bibtex-1.0, or maybe not

[Karl Berry]

    A url={} field would contain... a URL

You might think so. But no. Sometimes url= looks like
  url="\url{...}"
or
  url="\href{...}{...}"
or
  url="who knows what people come up with"

Because bib format is so free-form, the reality as far as I've seen in
TUGboat is that people use the fields, well, freely :).

I guess my basic thought is that despite the many lacunae in original
bibtex, they've necessarily been solved in a variety of ways in the
absence of continuing development. And so providing one solution in a
hypothetical bibtex 1.0 feels more likely to cause trouble than help.

    I think that 'everything of interest' is doing quite a lot of work in
    that sentence!

Agreed. And yet I think it's (pretty much?) true.

    The @alias and @modify ideas in the 1994 bibtex-1.0 article are
    thought-provoking, 

Agreed. @alias and @modify were also the two things I saw that looked
most interesting.  But clearly they are not anything paramount, since
we've gotten along without them for these decades.

    as is the idea of 'better documentation'. 

Fully agreed. That is the biggest hole in original bibtex in my
experience. I have often wished for a straightforward bibtex reference
manual instead of this enumerated list that passes for documentation.
But my impression from talking with Oren is that he's not going to write
a ton of new documentation when there's no new program.

I admit I've only cursorily looked into TameTheBeast and other attempts
to fill the documentation hole. I can't say I was immediately
grabbed. If anyone has any recommendations, I'd be curious to hear.

    still-not-yet-1.0 program.

I think the 1.0 version number is the bigger fetish :).
When a program has been 0.99x for decades, well, it's 1.0 in reality.
Seems pointless to pretend otherwise.

    One could imagine distributions including a `bibtex` program at v1.0
    alongside a bibtex99 program scrupulously unchanged from v0.99d, for
    those occasions where a precisely reproducible workflow is desirable.

It's imaginable, but in reality we don't do that in TL (or MiKTeX as far
as I know). Distributions aren't the place to do version control, IMHO.

Best,
Karl