Sorry for my tardy reply; it's a very long time since I've been able to keep up with the groff list.
On Mon, Feb 27, 2006 at 02:04:32PM +0100, Bruno Haible wrote: > Keith MARSHALL wrote: [...] > > >> 1) Although the man page starts with > > >> > > >> '\" t -*- coding: utf-8 -*- > > >> > > >> the groff driver is not intelligent enough to run preconv. I > > >> have to activate the -k option explicitly. > > > > > > This is a feature. I think it is best to not change this -- at > > > least not now. > > > > Maybe this is something that man, rather than groff, should do? > > > > Man already reads the first line of the man page, to check for the > > '\" t, etc. tag; this tells it which preprocessors to include in its > > command pipeline. I don't think it would be too difficult to have it > > also check for the coding tag, and add the -k option to the nroff > > options when required, > > I agree, Sorry, I had thought that groff, nroff or soelim were deciding > whether to run 'pic', 'tbl', etc. Since it's 'man' which decides this, > it's also 'man' which has to decide whether to run 'preconv'. As the maintainer of one of the two Linux man implementations (man-db), I agree. In particular, there's a useful cue that can be used on some systems in the absence of a -*- coding tag, which may not have been mentioned: the FHS specifies [1] a layout for /usr/share/man which includes an encoding specification in the name of the directory containing man pages. (The design of that encoding specification is IMO poor, in that it unnecessarily suggests decanonicalising "ISO-8859-1" to "88591", but its heart is in the right place and its deficiencies can be worked around.) Only man is in a position to get at this information. [1] http://www.pathname.com/fhs/pub/fhs-2.3.html#USRSHAREMANMANUALPAGES FWIW, I've been (rightly) under pressure for a while to update Debian's groff, but have found it very difficult to forward-port the notorious Debian Japanese patch; it's been a while since I last tried it, but I seem to recall encountering particular trouble in devps. Having caught up slightly on the state of groff, I'm seriously considering just ditching the whole mess and shipping with preconv instead, and doing the couple of hours' work required to teach man to use it. Would this be a totally insane thing to do? For that matter, if it comes to it, would shipping CVS groff right now be a totally insane thing to do? Thanks, -- Colin Watson [EMAIL PROTECTED] _______________________________________________ Groff mailing list Groff@gnu.org http://lists.gnu.org/mailman/listinfo/groff
