New Bug Reports/Feature Requests

[Miika Turkia]

I think I see the bug in XSLT transform, naturally hidden in common
templates and not on the actual file doing the conversion to
divelogs.de format. Looks like using only one decimal has been there
from the beginning. Makes me wonder if Richard has just been lucky
with an earlier version of Subsurface. Anyway, I'll send a patch
shortly for this.

miika

On Wed, Feb 24, 2016 at 12:27 PM, Rainer Mohr <mail at divelogs.de> wrote:
> OK, found the uploaded DLD file.
> This dive: https://en.divelogs.de/dive/1659944
> has <WEIGHT>9.1</WEIGHT>in the XML
> So converting this to lbs: Divide by 0.4536 gives me 20.06 lbs, which is
> diplayed as 20.1 lbs (I do round to one digit for displaying).
>
> So now the question is: At which point did the 9.1 get its value for putting
> it into the XML. If the subsurface conversion from the entered lbs to stored
> metric uses the same factor (which I do believe), it should be 9.072 kilos.
>
> I guess your subsurface file would now come in handy to check whats in
> there.
>
> Thanks,
> Rainer
>
>
> Am 24.02.16 um 11:13 schrieb Richard Houser:
>
> My id there is DiverMidMi, so feel free to look at the whole logbook. All my
> logs before this February display the values I see in Subsurface. Those were
> initially entered in 4.4.2. The new dives from this month show the excess
> weight. I can supply the actual Subsurface xml entries after I get home
> (probably tomorrow, last plane lands in a blizzard).
>
> The computer dives were one of these depending on the date:
>
> A300 CS -> ACI (Aeris proprietary app) -> CVS -> Subsurface -> export to
> divelogs.de option --- my first 10ish dives in Florida last year. I did have
> to do some really minor cleanup to the original XML (minutes vs. Seconds on
> samples if I recall)
>
> A300 CS -> Subsurface > export to divelogs.de (switching from Mageia
> Subsurface 4.4.2 to Subsurface 4.5.3 Appimage before the dives this month.
>
> On February 24, 2016 3:49:42 AM EST, Rainer Mohr <mail at divelogs.de> wrote:
>>
>> Am 24.02.16 um 07:42 schrieb Dirk Hohndel:
>>>
>>>  On Wed, Feb 24, 2016 at 08:10:04AM +0200, Miika Turkia wrote:
>>>>
>>>>  On Wed, Feb 24, 2016 at 12:52 AM, Dirk Hohndel <dirk at hohndel.org>
>>>> wrote:
>>>>>
>>>>>  On Tue, Feb 23, 2016 at 01:55:28PM -0500, Richard Houser wrote:
>>>>>>
>>>>>>  #4 Dives exported from subsurface to divelogs.de are now showing an
>>>>>>  additional 0.1lb more than what they show in subsurface.
>>>>>
>>>>>  Rounding error somewhere. Silly American
>>>>> pounds.
>>>>
>>>>  Both Subsurface and divelogs.de work in metric units internally. I
>>>>  just did a quick test and 5lb turns out to be 2.268 kg in Subsurface
>>>>  and is displayed as 2.2 kg on the divelist. Divelogs.de displays this
>>>>  as 2.3 kg (as is correct rounding). However, it seems that divelogs.de
>>>>  uses this rounded number when displaying the imperial weight (5.1
>>>>  lbs). What has changed, I have no idea, but Rainer might be able to
>>>>  clarify. (Us showing incorrect metric rounding is irrelevant regarding
>>>>  divelogs.de import/export as we use the exact value in calculations
>>>>  and export.)
>>>
>>>  That's odd, Rainer. Why would you convert the rounded value?
>>
>> I don't...
>>
>> Just entered a weight of 2.268 kilos in a dive at divelogs.de (which
>> I
>> do round to 2 digits, as I never thought a few grams would be of
>> importance).
>> This results in 2.27 kilos in the database and diplays as exactly 5 lbs
>> on the dive after setting my logbook to imperial. I have the feeling,
>> that the dive had 2.3 kilos when it got into divelogs.de (as this would
>> explain the 5.1 lbs). I don't do any additional rounding when reading
>> the XML of the DLD files, so I'd like to check what got exported to
>> divelogs.de
>>
>> Richard, could you please give me a link to the dive? How did the dive
>> get into divelogs.de? The exact path of where the dive went in its
>> digital life would help figure out where the rounding got screwed up.
>>
>> Thanks,
>> Rainer
>>
>
> --
> Sent from my Android device with K-9 Mail. Please excuse my brevity.
>
>