Stream: git-wasmtime

Topic: wasmtime / issue #6093 cranelift: Rename condition-code h...

Wasmtime GitHub notifications bot (Mar 23 2023 at 15:25):

jameysharp opened issue #6093:

          Yeah, I'm frequently confused between `inverse` and `reverse` too. I think both names are correct descriptions of what they do, but maybe later we can come up with alternate names that are less confusing. Maybe "negate" instead of "inverse", and "swap" instead of "reverse", for example.

_Originally posted by @jameysharp in discussion with @abrown and @meithecatte in https://github.com/bytecodealliance/wasmtime/pull/6086#discussion_r1146346401_

Wasmtime GitHub notifications bot (Mar 23 2023 at 15:26):

jameysharp edited issue #6093:

Yeah, I'm frequently confused between inverse and reverse too. I think both names are correct descriptions of what they do, but maybe later we can come up with alternate names that are less confusing. Maybe "negate" instead of "inverse", and "swap" instead of "reverse", for example.

_Originally posted by @jameysharp in discussion with @abrown and @meithecatte in https://github.com/bytecodealliance/wasmtime/pull/6086#discussion_r1146346401_

Wasmtime GitHub notifications bot (Mar 31 2023 at 00:43):

jameysharp labeled issue #6093:

Yeah, I'm frequently confused between inverse and reverse too. I think both names are correct descriptions of what they do, but maybe later we can come up with alternate names that are less confusing. Maybe "negate" instead of "inverse", and "swap" instead of "reverse", for example.

_Originally posted by @jameysharp in discussion with @abrown and @meithecatte in https://github.com/bytecodealliance/wasmtime/pull/6086#discussion_r1146346401_

Wasmtime GitHub notifications bot (Mar 31 2023 at 00:43):

jameysharp labeled issue #6093:

Yeah, I'm frequently confused between inverse and reverse too. I think both names are correct descriptions of what they do, but maybe later we can come up with alternate names that are less confusing. Maybe "negate" instead of "inverse", and "swap" instead of "reverse", for example.

_Originally posted by @jameysharp in discussion with @abrown and @meithecatte in https://github.com/bytecodealliance/wasmtime/pull/6086#discussion_r1146346401_

Wasmtime GitHub notifications bot (Mar 31 2023 at 00:43):

jameysharp labeled issue #6093:

Yeah, I'm frequently confused between inverse and reverse too. I think both names are correct descriptions of what they do, but maybe later we can come up with alternate names that are less confusing. Maybe "negate" instead of "inverse", and "swap" instead of "reverse", for example.

_Originally posted by @jameysharp in discussion with @abrown and @meithecatte in https://github.com/bytecodealliance/wasmtime/pull/6086#discussion_r1146346401_

Wasmtime GitHub notifications bot (Apr 03 2023 at 17:38):

jameysharp commented on issue #6093:

Name bikeshedding:

• negate or complement
• swapped or flipped

Wasmtime GitHub notifications bot (Apr 03 2023 at 17:45):

fitzgen commented on issue #6093:

Soft preference for flipped over swapped.

No preference on negate vs complement.

Wasmtime GitHub notifications bot (Aug 18 2023 at 13:05):

gurry commented on issue #6093:

If there's an agreement on a set of names I can implement this change.

Wasmtime GitHub notifications bot (Aug 18 2023 at 16:19):

jameysharp commented on issue #6093:

Sounds good! It's not important to pick the perfect names for this so I'm not going to bother looking for agreement from everybody. We just need names that are better than what we have now. :laughing:

Let's go with Nick's preference for "flip" to replace "reverse" since nobody else spoke up, and I'll pick "negate" to replace "inverse".

I think we should pick the verb form of these words, so use "flip" because that is the thing the function does, rather than "flipped" as the thing the function returns.

Wasmtime GitHub notifications bot (Aug 21 2023 at 16:04):

cfallin commented on issue #6093:

Sorry for missing this issue, somehow I didn't see the notification and only saw this after #6862 above -- I do have some thoughts on the bikeshed color if it's not too late though!

flip and negate are both still very confusing to me: flip could mean that we're flipping arguments, or the sense of the condcode itself; and negate sounds like it has something to do with twos-complement negation.

I might suggest flip_args for the first -- that makes it very clear that this is the condcode for the case where the args are flipped (commuted).

I might suggest complement rather than negate as there's precedent in Boolean logic: a complement really is the opposite-bit result (true for false and vice-versa), without connotations of integer negation.

Thoughts?

Wasmtime GitHub notifications bot (Aug 23 2023 at 15:58):

jameysharp commented on issue #6093:

After discussion in today's Cranelift meeting, our consensus is swap_args and complement. I'm sorry we didn't actually have consensus before you put a PR together!

Wasmtime GitHub notifications bot (Aug 24 2023 at 02:46):

gurry commented on issue #6093:

No worries @jameysharp. I'll make the change.

Wasmtime GitHub notifications bot (Aug 24 2023 at 04:24):

gurry commented on issue #6093:

Done

Wasmtime GitHub notifications bot (Aug 24 2023 at 05:51):

cfallin closed issue #6093:

Yeah, I'm frequently confused between inverse and reverse too. I think both names are correct descriptions of what they do, but maybe later we can come up with alternate names that are less confusing. Maybe "negate" instead of "inverse", and "swap" instead of "reverse", for example.

_Originally posted by @jameysharp in discussion with @abrown and @meithecatte in https://github.com/bytecodealliance/wasmtime/pull/6086#discussion_r1146346401_

Last updated: Dec 23 2024 at 12:05 UTC