Seabear dive computer integration in subsurface

[Jef Driesen]

On 2014-07-08 19:14, Dirk Hohndel wrote:
> On Tue, Jul 08, 2014 at 02:42:57PM +0200, Benjamin Kuch wrote:
>> I’m embedded software developer at Seabear diving technology. 
>> Currently we
>> have one dive computer on the market (Seabear HUDC). Three others will
>> follow soon. We would like to be integrated in subsurface. All our 
>> devices
>> have a mass storage mode in which the log data (.csv files) gets 
>> provided.
>> 
>> What has to be done to get integrated in subsurface?
> 
> Most dive downloading from divecomputers is done via libdivecomputer. 
> I'll
> let Jef respond how he envisions something like this could be 
> integrated.
> The other DC that I'm aware of that uses a storage mode based interface
> (the Uemis SDA) right now is integrated directly into Subsurface.
> 
> It depends a little bit on how exactly Seabear exposes the dive data to
> give a better answer. What's the directory structure in that virtual
> filesystem? How are the files named, what's the content?

With a mass storage mode, you don't really need libdivecomputer. You're 
not really downloading any dives, but importing files from disk. I think 
this is more suited to implement directly into the application as an 
file importer. Whether the file is stored on a regular harddisk or a 
dive computer with mass storage mode doesn't really matter. Trying to 
implement this as a libdivecomputer backend might be difficult, because 
libdivecomputer is designed to deal with communication protocols, not 
file imports.

The Uemis is different in this regard, because it (ab)uses the 
filesystem as a weird two way communication mechanism. I assume that's 
not the case here, and the Seabear device uses mass storage in a sane 
way. Similar to how the HW DR5 uses mass storage mode to export the 
dives as UDDF files.

Jef