View Single Post
Posts: 631 | Thanked: 837 times | Joined on May 2007 @ Milton, Ontario, Canada
#125
Originally Posted by qole View Post
I was thinking of a replacement USB kernel driver that would ignore / override the HW signal when you told it to go into "unsupported host mode". It would have to be 100% software controlled (the user would have to say, "go into host mode" and "go out of host mode") and probably charging and power supply would not be available... but would this kind of software hack be possible?
Might be possible but I'd be very skeptical realistically of that possibility; my experiences with low level hardware like that is that while software might have an option to adjust to tweak a setting on a chip, a hardware pin that's designed to perform the same sort of functionality always overrides since it is usually bound to the physical gate itself. I.e. if you have software to enable "host mode power" on the 5V line, then the software acts as a trigger for the hardware transistor/switch to make it happen; if that same transistor/switch has a hardware input to perform the same function, then using the software to "override it" doesn't work because it can't "undo" the hardware state. Even if you could get a driver to "ignore" that state, you'd probably still have a hard time getting the rest of the OTG host functionality to actually enable... just my thoughts though, I haven't studied the data sheets on the USB chipset though...

My assumption here is that the root cause of this is of course the USB charging thing. If you think about it, they probably had exactly the same issue with the N800 in that they couldn't say OTG because there was no software support to be "approved by the USB consortium" at time of launch, so they stuck with the old Mini-B connector; hardware still worked, just had the wrong connector and no official support (basically fly under the USB consortium radar and say "hey guys, if you can make it work, sweet, but we didn't know anything about it wink wink nudge nudge"). On the N900 it's a different story though because of the stupid USB charging... you can't just leave it all wired up for OTG Host mode, and have USB charging support without some proof that the software properly supports the OTG signalling; otherwise you could release the "OTG software update" to enable OTG later on, and since it no longer has to be verified/approved by anyone (technially... I'm sure Nokia would test things, but we're talking hypothetically here) then what if the software update is goofed and causes the chip to go into host mode while connected to charger/etc. Big problems ensue, and everyone who has a blown up device has the potential to turn around and say "Damn USB OTG blew up my device!".

So yea, really far fetched, but think about it from other more paranoid perspectives and you can see how it fits; therefore Nokia has to keep everyone happy by making it impossible for this to be updated later on.

(And as for the USB charging thing VS the 2mm jack... unfortunately that's a political thing too; Nokia signed on along with everyone else earlier this year as part of an EU innitiative I believe to standardize all mobile device/cell phone charges to be MicroUSB; good idea normally to save everyone a lot of money/hassle/cable messes/etc, but in this case a bit of a pain.. My biggest thought on this was "how would you power the N900 and use OTG at the same time?... powered hub, that stinks").

Just my take on things... oh and for the record I've got my USB OTG numbers for the N800 adapters... over 1000 sold... plus all the other ones people picked up from ebay, etc kind of means that there's at least a basis for OTG given the number of NITs relatively speaking...
 

The Following 6 Users Say Thank You to jolouis For This Useful Post: