update binaries for current libdivecomputer

[Jef Driesen]

On 2014-04-09 15:14, Robert Helling wrote:
> On 09.04.2014, at 11:40, Jef Driesen <jef at libdivecomputer.org> wrote:
>> On 2014-04-09 11:27, Robert Helling wrote:
>>> I just realized the problem in bug #487 goes away when I update to 
>>> the
>>> current version of libdivecomputer. Dirk, could you update the
>>> binaries on the webpage accordingly, in particular windows?
>> 
>> Which libdivecomputer version were you using? I suspect that the 
>> change that fixed your problem is commit 
>> b1574848b2c3ffcf7f0a8080c853453f5750ce2c. But this fix was already 
>> included in v0.4.1. So I assume your version is even older?
> 
> Hmm, I am not exactly sure. I am under the impression that when I
> wrote the bug report yesterday I had
> 
> Subsurface v4.0.2, built with libdivecomputer v0.5.0-devel
> (c5249861758305cff525a02e0a243f65b0750862)
> 
> of which the log says
> 
> commit c5249861758305cff525a02e0a243f65b0750862
> Author: Jef Driesen <jefdriesen at users.sourceforge.net>
> Date:   Fri Oct 4 16:37:24 2013 +0200
> 
> while the commit you mention predates it as
> 
> commit b1574848b2c3ffcf7f0a8080c853453f5750ce2c
> Author: Jef Driesen <jefdriesen at users.sourceforge.net>
> Date:   Mon Jun 17 22:09:28 2013 +0200
> 
>     Interpret the 7th bit of the event byte correctly.
> 
> Anyway, now that I am using
> 
> Subsurface v4.0.2, built with libdivecomputer v0.5.0-devel
> (6d4893a1467452d3a0e94e712c2d07b0597b9f3c)
> 
> I don’t see the problem anymore. It seems that the dive on which the
> download stalled has a double gas change event at 19:50 (identical
> event twice). I cannot see anything that mentions bailout.

If you go through the libdivecomputer history, then you'll notice the 
next commit (bed4d19acacdaba50a2c4d80f1c827aad0c406a9) causes a bailout 
event to be reported as a normal gas change event. So that's most likely 
the reason for the double gas change event, and thus also the failure 
with the older libdivecomputer.

> The original problem was reported by a friend of a friend that is
> using windows. I don’t have access to a windows machine so I cannot
> confirm myself that the problem is there with the currently offered
> binary.

I also can't check right now, but I suspect it will contain an older 
version, like v0.4.0.

Jef