On Mon, Nov 17, 2014 at 12:07 AM, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote: > Pascal de Bruijn wrote: > > > So those patches adapted for general use should be: > > > > https://encrypted.pcode.nl/files/temp/argyll17-dev-openbsd.patch > > Hi, a few of those patches need explanation before I will incorporate them: > Fair enough. > > - EXEMODE ?= 711 ; > + EXEMODE ?= 755 ; > > Why ? > I was surprised to see 711 there, as 755 is the standard mode for executables just about everywhere, as far as I know. Since I use a self packaged version of Argyll on Ubuntu, it's the Debian packaging code that actually fixes up these modes, so that's probably why I never noticed this before. > - if $(HOST64) { > + if $(OS) = OPENBSD { > + LibWinD = /usr/X11R6/lib ; > + } else if $(HOST64) { > if [ GLOB /usr/X11R6/lib : libX11.so ] { > LibWinD = /usr/X11R6/lib ; > > Why does BSD need this ? - ie why doesn't GLOB /usr/X11R6/lib : libX11.so > succeed > and trigger the LibWinD = /usr/X11R6/lib ; anyway ? > That's actually a good point. Leave that out for the time being. I'll retest without this. > > While I'm sure OpenBSD isn't a particularly important platform for you to > > support, considering the patches are trivial, I'm hoping you'd consider > > carrying them upstream. > > I'm happy to add all except the above two, until they are explained. > > I had a go at the BSD's and gave up when I came to the conclusion that > the USB capability is insufficient - there appears to be no way of > timing out certain operations for instance, short of building a special > kernel. > I can test some instruments to see how well it works once I get it to compile. That said, 1.6.3 already builds... > That said, I'm still getting a build failure in plot.c: > > > > https://encrypted.pcode.nl/files/temp/argyll17-openbsd-buildfail.txt > > > > Any ideas about that one? > > plot/plot.c:1847:22: error: X11/Xlib.h: No such file or directory > plot/plot.c:1848:23: error: X11/Xutil.h: No such file or directory > > Something amiss in locating the X11 headers. Right, so I'll have a dig a bit more :) Is there any chance you could incorporate the "good" bits already (into the dev zip), so I have a new basepoint to test from? Regards, Pascal de Bruijn