For the header, you can temporarily add a #error to the /usr/local one and
know whether it gets picked up or not.
--
Sent from a phone, apologies for poor formatting.
On 24 July 2018 16:48:33 Brian Callahan <[email protected]> wrote:
Hi Stuart --
On 7/24/2018 9:16 AM, Stuart Henderson wrote:
On 2018/07/24 15:11, Charlène wrote:
On Mon, 23 Jul 2018 17:05:34 -0400
Brian Callahan wrote:
Hi ports --
Attached is a new port, lang/mujs. MuJS is a lightweight, embeddable
Javascript interpreter.
Works for me on amd64, armv7, and macppc.
OK?
Did you check it doesn't affect the mupdf build? That has an internal
libmujs.a file, need to make sure that it picks up its own one in preference
to any installed /usr/local/lib/libmujs.
I did check this. MuPDF doesn't check for any installed MuJS. Instead,
what it does is unconditionally build the one.o object file that is
MuJS, archives that into a libmupdfthird.a with all the other third
party code, then links libmupdfthird.a into libmupdf.so.
However, because it doesn't bother checking, I can't rule out that mupdf
isn't picking up the mujs.h header file from an installed mujs during
build time, instead of the mujs.h header file included in mupdf. The
tweaks to mupdf to ensure this doesn't happen are minimal and I can send
something along if that's desirable. As I understand the way package
builds are done, this would not affect them.
Hi!
I've successfully tried some examples found in the html docs - by
the way the port doesn't install them - and also tested the
standalone interpreter: it's OK here on amd64.
Charlène.
The HTML in the docs/ folder is just a reproduction of the port's
HOMEPAGE. I'm not sure that's something we necessarily want to install.
~Brian
While testing is always welcome, when a developer is asking for an "OK"
for a commit, it is because they need approval from another developer,
it isn't something that somebody else can give.
