On Mon, Jul 28, 2008 at 08:29:58PM +0200, Markus Lude wrote:
> On Wed, Jul 16, 2008 at 08:11:44PM -0401, jared r r spiegel wrote:
> > On Wed, Jul 16, 2008 at 08:02:59PM -0400, Brad wrote:
> > > On Wednesday 16 July 2008 19:50:05 jared r r spiegel wrote:
> > > >   the pkg/PLIST in here has a .orig file it slurps up in 'make
> > > > update-plist', it's not a critical update, but corrects a local-specific
> > > > path in one of the default macro files.  don't know what the preferred 
> > > > way
> > > > of handling this is (just leave it, make a list of files to copy by 
> > > > hand in
> > > > the Makefile, make a command in the Makefile to remove that .orig file
> > > > before do-install, etc).
> > > 
> > > Just use @comment to exclude the file.
> > 
> >   yeah, whoops; sthen@ reminded me minutes after i sent it.
> > 
> >   i had forgotten the nice part of pkg_create(1) that says: "hey
> >   jackass, this is the right way to do that"
> > 
> >   attached (b/c they're so small) is updated tgz who inc'ls that @comment.
> 
> Please try the attached diff against -current. It should fix some minor
> problems like:
> 
> * use in-tree pcre
> * keep .tf files in $PREFIX/share/tinyfuge. You moved them to
>   $PREFIX/share/tinyfuge/lib
> * add all .tf files, some links were missing
> 
> One thing where I'm a bit unsure: I added a cast in src/socket.c to
> silence the warning:
> 
>   socket.c: In function `handle_socket_lines':
>   socket.c:2603: warning: assignment from incompatible pointer type
> 
> Both LHS and RHS are pointers to a struct. The difference between both
> is in one 2 members are const while not in the other struct. Is it
> better to keep the warning or to add the cast as in
> patches/patch-src_socket_c?
> 
> Builds fine here on sparc64. Only lightly tested so far. I'll do some
> more testing during next days.

There are some minor problems. I'll post an update after the tree
unlocks.

Regards,
Markus

Reply via email to