Re: Selectors API naming

On 12/20/06, Chris Wilson <Chris.Wilson@microsoft.com> wrote:
>
> I disagree.  Anyone who uses both getElementsByTagName and
> getElementByID encounters a pattern.

They do not encounter a pattern of return values, allowed arguments,
handling of strange arguments (e.g. document.getElementById(null)), or
error handling.

>
> >If the issue is only one of clarity and consistency, then I
> >heartily concur with Ian--that's what editors are for.
>
> How many web developers do you think read the entire spec rather than
> dive in looking for API calls?
>

I think most authors cut and paste code. People that consider
themselves web developers probably cut and paste code from sources
known to be comprehensive (W3Schools, MDC, MSDN, etc).

My point was that we should let spec editors deal with naming,
argument formats and order, etc. If implementors or authors come back
screaming, then that probably indicates an issue. I don't see evidence
of hardship. Just likes and dislikes from most people (on both sides).

I'll confess I'm completely shocked to discover people advocating the
longer names. I guess they smell like standard. I sort of assumed that
one benefit of getting *everyone* together would be that we could use
nice short names without feeling guilty. I don't think using matchAll
carries a cost, unless we're worried that The One True Query Syntax
will emerge and be the rightful owner of all generic-sounding names.

> >Recommendations seem to vary widely in that regard.
>
> HA!  Amen to that.

Well, well. Common ground. :)

-- 

Robert Sayre

Received on Thursday, 21 December 2006 01:41:56 UTC