cvs commit: ports/textproc Makefile ports/textproc/htmlc Makefile distinfo pkg-descr

Timothy Beyer beyert at fastmail.fm
Sat Mar 27 08:05:57 UTC 2010


At Sat, 27 Mar 2010 02:34:46 +0000,
Alexey Dokuchaev wrote:
> 
> Just a couple of more comments, now that stas@ gave a diff; hope you
> don't mind Martin.
> 
> On Fri, Mar 26, 2010 at 07:15:44PM -0700, Stanislav Sedov wrote:
> > On Sun, 21 Mar 2010 12:56:53 +0000 (UTC)
> > Martin Wilke <miwi at FreeBSD.org> mentioned:
> > +MAINTAINER=    beyert at cs.ucr.edu
> > +COMMENT=       A text file generator
> 
> I'm afraid COMMENT should be augmented here; it is way too generic to
> describe a port.
> 
Point taken, I assumed that a terse description was desired, I could submit a more complete description soon.

> > +#MAN1=         htmlc.1
> 
> Any reason for this commented out statement?
> 
I didn't originally comment the man page out, I was wondering about the same thing.

> > +HAS_CONFIGURE= yes
> > +CONFIGURE_ARGS=        --prefix ${LOCALBASE}
> > 
> > This looks weird.  Is this port PREFIX safe?
> 
> In fact, passing correct --prefix is one thing of GNU_ vs. HAS_
> CONFIGURE statements.  Does this port uses autotools-generated configure
> script, by chance?  If yes, you should have dropped bogus --prefix
> assignment (entire CONFIGURE_ARGS line) and just use GNU_CONFIGURE.  If
> configure script is not autotools-generated, but accepts --prefix, it
> should probably be ${PREFIX}.  If this particular port assumes PREFIX ==
> LOCALBASE, comment should be provided stating (at best, explaining) it.

Point taken, I can fix that .
> 
> > +do-install:
> > +       @${INSTALL_PROGRAM} ${WRKSRC}/compiler/htmlc ${PREFIX}/bin
> > +       @${INSTALL_PROGRAM} ${WRKSRC}/compiler/htmlc.byt ${PREFIX}/bin
> > +       @${MKDIR} ${DATADIR}
> > +       @${INSTALL_DATA} ${WRKSRC}/config/env ${DATADIR}
> 
> INSTALL_FOO statements should not be muted.
> 
> ./danfe


More information about the cvs-ports mailing list