I can't say for sure, but your problem is either a) you can't have a second writer with your hardware or audio driver, whether it's esd or not or b) consolehelper gives the first user exclusive access, blocking out the second.
Now why on earth would I run two versions of esd when one is more than capable of accepting multiple concurrent connections? Why was esd written in the first place? so that "crappy modern audio hardware" could be shared amoung processes. So the logical and simple solution would be to start a single esd, not at the time the first user logs in, but as a permanetly running system deamon, and have all audio processes talk to esd. I.e esd provides the sound sevice missing from the original X-server architecture. If I recall that was Rasters original idea but somehow it got lost along the way. Mostly no doubt because certainly applications didn't support esd or decided that they prefered to invent their own alternate approach.
Reading through the other postings it appears that with a bit of "bashing" I might be able to get my system to function in this manner.
Providing a user interface that normal human beings can use is "so last century"? I think you may have misread my point...
In this specific case, Your point was that the ctrl-alt-f7/f8 solution meant two X servers permanently running which you implied was a bad thing. My point was, that the other alternate (text based) logins where done using ctrl-alt-f1/f2...f6, so if your argument is valid that ctrl-alt-f7/f8 is the wrong way to go about it then you must also be arguing that the mingettys running to support all those mostly unused tty logins was also the wrong solution. However with todays massive amounts of memory and the kernels ability to swap out un-used processes, who cares?
Not me actually. P.