possible O2 PP display bug

[Dirk Hohndel]

Jan Schubert <Jan.Schubert at GMX.li> writes:

> On 02.03.2013 17:42, Dirk Hohndel wrote:
>> On Mar 2, 2013, at 8:38 AM, Jan Schubert wrote:
>>
>>> On 02.03.2013 17:23, Dirk Hohndel wrote:
>>>> Can you build from source (if not, tell me what platform you are on and I'll get you a binary)? The latest git should fix this - but it requires you to re-download from your divecomputer, I'm afraid.
>>> Dirk, shouldn't be that tricky to autofix this in the import? Will just
>>> have a look..
>> The case that should be easy is the one where 1.009 turns into 1.9 - you have a space there. But the problem here is that you can't tell if 1.90 came from an actual 1.90 or from a 1.090 :-(
>
> Things gets really complicated now as the fix provided by Linus seems to
> make things different than before. In the "past" we would store 1,3bar
> as "1,300" which is now only "1,3". I was assuming that we just filter
> out everything which maches the format "1.xx" and correct it with
> "1.0xx" but this is not possible anymore. I alreday had a modified

Now you cannot do that. Because 1.xy would also be 1.xy if x > 0

> double_to_permille funtcion bit I give up after recognizing that my
> separated test dive containing a spike looks different in the xml as the
> original one. So indeed, there is a need to reimport all dives and no
> chance to get these spikes fixed automatigacally for dives which are not
> on the dive computer anymore. There might be a chance for a context
> sensitive fixing by considering the previous and following PO2 but this
> is really not worth it.

Yep

/D