Ricardo Wurmus <rek...@elephly.net> writes:
> ng0 <n...@we.make.ritual.n0.is> writes:
>>> Is adding “bluez” and “libogg” really needed? They seem a little
>>> unusual for networking software.
>> It is not just networking software.
>> It's been to long since I found out all these dependencies, I just
>> decided to add them to guix master now. Bluez is for bluetooth support
>> of whatever feature (look it up in either documentation or the
>> Makefile), libogg is needed for audio playback I think. I could be wrong
>> about this one. You know, the longer you work on something the more you
>> find and without comments it all gets obscure. But I really think libogg
>> in addition to libopus is for something audio related.
> Well, that’s why I’m asking as I don’t have time to investigate the
> reasons for the suggested changes myself. I know what “bluez” and
> “libogg” do, but it’s a bit odd to have them as dependencies of GNUnet.
GNUnet has the ability to create an adhoc meshnetwork over bluetooth if
I remember correctly. I'll ask christian again, I never used this part.
greping gets you this:
ABOUT-NLS:157: bluez-pin |  
ABOUT-NLS:341: bluez-pin |      
ABOUT-NLS:525: bluez-pin |       
ABOUT-NLS:709: bluez-pin |     
ABOUT-NLS:893: bluez-pin |     
   |
ABOUT-NLS:1077: bluez-pin |       
 | 37
src/transport/gnunet-helper-transport-bluetooth.c:100: * In bluez library, the
maximum name length of a device is 8
Which is why I have to ask about this.
> GNUnet doesn’t encode/decode ogg audio files, does it?
It has this telephony feature (gnunet-conversation, despite the README
it no longer is experimental) and gnunet-helper-audio-record-gst
configure.ac:429: AC_DEFINE(HAVE_OGG,1,[Have ogg]),
configure.ac:432: AC_DEFINE(HAVE_OGG,0,[lacking ogg]))],
>> If you think comments are needed because someone in the future could
>> decide to cut down the "bloat", I can comment all the dependencies with
>> which part of GNUnet needs them.
> Commenting unusual inputs should be enough.
> ~~ Ricardo