First off: I can't be on IRC these days, because my Laptop broke. First the LCD went blank (circumvented that by using an external monitor), but now the gfx card broke - the screen's completely garbled. The laptop is currently in service and as soon as I have it back and repaired (cough I did a little investigation on the laptop on my own and hope they won't find out and void my warranty :P) I'll be on IRC again. Currently I'm writing from my Dad's laptop.
Now to my suggestion: I am not 100% sure, but as far as I know all that slider-volume-adjusting thingy in the Audio Wizard is only for voice activation (va), right? For new users, especially not computer-savvy people, all that configuring stuff might be a bit of a burden, they might feel mumble is too complicated to use, in comparison to e.g. TeamSpeak (fire it up, set a PTT key/va, connect and talk).
Therefore I would propose to remove that part from the startup Audio Wizard, so there is only the sound device selection and Audio Path detection/adjustment left. That would reduce the amount of fiddling needed right at the beginning of using mumble.
Of course we can't leave that part out completely. So I would suggest to have a little config option, which saves if the va Wizard has already been launched. If that's not the case (which would be the default option) and the user selects va instead of PTT, then the va Wizard could pop up, as soon as the "Apply" Button has been hit. When the va Audio Wizard has been started the first time, the config option will flip its status and the wizard won't pop up again in the future.
As I think that va is the default option in mumble, this would also require to change the default setting to PTT instead of va, of course. Users could then either figure out their way to the config screen to set their PTT key, or you add one more page to the startup Audio Wizard which asks for the PTT key (and no other key, that'd be too much fiddling again).
Think about it. I think this might make it a lot more comfortable for new people trying out mumble :).
What about a config wizard that first asks if the user wants to use ptt or va and then moves on accordingly?
I like the current practice of Mumble leading a new user through the wizard on the first start.
There is really not much you can do with a voice chat application that is not propperly configured.
But I agree with your point - starting Mumble for the first time may be an intimidating experience for casual users.
"What about a config wizard that first asks if the user wants to use ptt or va and then moves on accordingly?"
-> Yes, that sounds perfectly good to me (add continous mode to the list)!
However, if the user chooses to use ptt, mumble would still have to prompt for the va wizard if the user should decide to switch to va later on.
Regarding your statement about configuring everything @ startup: Well, mumble IS completely functional without the va setup, if you use ptt or continous mode.
As I am working in web development, I observed a similar movement there. About 2 years ago (and before), if you wanted to register at a new website, you were always presented with the complete profile form and you could fill out what you want. Nowadays (say "Hello" to Web2.0 :P), registration forms of new webservices often consist only of 4 fields: Nickname, password, email, email confirmation. Although there are many more profile options available, you can usually set them later. This is to make the entry easier. You want to get in as quick as possible - not filling out tons of forms and read lots of text.
On a completely different note: Does the audio path detection work properly? The ms detection seems to be ok, I have different results on different PCs/Laptops. But on every machine I can set the slider to the minimum (40ms) without any interruption in the sound (even on a real crappy onboard sound chip). The only exception is my Gentoo Laptop on which the tone stops playing completely (not being jittered, it just stops) at 40ms. So I'm not quite sure if that slider and soundfile is working correctly... But that'd probably rather be another thread and/or bugreport.
No, the slider volume adjust thingy isn't only for VA, it's also for the automatic gain. So without going through it, audio actually sounds worse. Possibly much worse. So it's still a necesarry step.
I also would like to keep VA the default; the VA isn't perfect but I would rather see someone work on it than change to PTT per default. PTT hinders natural communication and stops social interaction, whereas VA encourages spontaneity and unrestricted speech.
"No, the slider volume adjust thingy isn't only for VA, it's also for the automatic gain. So without going through it, audio actually sounds worse. Possibly much worse. So it's still a necesarry step. "
Alright, if that's the case, the whole discussion makes no sense anymore. Topic done :P
And I don't mind VA being the default... I just suggested changing it, because I thought that we could avoid that amount of configuring upfront.
I agree with you. Many 'noobs' who configure Mumble are a bit overwhelmed when they first install it. It's no big deal for you and me, but layman's have no clue.
I agree, I live with 'noobs'. (I hope they don't find this)