more ways not to help Debian release
joeyh, the next time you want to make an example of someone (not that there is anything wrong with this, I do it all the time...) please first do your homework and be sure to understand the issue. You rightly talk about communication, and this could have been quickly straightened out if you had just asked about it on IRC. So I have to write this instead of fixing real bugs in hotplug.
Let's start from getting the basic facts straight: hotplug 0.0.20040329-17 did not introduce any change related to firmwares loading, nor is it related in any way to possible fixes for #297481. This means that there was nothing wrong with this version of hotplug and no reason to keep it out of testing.
As I already explained in a previous post, this is a kernel bug exposed by udev 0.054-1 which I uploaded more than two weeks ago. The package is not in testing and there is no sign that it will migrate soon, because it's being held in the queue of some doorstop build daemons. Anyway, I consider this bug much less important than the other changes in the package which will fix real bugs in sarge, and this is why I did not open an RC bug to keep the new udev packages out of testing.
#297481 does not mean that firmwares loading is broken, it only applies to a very small number of drivers. Actually, I only know about only one driver exposing this problem and it will not be part of the release, so there are no RC-related concerns at all.
I have already written the code for a workaround, but before adding it to the package I would have liked to know if the sarge kernel packages were going to be fixed. I think this was a reasonable request, but considering the silence from the kernel team I think I will add the workaround to the next upload.