Well the sleep cycles idea didn’t work out too well, but then I could have ruined it by relying on an alarm to wake me up.. Bit of a neccessity though cuz if I got up at 12 today I wouldn’t fancy trying to shift everything back 5 hours in one night to enable myself to get up for work (and not feel like shit) tomorrow at 7.30..
Anyway back to the matter at hand. You see, among the problems that have had to be overcome to get bluetooth going, there is the fundamental issue that Windows XP SP2 is not designed to be able to handle bluetooth communication with devices other than keyboards and mice. Then there’s also the fact that the chipset manufacturer, Broadcom, refuses to make the latest drivers publicly available, referring you to the overall product manufacturer, whose driver release is about 5 revisions out of date. Fantastic eh.
By default, when you install the drivers that come with a bluetooth dongle Windows automatically reverts back to the proprietary bluetooth driver. In other words it ignores what you just did. Perhaps non-savvy users would be beaten at this point, who knows, because now you have to select the M$ driver, choose update, then manually proceed to pick out the one you just installed, and voila. Actually if you want to have a decent chance of getting things working like this, you have to follow this kind of quirky routine, which did need doing..
Now this was good enough to get my Orange SPV C500 working with ActiveSync over bluetooth, but the headset just didn’t wanna know. So next problem was to try and source an updated WIDCOMM driver that would be compatible with my dongle. I already said that Broadcom don’t share these, but I managed to download them from somewhere in Taiwan, and guess my way through the self-extractor as all the characters showed up as question marks :)
This driver revision was 220.127.116.111, a fair bit higher than the manufacturer supplied 18.104.22.168, but the driver .inf file didn’t contain any information about my particular device. At this point I thought I was going to have to hack together the old driver file and the new one so as to fool the system into thinking it was appropriate, but this seemed like too much hassle so decided to keep searching.. Fortunately, through trawling some forum posts I managed to track down a working version based on driver version 22.214.171.1240. For both generic driver sets, the install involved installing the drivers, then killing the bluetooth service, renaming & patching some system files, and patching some application files (in the latter case there was some kind of licensing set up that refused to co-operate with my dongle, but fortunately patching took care of this :). Oh yeah and even after that I had to delete and reinstall the bluetooth COM ports and perform a raindance before my phone would talk to ActiveSync again. Fun & games huh? So, fu Broadcom. I still have your latest drivers, so stop pulling this kind of shit.
So why don’t things just work? I have to say that, given my vast troubleshooting experience, I’m not at all surprised when they don’t. Oh no, that would be far too much to expect. But would manufacturers in other industries get away with this behaviour? I don’t think so. It seems that if you want to be at the cutting edge you have to accept that you’re going to be an active beta tester for the hardware companies involved. Don’t expect much help from customer support either, because generally they’re no wiser than you are :) Would this whole scenario be avoided by switching to Apple though? If I had the money I would already be using a 15″ Powerbook, and am currently still trying to justify the future expenditure.
It’s no wonder that bluetooth hasn’t taken off to the extent predicted, but when it’s this much hassle to get a headset working, well, is it any wonder. Boo.