@iuioiua We talked about #4082 and you said single file exports are overkill for cases. So I thought, maybe we should put these functions inside compare.ts as well, since they are sugary compare() functions.

@iuioiua We talked about #4082 and you said single file exports are overkill for cases. So I thought, maybe we should put these functions inside compare.ts as well, since they are sugary compare() functions.

Yes. In cases where functions don't incur any dependency costs on each other and share much overlap, it might make sense to have them in the same file. With equals() and friends, this might make sense to do. It also might not. E.g. if such a structure would be surprising to a newcomer, it might not make sense to do. WDYT, @kt3k?

@iuioiua

Yes. In cases where functions don't incur any dependency costs on each other and share much overlap, it might make sense to have them in the same file. With equals() and friends, this might make sense to do. It also might not. E.g. if such a structure would be surprising to a newcomer, it might not make sense to do. WDYT, @kt3k?

Sounds interesting/reasonable idea to me, but not immediately obvious if it's good or not. I'd like to keep discussing that in somewhere else.

@iuioiua

Yes. In cases where functions don't incur any dependency costs on each other and share much overlap, it might make sense to have them in the same file. With equals() and friends, this might make sense to do. It also might not. E.g. if such a structure would be surprising to a newcomer, it might not make sense to do. WDYT, @kt3k?

Sounds interesting/reasonable idea to me, but not immediately obvious if it's good or not. I'd like to keep discussing that in somewhere else.

I'll add it to the list in #3948

Another concern about naming: We have assertLess assertGreater assertGreaterOrEqual assertLessOrEqual in std/assert #3496

So these names might look inconsistent with them. What do you think?

Another concern about naming: We have assertLess assertGreater assertGreaterOrEqual assertLessOrEqual in std/assert #3496

So these names might look inconsistent with them. What do you think?

Thank you for bringing that up. I think we should align to assert names, I'll update the PR.

Another concern about naming: We have assertLess assertGreater assertGreaterOrEqual assertLessOrEqual in std/assert #3496
So these names might look inconsistent with them. What do you think?

Thank you for bringing that up. I think we should align to assert names, I'll update the PR.

Sorry for coming in late. I don't see a reason to align these function names with those from std/assert. They're different APIs with different purposes. But more importantly, I think removing the "than" word from these function names decreases their clarity. E.g. the purpose of less() in std/semver isn't immediately obvious. However, lessThan() in std/semver is clearer. What do we think?

@iuioiua
What do you think about the below?

• lessThan
• lessOrEqual
• greaterThan
• greaterOrEqual

@iuioiua What do you think about the below?

• lessThan
• lessOrEqual
• greaterThan
• greaterOrEqual

SGTM. To me, lessThanOrEqual() and friends sound better. But that's a weak opinion and not something I'd block any PR over. I also understand that lessThanOrEqual() may be too long.

SGTM. To me, lessThanOrEqual() and friends sound better. But that's a weak opinion and not something I'd block any PR over. I also understand that lessThanOrEqual() may be too long.

@kt3k how should we proceed?

@kt3k how should we proceed?

The below pattern looks best among candidates to me.

• lessThan
• lessThanOrEqual
• greaterThan
• greaterThanOrEqual

done

Mostly looks good to me. Thanks for updating.

Added an "s" to the *OrEqual() functions, to be consistent with equals().

Does that sound better to you? If I search lessThanOrEqual and lessThanOrEquals in google, the former returns more results (lessThanOrEqual looks more common as an API name to me). I think equal here is more of an adjective rather than a verb..

I just thought "a equals b".

• a equals b <- equals is a verb, so it has s suffix
• a is less than or equal b <- less than and equal are adjectives. equal doesn't have s

lessThanOrEqual or lessThanOrEquals

Maybe not a big deal, but I'd like to hear some more from community members

I know I said the names don't have to be consistent with std/assert, but now, the names are consistent with std/assert 😂

Either way, the names convey the purpose of the functions, which is the objective of a function name, with or without the "s".

TBH I didn't mind the abbreviated names. But I don't mind the longer names either.

I vote for lessThanOrEquals. To be very precise grammatically, it should be lessThanOrEqualTo, which is too long. So, if we're going to be grammatically imprecise anyway, lessThanOrEquals is semantically closer to lessThanOrEqualTo than lessThanOrEqual is because it keeps a strong sense of direction (which means the order of the arguments in this case).

TBH I didn't mind the abbreviated names. But I don't mind the longer names either.

I vote for lessThanOrEquals. To be very precise grammatically, it should be lessThanOrEqualTo, which is too long. So, if we're going to be grammatically imprecise anyway, lessThanOrEquals is semantically closer to lessThanOrEqualTo than lessThanOrEqual is because it keeps a strong sense of direction (which means the order of the arguments in this case).

Well, lessThanOrEqualTo is just one letter longer than lessThanOrEquals:) I think the length should not be the determining factor, because everybody is probably using autocomplete anyway.

That's true. But if we go for lessThanOrEqualTo, maybe we should also rename equals to equalTo. It feels a little too verbose to me. And when I see the word equals, my brain parses it as a symbol rather than a "grammatical word", so it's actually faster for me to read lessThanOrEquals than lessThanOrEqualTo. So I'm still in favor lessThanOrEquals. Sorry if I'm overthinking this 😅.

I now think it is worth simply aligning with std/assert. The names are short and simple and show their purpose. That's all that's needed. I'm sorry for the indecisiveness.

I now think it is worth simply aligning with std/assert. The names are short and simple and show their purpose. That's all that's needed. I'm sorry for the indecisiveness.

Done

Let's settle on greaterOrEqual and lessOrEqual for gte and lte replacements, which now have much support from core team (@mmastrac @satyarohith @lucacasonato @crowlKats @dsherret are in favor of them).

I changed less and greater to lessThan and greaterThan. They are a bit inconsistent with assertLess and assertGreater of std/assert, but less & greater look too less descriptive to me. I believe this change is a reasonable compromise between consistency and descriptiveness.