Extremely slow Petrel import + crash when saving

Jef Driesen jefdriesen at telenet.be
Thu Sep 19 12:26:53 UTC 2013


On 19-09-13 19:45, Henrik Brautaset Aronsen wrote:
> Jef Driesen wrote:
>> When I analyze the logfile, I see a repeating pattern. There is always
>> a number of write's (to send a request to the device) followed by a
>> number of read's (the receive the response packet). The read sequence
>> is always very fast, in the order of just a few milliseconds for the
>> entire packet. But writing appears to be terrible slow, with about 1.5
>> seconds for just a few bytes.
>>
>> Look for example at the first dive (lines 3594 to 7560). It has a
>> (compressed) size of only 3460 bytes but takes approximately 47
>> seconds to download. Since the data is send in packets of 144 bytes,
>> there are 24 data packets (and two more for the init/quit packets).
>> Thus of those 47 seconds, roughly 39 seconds (26 times 1.5s) are spend
>> sending requests!
>>
>> I wonder if this is some platform or driver related issue. Can you try
>> the download on another platform (Windows or Linux) to check?
>
> I got it running on an old Linux laptop, but when the download starts,
> the Petrel quits with "slip rx: unexp. slip end"

Arghh. That reminds me that I'm having exactly the same problem. On Linux, I 
also get that error when trying to download from the Petrel. The only other 
bluetooth dive computer I have available, a HW Frog, also fails to download on 
Linux. I'm sure both used to work fine in the past, but not anymore. On Windows 
downloading still works fine (just tried). Very strange.

I wonder if there is some bug in the bluetooth serial port emulation? I assume 
the bluetooth stack itself is working fine, because I have problems exchanging 
between my linux laptop and android phone.

If anyone knows how to diagnose bluetooth issues, please get in touch!


Now back to your Mac OS X problem, this Linux failure doesn't get us any further 
of course. I'm afraid the only suggestion I have left is to try on Windows.

Jef


More information about the subsurface mailing list