> On Jun 22, 2016, at 4:40 PM, Thomas Klausner <wiz%NetBSD.org@localhost> wrote: > > On Wed, Jun 22, 2016 at 11:29:27AM -0600, Brook Milligan wrote: >> The math/R package is failing to install with a message about missing the cairo library. This seems to be because the configure script is looking for both cairo and pango, not just cairo. >> >> The following patch fixes it for me: >> >> --- Makefile 13 Apr 2016 21:04:04 -0000 1.165 >> +++ Makefile 22 Jun 2016 17:25:12 -0000 >> @@ -151,6 +151,7 @@ >> .include "../../archivers/xz/buildlink3.mk" >> .include "../../converters/libiconv/buildlink3.mk" >> .include "../../devel/gettext-lib/buildlink3.mk" >> +.include "../../devel/pango/buildlink3.mk" >> .include "../../devel/pcre/buildlink3.mk" >> # uses callback interface >> .include "../../devel/readline/buildlink3.mk” > > I just noticed that I had read the diff the wrong way round. > > So I've tried to reproduce this, but can't. > > R builds for me in a sandbox, both without pango installed and with > pango installed, without a problem. > > What is the error in detail you're seeing? It is odd that ‘make package’ works for you. Attached is a partial transcript of the output of my ‘make package’. I have only included a few bits relating to cairo and the last few lines completely, which show the error. What do you see in place of these? Cheers, Brook
Attachment:
R1.out
Description: Binary data