[Tinyos Core WG] Meeting: September 27

[Prabal Dutta]

Some thoughts on TEP 117

- I2c -> I2C

- Need reference to where tri-state covered.

- Would grouping pins as ports be useful?  Would this be in a
hardware-specific TEP?

- Should GPIOInterrupt be an optional interface?  Have we seen any MCU
w/o interruptable pins?  Conversely, have we seen any MCU w/o capture?

- Some interrupts are level (not edge) triggered.  Is the current
abstraction the right one for level-triggered interrupts?
Level-triggered interrupts are useful for sharing the line among
multiple peripherals whereas this is difficult/impossible for
edge-triggered ones.  Perhaps a GpioLevelInterrupt and
GpioEdgeInterrupt?

- Does the caller of GpioInterrupt need to keep track of which edge
will cause the interrupt?  The "fired" event does not provide this
information and there are no accessors to query this state.  Should
either fired provide the edge state or should an accessor be added?

- S2.3, L5: awkward: "The interface makes not declaration of the precision "

- There should be a note about the timebase that the following is referenced to:
    async event void captured(uint16_t time);

- Is this correct? (should STOP -> START?):
I2C_START    - Transmit an I2C STOP at the beginning of the operation.

- It feels like there isn't enough context around the SPI (and I2C,
UART) sections.  I would suggest that these be moved to other TEPs
that can provide a more complete reference.


On 9/26/06, Philip Levis <pal at cs.stanford.edu> wrote:
> On Sep 26, 2006, at 10:32 PM, Jonathan Hui wrote:
>
> >
> > Checked in to the devel branch.
>
> OK. Just saw it. Here is the updated 117.
>
>
>
>
> Phil
>
>
> _______________________________________________
> Tinyos-2.0wg mailing list
> Tinyos-2.0wg at Mail.Millennium.Berkeley.EDU
> https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-2.0wg
>
>
>
>