[Wikipedia-l] Parsing TeX

[Toby Bartels]

Jan Hidders wrote:

>Toby Bartels wrote:

>>One of the first pages that I looked at was [[Wikipedia:How does one edit
>>a page]]. I didn't edit any pages seriously until reading that. Not
>>everybody is like this, but some people are. Now, if I hadn't known TeX
>>already, I wouldn't have been heartened to see a reference to it. What do
>>you propose to write there?

>That mathematical variables are preferrably enclosed in [$ and $] and that
>if you want to use sub- and superscript in them you can write [$x_n$] and
>[$c^2$].

And that you'd better not write "[$x% = x/100$]", because that won't work.
And you'd better not write "the right [[coset]] [$H\G$]".
And woe to the person that tries to say "the set [${a,b,c}$]".
These aren't just features that one would expect but that don't appear;
they're serious flaws in LaTeX -- *if* you don't know the easy
(but difficult to guess) ways around them.

>Apart from that there would be another section containing more explanation
>for writing complex math with [$ .. $] and [$$ .. $$]. This would probably
>be the same as what you would write for [[math: ..]]. We cannot get around
>mentioning LaTeX somewhere.

Well, that's just it.
We must mention LaTeX to make [$c^2$] work your way.
But we don't need to mention it for other solutions.

You have taught me well, Jan, perhaps too well.
I've learned from you why generic HTML is bad in wiki,
and while I still think that <i> is harmless,
I see now why <i lang="fr"> is a bad idea.
(Not that I used that before, but I might have.)
I see that we need wiki syntax for tables
not just, or even primarily, because that'd be convenient for *me*,
but because it'd be possible *at*all* for many others to read.
And I see that it's bad to have 2 systems (wiki and HTML) at once,
which is why I don't think that we should accept any HTML,
only wiki tags that resemble HTML but which we parse,
and which are documented as wiki and not as HTML at all.

Yet now it is you that argue for having 2 systems, wiki and LaTeX.
You seem to like TeX; well, I like it too, and use it often.
Heck, I've lately been creating images for Wikipedia using Xy-pic.
Nobody would ever try to do that if they didn't love TeX.
But LaTeX is a complicated system.  I know people that don't like it,
smart people, professional mathematicians, that don't want to deal with it.
Now, we may be forced to use it, or something as complicated as it,
to write the sort of complicated expression that it's so famous for.
But we can't let it infect [$x% = x/100$].

>>The point is that you keep criticising "my notation", but you don't seem
>>to have any idea what that is. If the notation above had been suggested by
>>a third party, I would have argued against it.

>Please, Toby, I get the impression you are taking this a bit too personally.

Hum ... no, if you called it "Jim's notation" instead of "Toby's notation",
I'd still be miffed.  (I sometimes do take things personally,
but rarely because my name is attached.  See a recent post about 24.)

>I'm not criticizing your notation per se, I'm criticizing the idea of having
>two separate markups for variables and more complex math expressions, as
>opposed to just one for math expressions. Whether you want to allow only
>$$x$$<sub>$$y$$</sub> or also $$x<sub>y</sub>$$ is not really relevant for
>that question; the arguments stay the same.

That's because you haven't seen my arguments against the latter ^_^.
But I agree that having 2 systems of notation is problematic.
I just find it even *more* problematic to have 1 system of notation
that will mystify anybody that tries to use a percent symbol or a brace.

>I don't remember any discussion about, assuming that we should use <var>,
>how <var> should be used with variables with subscripts. But again, IMO this
>is all besides the point so I'll be happy to discuss this with you in
>another discussion when I have some more time. Right now I'd like to limit
>this dicussion to the question at hand.

That's not what I meant; what's relevant to me is the argument against <var>.
If that's really not right, then it could put a new perspective
on *my* view of the matter, even if it doesn't affect *yours*.
But since you and I are the only people discussing this now,
and I'm standing in opposition to you, still you should care about that.
So if you want to put it on my talk page instead, then that's OK by me,
but I'd still like to hear what the objection to <var> is.


-- Toby Bartels
   <toby+wikipedia-l at math.ucr.edu>