Status of the word May

[dcbullock]

Lars Borg:

Summary:

From the ST minutes 2020-09-01

Status of the word ‘MAY’ in the foreword: eliminate and need not implement them as described. Some good words in email from Thomas. Some objection because it splits the definition of the word may at a certain point. Decision – review the OM wording. Default position is no change (especially if agreements cannot be made to change existing wording

Discussion thread copied below and strong feelings that we should not be able to make a standard where the word may is used to bypass the word shall.

Exactly our thoughts.

But we got pushback from 2-3 people so far.

Worth noting that ISO says ‘may’ states something is permitted, but it goes no further.

While SMPTE adds:

A conformant implementation need not implement optional provisions ("may") and need not implement them as described.

Should we revisit this in our aim to align with ISO?

Lars

[dcbullock]

@tbause

Both Bruce and I were on vacation last week.

• I agree with your and Raymond’s view. “You may take one cookie” means you have permission to take one (1) cookie or you may not take a cookie (0). But not two or three. The quantifier is “one”. I do have this situation with my daughter all the time and if she takes more then one cookie I tell her to put the excess cookies back because she was not permitted to take two or three As you said, seems to be a no-brainer.
• A standard should provide everything needed to implement it. So if people interpreted “You may take one cookie” as being able to take two or three then they are wrong as they are creating a lot of ambiguity and confusion for implementers. One could then also question “Cookie” and interpret it as a bag of “Gummy Bears”. A standards needs to specifically state what is required. If those people want more cookies it should say “You may take one, two or three cookies but no more”
• One could be more precise by saying “You may take one cookie but no more”. That way you put an end to interpretation and naughty children who think they can take the whole cookie jar.

Bottomline, “You may take one cookie” Is a clear statement and it does not give permission to take two or three cookies or gummy bears instead.

Best Regards

Thomas

[dcbullock]

Lars Borg:

An absurd issue has come up in regards to ST 2094-10, and I need ST’s help to clarify this.

Several members now believe that the word “may” allows the implementor to do anything including un-permitted actions.

Here is a snippet from an email thread:

My issue with “may” is that, for example, “the value may be X” does not forbid the value from being Y, whereas “the value shall be X” does forbid it from being Y.

I find this interpretation absurd.

If you tell your kid that he “may take one cookie” and he takes two, you will not accept his smart-ass response “you didn’t say I couldn’t take two”.

If a standard says “may include a sub-descriptor of type Charlie”, then this does not give you the permission to include two other objects of type Brown.

Neither is a useful interpretation of may.

Any skilled standards implementor will understand that one means one and not two, and Charlie does not mean Brown.

It makes common sense.

Most standards rely on this constraint and understanding.

Our boilerplate says:

1. The keywords "may" and "need not" indicate courses of action permissible within the limits of the document.

2. A conformant implementation need not implement optional provisions ("may") and need not implement them as described.

It seems that our members now twist point #2 to imply that you can take two cookies.

This needs to be addressed.

How?

Can we get a clarification from ST or HQ?

Please help

Mahalo,

Lars Borg