...pass a -V option to say what channels you want to show. (multiple -V or duplicates of channel symbols are handled gracefully). "-V" for Visible?

e[.]g.
aumix -V wvc # would show only the vol,pcm and cd

It could be ported to the curses mode I suppose, I might have a look at that tonight. And, if I have time, some documentation (man page update) would be useful I suppose ;)

I want aumix as minimal (optionally of course) as kmix can be). There aren't many option letters left to cover things like "show balance channel", "show menu", "show spacewasting labels"!

PPS: Might try to squeeze them into -V as well

...dynamic handling of channels (like it is, if someone change[s] the volume outside of aumix, it knows that and update[s] the mixer). It is needed for example for SBLive, where one can add support for bass/treble control[s] at runtime by loading a patch.

It only demonstrates an idea and it is functional only in ncurses, as I am not very experienced with GTK[.]

...adds three new things that can be hidden (and later shown again) to the View menu in the GTK+ interface:

• Menu, for the menu bar itself (it can be shown again by pressing the hotkey)
• Balance, for the balances
• Numbers, to get rid of all those confusing numbers.

...a small shell script to mute.

Here's a first attempt at adding long options to aumix. I made options for all the mixing channels that couldn't be adjusted from the command line before. The rest of the new options are just long versions of some of the old ones. So far this only works on Linux. On FreeBSD there's a separate lingnugetopt, typically installed in /usr/local/lib/, that contains the getopt_long() function. The configure script is supposed to find this, as well as getopt.h (typically in /usr/local/include/) but so far

./configure --libdir=/usr/local/lib --includedir=/usr/local/include

fails to find either of them. For systems that have neither the GNU libc nor a separate library with getopt_long(), there are two obvious possibilities: we could distribute getopt.c from the GNU C library and compile the function into aumix, or we could disable the long options.

At the end is an unrelated change to gtk.h.

updated Spanish translation

updated Galician translation

an updated Ukrainian [translation]

I just noticed that aumix builds without sysmouse support on my [FreeBSD]-CURRENT box, because <machine/console.h> has become finally obsolete.

Pixel, a developer from mandrakesoft, has authored the following patch, because we'll use devfs by default for the upcoming 8.1.

I saw in the docs that if you use [Linux's] devfs, you have to specify "-d /dev/sound/mixer" on the command line. How about automatically trying that one if opening the standard /dev/mixer fails? Something like the patch below should do it. The logic used is if no device is explicitly given as a parameter, to first try /dev/mixer, and if that fails, then try /dev/sound/mixer (only on [Linux] systems).

I use the DISPLAY var in profile and get problems running aumix (interface=IN_GTK depends [on] getenv(DISPLAY) ). This my little patch resolves my problem :-).

...so that stdin/stdout is used as the load/save file. It is not tested...

...the stuff I (and my predecessors) improved in the Debian package. I stripped the Debian-specific stuff out of the patch...

...translation for Simplified Chinese. It is for the latest 2.7 version. I have tested it on Mandrake 8.1, all works fine except the channel names are still in English. The file use GB2312-1980 encoding, and $ iconv -f gb2312 -t utf8 .... can convert it to UTF8.

a Dutch translation

a patch making aumix quite useful on terminals with only 10 characters on a line. As I do not a have a mouse, I have not changed the mouse event handling code. PlaceCursor() routine is slightly changed, as are some timeout()'s.

improved Swedish translation

Debian bug 139410

fixed option parsing bug, where aumix handled load and store before scanning for a file name

This was also reported by Pedro Zorzenon Neto as Debian bug 140610

Newer autoconf requires that arguments with internal whitespace be quoted, so we patch up configure here for when the makefiles run autoconf later.

[in the German translation] the missing charset specification leads to the effect that umlauts are not printed correctly in a locale where the characters do not match the encoding in the PO file. Please correct the header of the PO file to this effect (don't forget to unfuzzy it). You might also just take the attached new PO file.

Debian bug 138892

a Dutch translation, Debian bug 111300

machine/soundcard.h was only a compatibility symlink to sys/soundcard.h and has just gone away in -CURRENT, cf. PR #36966

Brazilian Portuguese translation (found in aumix-2.7-9cl.src.rpm).

The aumix 2.7 release has a bug which makes it blind for digital-input devices, at least the spdif on my TB-Fiji.

The reason for this behavior is that aumix assumes that every device must have a slider, which is not true for digital inputs, where level-adjustment is done by the digital source device.

I did a patch wich corrects this problem with my card, it may work with other spdif-soundcards too. It works with gtk and ncurses including saving/loading aumixrc. Querying devices in noninteractive mode works too. Selecting a digital recording device in noninteractive mode is not supported by aumix. In gtk mode a dummy slider is shown which resets to zero after adjusting.

fixed to compile with new autoconf