[latexrefman] Command vs macro

Vincent Belaïche vincent.belaiche at gmail.com
Sun Apr 12 20:56:31 CEST 2020


I must admit that I have not yet reverted the two occurrences where you 
(Karl) consider that we should keep « command».

Well, I agree, only « fragile commands » is a defined phrase.

But then in the sentence «  to make more commands be safe or…» (current 
text still has « macros » instead of «commands»), what does «safe» mean, 
why not just « robust » ?

I think that an example is the definiton of \caption, where \label is 
redefined as \let\label\@gobble to play no havoc.

BTW, this doing makes it impossible for packages like cleveref to 
redefine \label with an optional argument to work properly in some 
cases. LaTeX kernel should have some \neuratalize at label hook instead of 
directly \let\label\@gobble, then \neuratalize at label would be defined as 
\def\neutralize at label{\let\label\@gobble}, so this way cleveref would 
just have to redefine \neutralize at label to make neutralized \label to 
also admit an opt arg.


Le 31/03/2020 à 00:57, Karl Berry a écrit :
>      http://svn.gnu.org.ua/viewvc/latexrefman/trunk/latex2e.texi?r1=760&r2=777
> I agree with the first two changes,
>    ... any macros in @var{string} will be expanded ...
>    ... the macros are unexpanded until shipout ...
> But not with the other two:
>    ... on fragile macros ...
>    ... make more macros be safe ...
> As far as I've seen, the fragile concept is nearly universally used in
> the phrase "fragile commands". I can't recall ever seeing "fragile
> macros". And "commands" is not incorrect. So I think it is more
> confusing than helpful; I doubt anyone would make the connection with
> expansion vs. execution. If we want to say something about that, I think
> we should do it explicitly. --thanks, karl.

More information about the latexrefman mailing list.