AW: OSTC extra info (deco model and deco model info) not downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)

Fri Apr 29 09:37:17 PDT 2022

Give me 1 good reason why I should sideload something, when subsurface is working all right for me? Never change a running system…

________________________________
Von: harv at ruin.nu <harv at ruin.nu> im Auftrag von Michael Andreen <harv at ruin.nu>
Gesendet: Friday, April 29, 2022 6:29:26 PM
An: subsurface at subsurface-divelog.org <subsurface at subsurface-divelog.org>; Martin Gröger <martingroeger at hotmail.com>
Cc: subsurface at subsurface-divelog.org <subsurface at subsurface-divelog.org>
Betreff: Re: OSTC extra info (deco model and deco model info) not downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)

Not sure what's going on at your end. Only thing I can recommend is that you try a recent daily build and check that it still works.

https://subsurface-divelog.org/downloads/daily/

/Michael

On Friday 29 April 2022 17:49:41 CEST Martin Gröger wrote:
> sorry - but I see no bug. subsurface does what I want it to.
> is shows me the data of my ostc and my shearwater as it should. so what....
> ________________________________
> From: harv at ruin.nu <harv at ruin.nu> on behalf of Michael Andreen <harv at ruin.nu>
> Sent: Friday, April 29, 2022 2:43:06 PM
> To: subsurface at subsurface-divelog.org <subsurface at subsurface-divelog.org>
> Cc: subsurface at subsurface-divelog.org <subsurface at subsurface-divelog.org>; Martin Gröger <martingroeger at hotmail.com>
> Subject: Re: OSTC extra info (deco model and deco model info) not downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)
>
> The bug was in Subsurface. The OSTC is reporting the right thing, no one is using sat/desat, but Subsurface 5.0.7 and 5.0.8 reads some unrelated information instead of the deco model. It reads OC and thinks that means ZH-L16 and when it reads CCR it thinks that means ZH-L16-GF.
>
> You can check the fix yourself if that helps:
>
> https://github.com/subsurface/libdc/commit/5f3a0f0b8a951030cbeb61b7b2d3416b02229bc7
>
> /Michael
>
>
> On Friday 29 April 2022 14:00:48 CEST Martin Gröger wrote:
> > why should my ostc report a sat/desat if I'm gonna use GF instead? and that is what I see in subsurface and so all good!
> >  the use of sat/desat is "possible" but if you want to plan an non-ndl dive with common software (subsurface, multideco,...) the sat/desat settings are rubish and leads you to...
> > ________________________________
> > From: harv at ruin.nu <harv at ruin.nu> on behalf of Michael Andreen <harv at ruin.nu>
> > Sent: Friday, April 29, 2022 8:52:42 AM
> > To: Martin Gröger <martingroeger at hotmail.com>
> > Cc: subsurface at subsurface-divelog.org <subsurface at subsurface-divelog.org>
> > Subject: Re: OSTC extra info (deco model and deco model info) not downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)
> >
> > There was a bug in how Subsurface parsed the ostc log and it had nothing to do with saturation/desaturation setting on the dive computer. Everyone got "Saturation 77, Desaturation 78" because the index was used instead of the data itself. That's an even older bug, so likely no one using Subsurface is actually the non-GF algorithm on their OSTC.
> >
> > If you actually do get GF extra info for OC dives with 5.0.7 and 5.0.8 then I would recommend that you test with a daily build of Subsurface, since that indicates that something is off in the logfile.
> >
> > /Michael
> >
> > On Friday 29 April 2022 05:57:01 CEST Martin Gröger wrote:
> > > yep - absolute shure
> > >
> > > but _maybe_ it is so because I have "saturation/desaturation" set to 100/100 (as it should be if you use GF). to use both setting in my eyes makes no sense. IF you are using both settings maybe the sat/desat "overrules" the GF in the aditional info. that what it looks like. but it could be cleared in the HW-forum....
> > > ________________________________
> > > Von: Michael Andreen <harv at ruin.nu>
> > > Gesendet: Donnerstag, 28. April 2022, 22:31
> > > An: Martin Gröger <martingroeger at hotmail.com>
> > > Cc: subsurface at subsurface-divelog.org <subsurface at subsurface-divelog.org>
> > > Betreff: Re: OSTC extra info (deco model and deco model info) not downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)
> > >
> > > That's odd and shouldn't be possible. Are you sure you actually downloaded the dive with 5.0.7 and it wasn't downloaded from the dive computer with a previous version and then just displayed with 5.0.7?
> > >
> > > With 5.0.7 and 5.0.8 on OC you should get "ZH-L16" and "Saturation 77, Desaturation 78". And CCR you should get "ZH-L16-GF" and "GF 40/80".
> > >
> > > With older versions, and current master, you should get "ZH-L16-GF" and "GF 40/80".
> > >
> > > Anyway, the bug is fixed so future versions should get the right information in either case.
> > >
> > > /Michael
> > >
> > > On Thursday 28 April 2022 22:07:37 CEST Martin Gröger wrote:
> > > > hm...
> > > >
> > > > interesting, 'cause I dive OC and the extra info says
> > > > "ZHL16" and "GF40/80" as my ostc is set
> > > > (subsurface 5.0.7 /hwos 3.17)
> > > >
> > > >
> > > >
> > > > ________________________________
> > > > From: harv at ruin.nu <harv at ruin.nu> on behalf of Michael Andreen <harv at ruin.nu>
> > > > Sent: Thursday, April 28, 2022 7:19:55 PM
> > > > To: Martin Gröger <martingroeger at hotmail.com>; Subsurface Mailing List <subsurface at subsurface-divelog.org>; JB2Cool <jb2cool at gmail.com>
> > > > Subject: Re: OSTC extra info (deco model and deco model info) not downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)
> > > >
> > > > I don't have access to the screenshot, but it shouldn't be related to the OSTC firmware. The core problem was a bad merge between changes to the OSTC parser in Jef's master branch and the Subsurface specific additions in the Subsurface-DS9 branch. In short the dive mode (e.g. OC or CCR) got mixed up with deco model (old ostc buhlman or gradient factors). If Martin is diving CCR then he would get gradient factors in the log, while anyone diving OC would get the old deco model (with hardcoded saturation and desaturation parameters).
> > > >
> > > > /Michael
> > > >
> > > > On Thursday 28 April 2022 18:48:14 CEST JB2Cool via subsurface wrote:
> > > > > Martin,
> > > > > What OSTC firmware are you using? I wonder if you'll see this if you
> > > > > upgrade to 3.18 i think that's when i started seeing this,
> > > > >
> > > > > Jason
> > > > >
> > > > > On Thu, 28 Apr 2022 at 17:32, Martin Gröger <martingroeger at hotmail.com>
> > > > > wrote:
> > > > >
> > > > > > Hm…
> > > > > >
> > > > > > I can’t confirm that issue (see screen)
> > > > > > data downloaded subsurface desktop 5.07/OSTC plus
> > > > > > deco model & gf all ok
> > > > > >
> > > > > > Keep on howling
> > > > > >
> > > > > > grey
> > > > > >
> > > > > >
> > > > > >
> > > > > > *Von: *JB2Cool via subsurface <subsurface at subsurface-divelog.org>
> > > > > > *Gesendet: *Montag, 4. April 2022 10:22
> > > > > > *An: *Subsurface Mailing List <subsurface at subsurface-divelog.org>
> > > > > > *Betreff: *Re: OSTC extra info (deco model and deco model info) not
> > > > > > downloading correctly in newer Subsurface releases (5.0.7 and 5.0.8)
> > > > > >
> > > > > >
> > > > > >
> > > > > > Someone replied over at the Heinrichs Weikamp forum (
> > > > > > https://forum.heinrichsweikamp.com/viewtopic.php?p=23678#p23678) saying
> > > > > > that it looks like libdivecomputer is using the wrong memory addresses for
> > > > > > this information (this was correct previously). A pull request (
> > > > > > https://github.com/subsurface/libdc/pull/32) has been generated that
> > > > > > hopefully reverts this.
> > > > > >
> > > > > >
> > > > > >
> > > > > > I'll try to build locally with those adjustments and see if this
> > > > > > straightens things out for me.
> > > > > >
> > > > > >
> > > > > >
> > > > > > Jason
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Sun, 3 Apr 2022 at 20:33, JB2Cool <jb2cool at gmail.com> wrote:
> > > > > >
> > > > > > Good evening,
> > > > > >
> > > > > > I’ve downloaded some dives today and realised the 'extra info' on my OSTC
> > > > > > is not downloading correctly. I originally reported this to Heinrichs
> > > > > > Weikamp here (https://forum.heinrichsweikamp.com/viewtopic.php?t=22302)
> > > > > > as these were the first dives I’ve done on this new OSTC firmware but
> > > > > > subsequent troubleshooting makes this look like the issue is at the
> > > > > > Subsurface or libdivecomputer end.
> > > > > >
> > > > > >
> > > > > >
> > > > > > I’ve downloaded some older AppImages and this works fine on 5.0.5 and
> > > > > > 5.0.6 but shows the undesirable behaviour on 5.0.7 and 5.0.8.
> > > > > >
> > > > > >
> > > > > >
> > > > > > Extra info on dives with older versions of Subsurface look like:
> > > > > >
> > > > > >
> > > > > >
> > > > > > FW Version 3.17
> > > > > > Deco model ZH-L16-GF
> > > > > > Deco model info GF 30/85
> > > > > >
> > > > > > Now with the newest 2 versions of Subsurface extra info looks like:
> > > > > >
> > > > > > FW Version 3.18
> > > > > > Deco model ZH-L16
> > > > > > Deco model info Saturation 77, Desaturation 78
> > > > > >
> > > > > > It’s still doing this for dives that have downloaded fine in the past
> > > > > > where I know the dive was completed with the gradient factor model. This
> > > > > > only seems to affect newly downloaded dives, if something was downloaded
> > > > > > correctly in the past on a ‘good’ version then it will still display
> > > > > > correctly on a ‘bad’ version but if you were to download the dive again on
> > > > > > a ‘bad’ version then the issue would present itself (I may not be making
> > > > > > myself clear here).
> > > > > >
> > > > > >
> > > > > >
> > > > > > I've hopefully captured the relevant logs/dumps that will help with
> > > > > > troubleshooting.
> > > > > >
> > > > > >
> > > > > >
> > > > > > If I can provide screenshots or more information I'm of course happy to do
> > > > > > that.
> > > > > >
> > > > > >
> > > > > >
> > > > > > Regards
> > > > > >
> > > > > > Jason
> > > > > >
> > > > > >
> > > > > >
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20220429/324a7512/attachment-0001.htm>