On Sat, Feb 25, 2012 at 9:02 AM, Hans de Goede <j.w.r.degoede(a)gmail.com> wrote:
On 02/25/2012 03:52 PM, Richard Shaw wrote:
> On Fri, Feb 24, 2012 at 4:37 PM, Nicolas Chauvet<kwizart(a)gmail.com>
> wrote:
>>
>> 2012/2/24 Richard Shaw<hobbes1069(a)gmail.com>:
>>>
>>> Ok, I tried the options and they seemed to work except for libmad...
>>>
>>> checking mad.h usability... no
>>> checking mad.h presence... no
>>> checking for mad.h... no
>>> checking whether to dlopen mad... yes
>>> ...
>>> mp3........................dyn
>>> id3tag....................yes
>>> lame......................yes
>>> dlopen lame...............yes
>>> mad.......................no
>>>
>>> Hmm... where to go from here.
>>
>>
>> Please forget about "--enable-dl-lame --enable-dl-mad" I never talk
about
>> that.
>> This is not the the right path for fedora.
>
>
> I understand that we wouldn't use those options for the RPM Fusion
> package, but wouldn't the Fedora package need these options enabled in
> order to look for the additional plugin libraries?
>
Sorry for dropping into this thread late, and esp.
after not having read it all. But has someone considered to
add the relevant lame / mad header files as extra sources
to the existing Fedora package and to build the existing
Fedora package with dlopen support ?
I'm glad you did drop in!
Well, one thing I was trying to figure out is if we still needed the
headers. From the configure output it looks like we don't for lame but
do for mad. I'm going to try building test packages of both "new"
Fedora and RPM Fusion packages and see if I can get it to work. Once I
figure that out I'll submit a bug against Fedora sox with a git diff
and we'll see what happens.
Any help or advice would be appreciated!
Richard