On 2014-02-01 16:59, Js wrote:
Jef Driesen <[email protected]> wrote:
On 2014-01-28 16:11, Dirk Hohndel wrote:
I wonder... Didn't the header length change? If he now has the new
firmware with the longer header, will the old dives be downloaded
with
the new or the old header?
That could cause parsing errors...
Yes, there are two variants with a different length, but there is a
version number embedded in the header to figure out the correct
length.
but the header looks correct to me. I didn't notice any problem there.
The problem is in the profile data. When parsing the section with the
sample divisors, the corresponding sample lengths are all wrong. For
example a temperature sample is supposed to be 2 byte long, but the
length stored in the data is 3 bytes. The libdivecomputer parser fails
there already, but if modify the code to accept 3 bytes, it just fails
at some later stage.
The fact that the ostc itself also has problems displaying these
dives,
is for me a strong indication that there is a problem with the data
rather than the parser.
I managed to download the dives to diving log 5.0 and from there
upload to en.divelogs.de then download to subsurface.
This makes me wonder why divinglog can parse the data successfully and
we (and the ostc itself) can't. Strange.
Jef
_______________________________________________
subsurface mailing list
[email protected]
http://lists.hohndel.org/cgi-bin/mailman/listinfo/subsurface
