[Groff] Re: a couple of fixes to html.cc

[Werner LEMBERG]

> I've fixed a number of bugs in your manual page example with this
> patch. I'm now up to the OPTIONS section I believe :-) (incidently I
> use the -mhtml macros after the -man macros to eliminate the
> hyphernation problem)

Gaius, the next thing you will see in the CVS (as soon as it is
running again) is the whole html.cc file you've sent to me.  Please
check it again -- if it is OK, I will release the groff package.

BTW, in case you send patches please update your copy of the ChangeLog
file from the CVS also -- I sometimes edit it slightly.

> However I thought I'd email this patch in the hope that it
> can be included in the ship before it leaves port?
> Hopefully this is one advantage of being 5 hrs ahead
> of the USA? :-). The diff is against the repositary 8:45am GMT.

No chance -- I'm sitting in Dortmund, Germany :-)

> On another point, I wonder whether when we install groff if it
> should create symbolic links from the g equivalent binaries to the
> old troff names.  On my machine (RH-5.2) the geqn, gtbl, grefer,
> gseolim all disappeared after I removed them from the previous
> release (/usr/bin/g...) and rebuilt groff (when it installed them
> into /usr/local/bin).  Is this expected?

Yes.  it is handled by the GROFF_G macro in aclocal.m4 -- the
configure script tests whether a program called `troff' is installed
and whether it works.  Only if yes, the prefix `g' is used.  I'm not
really happy with this behaviour since you can't influence it.
Additionally, it is inconsistent since some drivers always have a `g'
at the beginning.

Anybody on the list who has suggestions how to improve this?  I mean
the inconsistency, not the automatic process of checking the `g'
prefix (which can be changed rather easily using the AC_ARG_PROGRAM
autoconf function).  My idea is to always use `g'.


    Werner