Please drop the DVI-with-pdfTeX hack! (was: [tex-live] [de.comp.text.tex] Re: hfoldsty und microtype - font problem)

Frank Küster frank at kuesterei.ch
Mon Sep 5 19:33:38 CEST 2005


Hi all, especially Sebastian,

we have discussed this last week and decided to delay further discussion
until Sebastian is back again.  I hope he's arrived now and is well.

Sebastian, maybe you want to read the mails in the first part of the
discussion (subject still in this mail);  the question is whether the
"hack" to undefine pdfTeX primitives and store them in \normalPRIMITIVE
should be kept or dropped.  It came up because a package author wondered
whether he was supposed to extend his check, although it was correct, to
check \normalWHATEVER instead.

My opinion is that it should be dropped:

Frank Küster <frank at kuesterei.ch> wrote:

> karl at freefriends.org (Karl Berry) wrote:
>
>>     Maybe, TL 2005 should also stop using these "new hacks" now!?
>>
>> Sorry, I continue to strongly disagree.  It will break thousands of
>> users documents -- like every one that tests for pdftex up until last year.
>
> If you don't break them now, when are you going to break them?  If you
> don't plan to break them anytime, have you already asked the pdftex
> people to rename their primitives, so that they are usable again?
>
> I don't know how long ifpdf.sty exists, but for sure it is longer than
> just last year.  Ever since then anybody who wanted to do it right could
> do it right.  And people where told to do it right in mailing lists and
> newsgroups.
> [...]
>
>> There is no good solution at this point.  
>
> I would say the clear cut solution is good.
>
>> Every compatible solution has
>> been broken.
>
> Indeed.  So there's no point in trying to stay compatible.  You can as
> well adopt the clean solution and not fiddle with pdfTeX primitives.

I have some additional question here.  First of all, ifpdf.sty in teTeX
does not test for \normalpdfoutput, but just for \pdfoutput.  Doesn't
that mean that even ifpdf.sty hasn't been adapted to work with this
hack?

Second, citing Karl Berry answering to Morten Høgholm:

,----
|     something that only ever worked due to shear luck. 
| 
| That is not my view of the situation at all.
| 
`----

What is your view instead, Karl?  

Regards, Frank
-- 
Frank Küster
Inst. f. Biochemie der Univ. Zürich
Debian Developer



More information about the tex-live mailing list