell the dock workers which train to load the crates onto so that they ended up in the right part of the country. This was before the widespread use of ISO containers on that route. Most dock workers were illiterate. It was in 1975. The customer specified the colours to be used on the purchase order. John

A guy in another discussion on this topic said *railroad* cars are color-coded in some way. But, I can't understand why you'd need to "scan" railroad cars/trains as it seems like a rather heavyweight task to connect a particular car to a particular train -- not something done "casually".

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required