Re: [gpsd-dev] gpsfake, gpsd, printing, throttling?

[Gary E. Miller]

Yo Jiwon!

On Thu, 18 Apr 2019 04:43:02 -0400
Jiwon Joung <address@hidden> wrote:

> I ran gpsfake w/  vanilla gpsd and it  was successful.

Cool.  So gpsd is good.

> I edited gpsd to include some printf's and some dummy calls to a
> function in an  external library.
> Now I get the error:  "gpsfake: socket error [Errno 104] Connection
> reset by peer.".

Hard to debug your code for you without seeing your code, or even a
good debug log from you.

BUT: printf() is a bad idea.  gpsd is multi-threaded, and stdout is not.
Best to use gpsd_log().

> However, this happens  after a couple dozen calls  to my dummy calls
> (args to the call don't change and yet the first few calls to my
> function succeed...)

Many ways to break sockets...

> Any thoughts on why this might be the case? I'm kind of at a  loss.

I've got a hundred ideas on how you broek it, but no data to back up any
of them.

As always, when in doubt, run gpsd with a lot of debug.  To do
that with gpsfake:
        # gpsfake -o "-D 4" testdata.log

That is likely too much data, too fast, so run it with nohup to get a
file of the output.

> Is there some timeout in gpsd?

All over the place.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can’t measure it, you can’t improve it." - Lord Kelvin

pgp7Tb0iEiu1W.pgp
Description: OpenPGP digital signature