devel September 2015

devel@libdivecomputer.org

3 participants
4 discussions

[PATCH] Skip unconnected ppo2 sensors
by Anton Lundin 09 Sep '15

09 Sep '15

The HwOS devices logs ppo2 sensor values even in Auto / Fixed SP mode, but they log them as reading zero, if there are no sensors there. This skips us reporting those zero values. Signed-off-by: Anton Lundin <glance(a)acc.umu.se> --- This shows one of the flaws with re-using the DC_SAMPLE_PPO2 value. Should we report unconnected sensors? how should a application actually handle them? Maybee we should look at the ppO2 value too, but for now this POC only looked at the mV values. src/hw_ostc_parser.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/hw_ostc_parser.c b/src/hw_ostc_parser.c index 6db9d27..954f6f7 100644 --- a/src/hw_ostc_parser.c +++ b/src/hw_ostc_parser.c @@ -885,6 +885,9 @@ hw_ostc_parser_samples_foreach (dc_parser_t *abstract, dc_sample_callback_t call value = data[offset + j]; } else { value = data[offset + j * 3]; + // Skip zero mV sensors, as they are most likely unconnected + if (array_uint16_le (data + offset + j * 3 + 1) == 0) + continue; } sample.ppo2 = value / 100.0; if (callback) callback (DC_SAMPLE_PPO2, sample, userdata); -- 2.1.4

2 7

Re: Mares Smart Apnea
by Jef Driesen 09 Sep '15

09 Sep '15

On 2015-09-04 12:25, Giorgio Marzano wrote: > Right now, I noticed a bug in the mares_iconhd_get_model function, line > 120 > of mares_iconhd.c, where sizeof should be actually strlen. > > if (memcmp (device->version + 0x46, models[i].name, /*sizeof*/ strlen > (models[i].name) - 1) == 0) { This is not a bug. The name field is a 16 byte array, not a string. The initialization with a sting literal is convenient. The remainder of the array will automatically get padded with zero bytes, which is exactly what we need. The correct fix is to add a new entry containing "Mares Smart". Then, the next question is: what's the correct model number? According to the model number in the memory dump (e.g. byte at offset 0), it's 0x10 or SMART: {"Smart", SMART}, + {"Smart Apnea", SMART}, {"Icon HD", ICONHD}, But then how do we distinguish the normal smart from the apnea variant? That will be important, because both appear to use a different data format. The strange thing here is that the normal smart already had a freedive mode. The main difference is that we now not only get a summary of each freedive in the session, but also a detailed profile. I wonder if there is some kind of format indicator in the header that we can use to distinguish between dives with the enhanced apnea format and the more limited normal format. Jef

2 7

Re: Mares Smart Apnea
by Giorgio Marzano 09 Sep '15

09 Sep '15

Again on the Mares Smart Apnea topic. I think I have found a problem on the current SW implementation and I need some advice. Sorry for the long message. I have dumped my DC logbook, and some cross checking between the code and the expected results from the mares SW. I think I have found some problem in the code. In the function "mares_iconhd_extract_dives" it seems to me that the code points to the very end of the buffer, then it starts crawling back to the beginning of the last immersion: offset = layout->rb_profile_end - layout->rb_profile_begin; unsigned int nbytes = 4 + headersize + nsamples * samplesize; ... // Move to the start of the dive. offset -= nbytes; ...but it seems that on the Smart Apnea there is not a constant samplesize: one immersion is a variable length (nsamples) sequence of dives , each one having one header (dive time, surface time, temperature, ...) and a stream of "dive time" depth measurements. If this is true, the current approach can't work, as long as it is no longer possible to compute "nbytes". Am I wrong? Is there any reason to start from the end and not from the beginning (the "end of profile ring buffer" eop) and then move forward? I am getting familiar with the memory layout of the DC but still have many many doubts about the library code and architecture. Any help would be very appreciated. I can provide the annotated dump, but I am not sure what the mailing list netiquette says about big attachments. Thank you in advance Giorgio EXAMPLE: *** start immersion 2015 08 31 10.08 18 25 : LEN (in bytes) 00 00 : ?? *** sample 1 *** 0c 00 : maxdepth(1.2) sample 1 13 00 : dive time (19 secs) 00 00 : surface time (0) fc 00 ?? fc 00 ?? 04 00 ?? 17 00 ?? 0b 00 depth //19 depth measurements: 1 per second 0c 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0c 00 depth 0c 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0b 00 depth 0c 00 depth 0b 00 depth 0b 00 depth 09 00 depth **** sample 2 *** 11 00 maxdepth (1.2) 2d 00 divetime 3c 00 surftime fc 00 ?? fc 00 ?? 05 00 ?? ...

2 3

Individual O2 sensor readings
by Anton Lundin 03 Sep '15

03 Sep '15

Nowadays Subsurface have a way to represent individual o2 sensor values, but none of the libdivecomputer backends export such. The diverite_nitekq and shearwater_predator backends have support for exposing DC_SAMPLE_PPO2 samples, which looks to be the calculated / voted PPO2. The hwOS-based HW computers have such sensor values stored, and probably the Shearwater ext devices to. I was thinking about implementing support for the hwOS based devices, and was thinking if anyone have some preference on how such values should be represented in libdivecomputer? Should we go for DC_SAMPLE_PPO2_SENSOR1 or something like that to expose the individual sensor values? Or do anyone else have some other, better, idea? //Anton -- Anton Lundin +46702-161604

2 6

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

devel September 2015