Re: [AUCTeX-devel] Moving across formulas

[Tassilo Horn]

Marcin Borkowski <address@hidden> writes:

> One question.  What I currently do is a simple
>
>   (TeX-re-search-forward-unescaped "\\\\[])]" nil t)
>
> This /mostly/ works.  It doesn't work in cases like this:
>
>   -!-\( |x|=x \text{when \(x\ge0\)}\)
>
> but has the advantage of being fast.

What does "does not work" mean here?  I guess you mean that your
function just scans to the next occurrence of \) which is the one in the
\text here, right?

> So, do we prefer a faster version which is not entirely correct
> ("worse is better"), or is it better to have a correct (even if
> slower) version?

Obviously, that depends on how slow "slower" is.  But in general, I'd
prefer a correct version.

> (Since I have now the infrastructure for moving tokenwise,
> /implementing/ the correct version is not a problem.  Its speed might
> be.)

Then let's try it out.  I see no reason why it can't be fast enough.

> I'd really like to settle the FSF CA thing ASAP, though I have one
> more question to the FSF clerk.  I hope I'll be able to sort it out
> soon.

Great, we're awaiting your contributions! :-)

> (That does not mean that I'm no longer critical about FSF/RMS.  I
> still am, though less than earlier.  But this is another story, and
> I'm going to write a bit about my stance in a future blog post.)

I'm also genuinely interested in reading that.

Bye,
Tassilo