>>>>> "Eli" == Eli Zaretskii <[EMAIL PROTECTED]> writes:
Eli> On Sun, 5 Sep 1999, =?ISO-8859-1?Q?Torbj=F6rn Lindh?= wrote:
>> The following patch only works if you have perl (but why
>> wouldn't you?).
Eli> No, please don't introduce Perl into this equation. Many
Eli> people, especially on non-Unix platforms, don't have Perl
Eli> installed, and IMHO they shouldn't be required to. I also
Eli> believe that GNU standards say not to rely on Perl except for
Eli> maintainer's parts of the package.
I utterly disagree; but that argument is old, so I will not venture
into it as I cannot hope to sway GNU on this...
Eli> I think you can do the same with Awk. I would also suggest
Eli> to make this pass only if it is required (e.g., grep the
Eli> files for those ^^## patterns). That way, those who use
Eli> 7-bit ASCII won't need even Awk.
I absolutely believe that it can be done with awk, (although awk is
horrible and ugly:-), but I have since discovered another problem; the
.toc-file (table of contents) is read immediately after it is written,
so the problem must be solved within TeX.
My old way of discarding with the inane 8-bit-distruction in TeX has
never caused any other trouble than causing it to fail the trip test;
that is of course not acceptable for a wider audience.
I noticed that texinfo.tex contained a fix by [EMAIL PROTECTED]
that enabled TeX to read those ^^e4 sequences, but it had been
commented out as it broke all other uses of circumflex. I might come
back with some sort of TeX macro that will do the job properly.
In essence, I hear you, but for now you can disregard my patch as
further use proved it to be broken.
