Paladin of the Pegasos
Posts: 1178 from 2003/3/13
From: Pinto, Madrid ...
Quote:
platon42 wrote:
It is in a very prototypic state. It's got a good HCI layer, an ACL layer, a basic L2CAP layer...
I think I've grasped a bit of what you've explained here, and realized that one has to be a Bluetooth expert to understand the value of your job. Impressive.
Quote:
Ah yes, the stack is very cleverly optimized, needing very few memory copying operations, sometimes even zero copy between the lots of layers the data is moved through. It's highly asynchroneous, meaning that there's no messy blocking or waiting like in dead-stupid stacks like the Linux BlueZ.
This is interesting for the unavoidable, predictable situation of "we can't have Platon's stack, but hey, here's Linux to the rescue, let's port its stack". It's been said here many times, better live without some things for a while, than have them ported from alien systems.
Quote:
it's >460 KB of worthless source code, gathered in steps of a few weeks with gaps of 12 months over the last two years.
:-O
Quote:
No, I don't see sense in releasing it into public -- I doubt anyone would have the know-how to continue it. There are *very* few Bluetooth experts out there, and the probability that these would care for AmigaOS/MorphOS is 2^-65535.
Understood.
Quote:
Maybe at some point, I can turn this into some profitable commercial project on a non-Amiga system.
Perhaps that work for another, profitable system, could help in having this much expected Amiga and MorphOS version. I guess you can make it very platform independant.
Quote:
Or, I will indeed finish some "light version" of this for AmigaOS/MorphOS. But my guess is: not this year.
What worries me most is that, seeing how things develop in the tech gadget market, chances are that you might finish your Bluetooth stack when no one else uses it because some other new wireless standard is "hot".
