On 2025-08-08 19:15:07 +0300, Eli Zaretskii wrote: > > Date: Thu, 07 Aug 2025 12:55:20 +0300 > > From: Eli Zaretskii > > Cc: gdb-patches@sourceware.org > > > > > Date: Thu, 7 Aug 2025 11:05:24 +0200 > > > From: Vincent Lefevre > > > Cc: gdb-patches@sourceware.org > > > > > > On 2025-08-07 08:50:14 +0300, Eli Zaretskii wrote: > > > > That's true, but there are TeX experts there. If you know of a better > > > > place, we could discuss this there. I unfortunately don't know enough > > > > about this to make a decision about the right solution. > > > > > > FYI, the TeX Users Group has mailing-lists: > > > > > > https://tug.org/mailman/listinfo > > > > > > texhax seems to be the one for general TeX questions. > > > > Thanks, posted a question there. > > Given the responses there, can you post (and hopefully test) an > improved patch? Note: my old patch was based on what was done in PARI. I've attached a new patch. It still uses \ifx\pdfoutput\undefined because the suggested \ifnum\pdfoutput<1 gives an error. For dvi and pdf, I could not see any issue. For ps, I get with "make refcard.ps": DVIPS refcard.ps dvips: both landscape and papersize specified: ignoring landscape The dvips message is due to the "-t landscape" in the Makefile (from Makefile.in): refcard.ps : refcard.dvi $(ECHO_DVIPS) $(DVIPS) $(SILENT_Q_FLAG) -t landscape -o $@ $? It is just redundant and could possibly be useful in case the \special isn't recognized in some settings, but you may choose to remove this option. In both cases, there are no issues with "gv refcard.ps". -- Vincent Lefèvre - Web: 100% accessible validated (X)HTML - Blog: Work: CR INRIA - computer arithmetic / Pascaline project (LIP, ENS-Lyon)