<p dir="ltr">I already started cleaning the preferences, should be fixed this week. </p>
<div class="gmail_extra"><br><div class="gmail_quote">Em 30 de jul de 2016 16:27, "Dirk Hohndel" <<a href="mailto:dirk@hohndel.org">dirk@hohndel.org</a>> escreveu:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi<br>
<br>
So the discussion a little while ago seemed to focus on two points:<br>
<br>
a) continue the desktop app in maintenance mode<br>
b) get the mobile apps to a reasonable state (and release the iOS version, finally)<br>
<br>
This email is about a), I'll send a separate one about Subsurface-mobile.<br>
<br>
<br>
We have a ton of small fixes and improvements in master. And it seems<br>
silly to try to backport them all to the 4.5-branch. So I'm thinking it<br>
would make more sense to clean up master and do a 4.6 release instead.<br>
<br>
Linus and I have been using master on our last trip and I didn't really<br>
find a lot of things that are outright broken. One big thing that is<br>
currently not working as designed is the preferences. Tomaz promised me<br>
yesterday that he'd be taking a look at finishing the transformation<br>
there.<br>
<br>
But what else is missing to get master ready for a release?<br>
<br>
Clearly I need to document the two new features that I implemented and<br>
never talked much about (Linus may have given me an earful about that).<br>
<br>
(1) Subsurface master for a while now has supported multi user mode.<br>
<br>
Admittedly in a slightly limited fashion, but for me it has worked really<br>
well for testing.<br>
<br>
./subsurface --user=Joe<br>
./subsurface --user=Jane<br>
<br>
This uses two different sets of preferences. So units can be different,<br>
default file can be different (but careful, the default file is still<br>
based on the system username under which you run the app, so that's the<br>
same -- but come to think of it, that's something I really should change,<br>
too), cloud credentials can be different. It's not perfect, but I think<br>
it's a start.<br>
<br>
Comments, patches welcome.<br>
<br>
(2) Subsurface master now supports longer timeouts for cloud storage.<br>
<br>
When in Palau, 5 seconds was usually just a little too short for a<br>
successful connection to the cloud storage server.<br>
<br>
./subsurface --cloud-timeout=NN<br>
<br>
allows you to tell Subsurface how many seconds to wait for a response.<br>
Additionally, once it times out once, it no longer tries every single time<br>
but instead just switches into local cache only mode.<br>
<br>
<br>
With this out of the way, here's my TODO list for 4.6, and I hope that<br>
others will add to it with things that you are aware need fixing before we<br>
release 4.6:<br>
<br>
- Preferences handling: right now this is rather broken; some preferences<br>
aren't saved correctly at all. The "Apply" vs. "Save" in the preferences<br>
dialog doesn't work as expected, etc.<br>
<br>
- Offline mode: as explained above, when the cloud server times out we now<br>
switch Subsurface into offline mode. I think we should allow the user to<br>
try to take things online again, manually, without restarting Subsurface.<br>
I'm planning to work on this today after sending this email.<br>
<br>
- Mutiuser mode: change the name of the local default file if --user=NAME<br>
is given to NAME.ssrf<br>
<br>
- The odd error messages when saving to cloud storage that have been<br>
discussed on the mailing list in the last few days. That needs to be<br>
traced so we know what's going wrong. Since some people reported that it<br>
sometimes actually saves things correctly, my guess is that the error<br>
handling is broken in weird ways?<br>
<br>
- Image handling: Robert, I seem to remember that this was currently not<br>
in a sane state. Can you let me know where we are with this?<br>
<br>
- Dive planner: what's the status on that? Are we good?<br>
<br>
- Handling rebreathers: same question. Linus had some comments on that, I<br>
don't dive rebreathers so I simply don't know what the status is.<br>
<br>
- Qt 5.7: we need to get rid of the dependency on WebKit. Yes, I guess we<br>
could manually compile it and bundle it for our binary packages, but<br>
people who compile from source will run into all kinds of pain. I think<br>
this will also cause pain for Marble (but haven't looked into this at<br>
all, yet).<br>
<br>
<br>
That's the list that I'm aware of, feel free to add if there are things<br>
that I'm missing. Right now I have no time line in mind since it's very<br>
hard for me to guess how much time I'll have to work on this in the next<br>
month or two, but also because some of this depends on others (notably,<br>
Tomaz and Robert), who also had issues finding time for Subsurface,<br>
lately.<br>
<br>
Thanks<br>
<br>
/D<br>
_______________________________________________<br>
subsurface mailing list<br>
<a href="mailto:subsurface@subsurface-divelog.org">subsurface@subsurface-divelog.org</a><br>
<a href="http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface" rel="noreferrer" target="_blank">http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface</a><br>
</blockquote></div></div>