Divecomputer-reading from Mobile devices (Was: Divemate Fusion)

[Dirk Hohndel]

> On Apr 10, 2016, at 12:03 AM, Jef Driesen <jef at libdivecomputer.org> wrote:
> 
> On 10-04-16 02:20, Linus Torvalds wrote:
>> On Sat, Apr 9, 2016 at 5:15 PM, Dirk Hohndel <dirk at hohndel.org> wrote:
>>> 
>>> As expected, the individual dives run roughly 4k per hour dive time. This is with the libdivecomputer default XML output (to make my life easier I just used dctool for the download). I think our XML for example is much denser... so I think the amount of data will be manageable.
>> 
>> I'm not sure that we're _much_ denser, but yeah, we should be a bit better.
>> 
>> It might still be a good idea to just eventually compress the data
>> with zlib or something fairly trivial like that, but that's just a
>> small tweak on top of whatever protocol..
> 
> Or transfer the raw dive data, and do the parsing in subsurface? The raw data is likely the most dense format.

So many options. But you are right, that would almost certainly be the densest way to do this.

/D