Hi Mehmet,

On Fri, Sep 11, 2009 at 10:08 AM, Mehmet Akif Antepli
<[email protected]> wrote:
> Hi Miklos,
> Sorry for my late reply.
> i had a chance to try your test code several times, but the leds never
> toggled,  which means
> ArbiterInfo.inUse() == true, it never turns out to be false!
> Then the overall code is the same as in rev1.10. But my network works quite
> well !  I did not come accross with any sudden disconnection problem that i
> mentioned in my first e-mail. That's so weird, isn't it?

Yes indeed it is very weird. There is a possible explanation: the code
path got shorter with those two lines removed which makes the SPI data
transfer faster which could cause some weird race condition somewhere.
Would be very hard to debug that, but if that is the case then my
comments in the file seem to be wrong. I would really appreciate if
you could put together some stripped down test case which shows the
problem (it is not too urgent, but I would like to get to the bottom
of this).

> When I use rev.1.9,  the problem occurs again.

I am confused, the rev 1.9 is the original that is supposed to be
working the same way as rev 1.11

> Anyhow, i noticed that the new rev.1.11 on the CVS, it seems to work for
> now,

Yes, it is reverted to the rev 1.9 behavior.

Miklos

_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to