Slashdot post but no web store?

[Kai Römer]

2008/6/27 Marcus Bauer <marcus.bauer at gmail.com>:
[...] cut a lot!

>> Our tests on GPS TTFF where quite comparable with some of the best devices
>> available. However note there is some data GPS *has* to download from sat
>> for "virgin" fix, which may take as long as 12min due to low bandwidth the
>> sats offer (just takes this time for them to send one complete set of ephem
>> and alm),
>
> Under good conditions all modern GPS chips have a cold start TTFF
> ("virgin fix") of ~45secs. 12min is the value for 8 to 10 year old
> chips.
>
>> On tests done at Taipei the TTFF was around ~40sec without(!)
>> AGPS, IIRC.
>
> All reports on the MLs spoke about minutes or not getting a fix at all.
>
>> So this is a clear sw-issue.
>
> What could be the problem then? Usually you power the GPS up and you
> listen on /dev/ttySAC1 for the NMEA sentences. That's how it works on
> the Neo1973 and every other device. What software tweaking can be made?

So far i have tested 6 OpenMokos in Germany. With the internal antenna
i never got a fix in less than 1min 20 secs if ever (2 hours) (Perfect
conditions/ sirf3 need 36 sec cold boot). Even when i use AGPS. As
soon as i attach a external antenna it easily finds a lot more
satellites and connects withing 40 to 50 seconds.  So i think its
neither a software issue nor a satellite visibility problem but a
problem with the internal antenna or the wiring to this internal
antenna.

Can someone test this too and affirm my results?