Hi Doug, Looks like today's my day to reply to years-old mails.
At 2021-01-03T17:26:35-0500, M Douglas McIlroy wrote: > I like the idea of an rfc1345.tmac file, and would be happy to operate > under the false assumption that it exists. I think it would be > reasonable to abandon the old AT&T accent strings while we're at it, > even though I have plenty of groff source that uses them. > > Corollary 1: So that calls for abandoned strings don't just silently > get ignored, warnings should be issued for unknown strings--preferably > only once per distinct name. groff has for many years issued such diagnostics when given the `-w mac` option. `-ww`, which turns on more-than-"all" warnings, implies it. > Corollary 2: If Corollary 1 is adopted, then for consistency warnings > should be issued similarly for undefined macro calls Since the same dictionary is used for requests, strings, macros, and diversions, `-w mac` works the same for all of these. Regrettably, the diagnostic itself calls them all "macros". > and undefined escape sequences. groff has also for many years issued such diagnostics when given the `-w escape` option. `-wall` implies it. This stuff is documented in the troff(1) page. $ printf 'hello\\= world \\*x\n' | groff -ww -z troff:<standard input>:1: warning: ignoring escape character before '=' troff:<standard input>:1: warning: macro 'x' not defined > Yours radically, > Doug I look forward to doing some radical things, but James Clark beat me to these. Regards, Branden
signature.asc
Description: PGP signature
