Re: Syntactic variability

I'm reasonably neutral on this - perhaps I'd just not like '=' on the 
grounds that most production rules I've come across use either ':' or 
':='. I wouldn't have thought it would be much of an issue to support 
both pairs of synonyms, though of course 'roundtripping' might well suffer.

John

On 18/04/2022 10:18, Norm Tovey-Walsh wrote:
> Hello,
>
> I’ve been toying with creating an issue for these ideas, but we need to
> be reducing the number of issues at this point, not increasing them, so
> I keep talking myself out of it.
>
> There is unnecessary syntactic variability in ixml that I don’t really
> understand. We allow either “:” or “=” as a rule separator and we allow
> either “;” or “|” as a alternative separator.
>
> I don’t think we’re doing our users a service this way. I’m prepared to
> believe that there are users who favor “:” and “;” over “=” and “|” (and
> perhaps even other pairings) but I have a hard time believing that it
> would be make-or-break deal for anyone: “I love the idea of ixml, but I
> refuse to use “=” and “|” so I’m not going to use it.”
>
> I tend to use, and perhaps even prefer “:” and “;”, but I propose that
> we adopt “=” and “|” exclusively.
>
> Using “=” would eliminate the ambiguity caused by colons in nonterminal
> names, whether we adopt a proposal to allow that for version 1.0 or
> v.Next.
>
> Using “|” would reduce the syntactic similarity of “sequence” from
> “alternate”. On several occasions, I have used “,” where I meant “;” and
> it’s hard to see. I don’t think I would be as likely to use “,” where I
> meant “|” and if I did, it would be easier to see the difference. This
> is especially the case in character classes, where I’m drawn to
> [',', '.'] instead of [','; '.']. I’d be better off with [',' | '.']
>
>                                          Be seeing you,
>                                            norm
>
> --
> Norm Tovey-Walsh
> Saxonica


-- 
*John Lumley* MA PhD CEng FIEE
john@saxonica.com
on behalf of Saxonica Ltd