Re: [gpsd-dev] PPS over USB

[Ed W]

On 08/05/2012 09:33, Eric S. Raymond wrote:
> Ed W<address@hidden>:
> > I would have expected the kinds of problems you are describing above
> > to be more easily studied/tested using models in a lab?
> Please tell me you are joking, because otherwise you win the prize for
> the fscking stupidest thing ever said on this list.

No and I'm getting bored with your overtly hostile attitude.  The idiot 
is the one who thinks that having a repeatable lab situation is of *no 
use* to debugging a random non repeatable real world problem...  I think 
we all get that the real world doesn't match any model - doesn't make 
mathematics an entirely redundant discipline though.


> Models are only as good as their assumptions, and models of systems
> with lots of poorly-understood causal linkages (like, say, chaotic
> resonance in networks) are notoriously prone to becoming greased
> slides to la-la land that damage comprehension rather than aiding it.

Sure - but the real world isn't repeatable, so many people start by 
assuming the problem is spherical and moving in a vacuum...  Once that 
works you can try and deal with the vagueness of real world situations.


> So, hell no, we're not going to study models instead of measuring
> what's really happening.  That would be idiotic.

Yeah, really idiotic...

Well, just remember that the tools you need to process your data are 
based on a bunch of mathematic models...


I guess lets just get back on topic to gpsd since you like to pick 
fights with me and it generally seems to deteriorate to you calling me 
names.  I still owe you some results on the stability of the Venus gps 
NMEA, got some other urgent tasks, but hoping to collect those shortly.

Note to Gary: Chrony has additionally logging I just discovered, if you 
enable "refclocks" in your logging you can dump the uncooked 
measurements from your new PPS gps - this may be interesting for 
plotting stability graphs.  It appears if you just use the statistics 
plot then you are plotting the averaged values


Good luck

Ed W