call printfflush() Kevin
On Thu, May 8, 2008 at 9:57 AM, Nitin Sharma <[EMAIL PROTECTED]> wrote: > Hi Kevin, > > The setup works perfect now, except for one thing which i havent been able > to figure. Earlier version allowed us to flush the PrintBuffer timely like > 500 ms or so . However, in this one, How would i flush it when i have no > more to add to a buffer and the buffer hasnt been filled yet? > > Have been adding dummy messages to fill it up, but is there a better way? > > Thanks. > Nitin > > > > On Thu, Apr 24, 2008 at 6:40 PM, Kevin Klues <[EMAIL PROTECTED]> wrote: > > > Yes, that is right, but youshouldn't copy the MainC in the printf > > directory into tos/systtem otherwise the serial interface will alwyas > > be enabled even if you dont want / need it to be. Just leave it where > > it is in the print directory and it should be included only when you > > actually include this directory via your makefile. > > > > Kevin > > > > > > > > > > On Thu, Apr 24, 2008 at 4:32 PM, Nitin Sharma <[EMAIL PROTECTED]> wrote: > > > Thanks Kevin. > > > > > > I downloaded and updated the library, and the MainC in tos/system too. I > got > > > printf sending over to the serial without wiring the PrintfControl or > > > PrintfFlush interfaces since it was wired in MainC? Is that how we are > > > supposed to use it? > > > > > > Nitin > > > > > > > > > > > > On Fri, Apr 11, 2008 at 10:34 AM, Kevin Klues <[EMAIL PROTECTED]> wrote: > > > > So the existing printf library is broken in the sense that if it is in > > > > the process of flushing packets out the serial line it will fail all > > > > printf commands that are called on it. I have an alternate > > > > implementation that doesn't do this, but hasn't made its way into the > > > > repository yet because the interfaces it provides are somewhat > > > > different than the existing ones and would break existing code. The > > > > only 'real' reason it hasn't made it in yet is because I don't want to > > > > just throw it in there without any documentation as to how its > > > > different than the existing library -- I don't want people's existing > > > > code to break without at least some documentation on how to fix it. > > > > > > > > The new implementation is much better though, and will make its way in > > > > once I have time to write up the documentation. For now you can grab > > > > it from here though if you want. > > > > > > > > http://sing.stanford.edu/klueska/printf_test.tar.gz > > > > > > > > Kevin > > > > > > > > > > > > > > > > > > > > On Fri, Apr 11, 2008 at 8:19 AM, Nitin Sharma <[EMAIL PROTECTED]> > wrote: > > > > > I was just wondering how you guys manage flow control for testbeds > > > connected > > > > > to a logging PC using a USB hub or something. > > > > > > > > > > I am currently using Xubuntos for the PC. The Hub seems to wear more > i > > > use > > > > > it, and the behavior of the logged packets using Serial Logging like > > > Printf > > > > > library seems to be giving not-so-good responses. > > > > > > > > > > Any advices/suggestions to improve USB performances? > > > > > > > > > > Thanks. > > > > > > > > > > -- > > > > > Nitin Sharma > > > > > _______________________________________________ > > > > > Tinyos-help mailing list > > > > > [email protected] > > > > > > > > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help > > > > > > > > > > > > > > > > > > > > > -- > > > > ~Kevin > > > > > > > > > > -- > > > Nitin Sharma > > > > > > > > -- > > ~Kevin > > > > > > -- > Nitin Sharma -- ~Kevin _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
