On Tue, May 24, 2011 at 7:36 AM, Greg Stein <gstein_at_gmail.com> wrote:
> On Tue, May 24, 2011 at 09:39, Mark Phippard <markphip_at_gmail.com> wrote:
>>...
>> Wasn't it JavaHL that failed to compile? If we want developers to
>> always compile the Java bindings then why don't we remove it as a
>> separate build target and just make it part of our default?
>
> You're right. We should probably Just Do This.
>
> We've gone back and forth on bindings for a long time now. I don't
> really see much of a problem if we just built them by default (any
> bindings that are configured for the local machine, that is).
>
> Ten years ago? Yeah... compiles weren't exactly speedy. But today?
> Pfft. Compile it all. :-)
I don't think I'm concerned with compile time (though JavaHL doesn't
play nicely with 'make -j'), but rather output. The swig bindings
spew a lot of superfluous output, and that can mask warnings / errors
in the core C libraries.
-Hyrum
Received on 2011-05-24 17:10:05 CEST