Naming things

[Steven Schveighoffer via Digitalmars-d]

On 6/20/15 5:27 AM, Vladimir Panteleev wrote:
> Naming things
>
>        There are only two hard things
>        in Computer Science: cache
>        invalidation and naming things.
>        -- Phil Karlton
>

I think the issue we are struggling with here is that:

abbreviate -> abbreviated

This makes some sense. However, the past participle of "set" is "set".

So "set the extension" -> "a set extension" doesn't work, because "set" 
doesn't change. Our enemy here is the English language :)

If the original was named something like "modifyExt", then "modifiedExt" 
would be fine.

And my understanding of the pushback from Walter about renaming really 
has to do with avoiding breaking code for the sake of renaming. At this 
point (before setExtension has ever been released), it's "what is the 
best name". No code should be broken, the renaming objection shouldn't 
apply. I'm 100% in favor of not having both setExt and setExtension to 
mean different but similar things.

withExt seems better, and reasonably informative.

path.withExt("abc") -> "use this path, but with extension '.abc'"

And 'with' doesn't work with every possible updated version, we have to 
work around the quirks of English here.

But really, the egregious error is the slightly different yet identical 
names. It's like having setExt and set_ext mean different things. This 
also reminds me of std.regex vs. std.regexp. I never knew which one was 
the "new" version.

If there is some other idea besides setExtension (or some prefix of 
that), I think we should go with that. We could use some analog to 
'set', like 'modifiedExt' or 'changedExt' if that sounds better. It just 
shouldn't be the same exact name, with differing levels of abbreviation.

-Steve