[OS X TeX] Strange error

Ross Moore ross at ics.mq.edu.au
Mon Nov 12 19:06:07 EST 2007

Hi Morten,

On 13/11/2007, at 9:54 AM, Morten Høgholm wrote:

> On Mon, 12 Nov 2007 23:41:30 +0100, Ross Moore wrote:
> Hello Ross,
>> What I don't understand is the subtle difference that using
>>   \begingroup ... \endgroup  makes in this context.
>> How is it different to  { ... } ?   or  \bgroup ... \egroup  ?
>> Most curious.
> And rightly so! :-)  For some reason, DEK decided to implement {...} 
> to act as a shorthand for \mathord{...} in math mode. And so nice TeX 
> programmers like Didier Verna get cought by it sometimes when all they 
> want is grouping

This applies only in math-mode, right.
Hmm. I've been using \bgroup ... \egroup a lot, especially in Xy-pic,
without any noticeable bad effects.

> but they wind up with more than they wished for. \bgroup and \egroup 
> act as { and } in this case and produce an Ord.

Yep. Understood.

> Somewhat related are the well-known (for those who know too much!)
>   \iffalse {\fi \iffalse`}=0 \fi
>   \iffalse`{=0 }\fi
> when it comes to opening special groups in alignments without 
> accidentally introducing extra math atoms.

I've seen these, and figured they had something to do with keeping
the braces balanced for  .docstrip  documentation purposes.
Now it'll stick in my mind, in case I ever have difficulty programming
alignment macros in math.

> Cheers,

Thanks for the instructive lesson.

> -- 
> Morten



More information about the MacOSX-TeX mailing list