new statistics code merged

Sun Jan 3 15:04:09 PST 2021

Hi,

Short testing on Ubuntu 20.10 with the AppImage:

$ ./Subsurface-4.9.10-232-g77bc0c7c9396-x86_64.AppImage
qrc:/qml/statsview.qml:3:1: module "QtCharts" is not installed
QObject::connect(QtCharts::QChart, StatsView): invalid null parameter
zsh: segmentation fault (core dumped) 
./Subsurface-4.9.10-232-g77bc0c7c9396-x86_64.AppImage

Do I need some additional dependencies to get the AppImage to run? I
thought the idea was that there are no need to install dependencies when
using an AppImage built app. But I could be wrong on my assumptions.

I also tried the 4.9.10.232 from launchpad, and with this download it
works. I found the statistics (after some searching) and clicked around
a little bit. I really like the feature, but I guess I need some more
time and motivation (read possibility to go diving) to really dig into it.

One small issue I found: When "Data" is set to "none", I can't select
the binning for the "Base variable". Maybe that's the same bug that Dirk
mentions in the mail below? The same problem is also present when "Data"
is set to "Date" or "Weight system" (and possibly others), only the
default binning for "Base variable" is selectable (for example 2l for
SAC or yearly for Date).

Thanks for all the work, Willem, Berthold and Dirk!
Christof

On 03.01.21 23:27, Dirk Hohndel via subsurface wrote:
> Hi Everyone
>
> For the last year or so, Willem and Berthold have been kicking around
> a complete rewrite of the statistics code in Subsurface.
> And for the last three months or so, Berthold has really dug in and
> created a rather stunning piece of work.
> Over the past two weeks we cleaned things up and merged it, and I just
> finished of the (for now) last part of that, bringing in the UI.
>
> This will become Subsurface 5.0 in a few weeks, but until then a LOT
> of testing will be needed, and frankly, a lot of feedback.
> It's really hard to create something this powerful in a vacuum, but we
> also know from experience that it's very hard to get people to comment
> on ideas and drafts when they can't actually play with the software.
> So what we have now is what Berthold, Willem, and I managed to pull
> together based on ideas that were discussed here and on GitHub for
> quite a while. But it's almost certainly not the final version of what
> we'll ship as Subsurface 5.0
>
> What we need from all of you is testing and honest feedback. What do
> you like? What do you wonder about? What do you hate?
> We want to make this better, easier to use, more useful in itself.
>
> Once we have consensus (or at least once we made the decision that we
> are done), all this needs to be added to the user manual, needs to be
> translated, needs to be gotten ready for a release. And THEN we can
> talk about a release time for Subsurface 5.0.
>
> But for now the most important thing is to get this into people's hands.
>
> I tried to test the binaries on the main platforms and they work for
> me - but of course the challenge there (with the exception of maybe
> Windows) is that I have a ton of things installed on my Linux boxes
> and Macs that may not be installed on yours - so there's always the
> risk that some of these don't work out of the box for you. If that
> happens, please let me know so I can track this down. So even if you
> usually compile your own binaries, it would help me if you tried the
> test builds on Launchpad (Ubuntu) and OBS (openSUSE and Fedora). And
> of course the Windows and macOS builds.
>
> Look for versions that are at least 4.9.10.232 - as I write this
> macOS, Windows, and the AppImage are on
> http://subsurface-divelog.org/downloads/test
> <http://subsurface-divelog.org/downloads/test> and the various builds
> on OBS and Launchpad are mostly completed or at least well on their
> way and should be done, soon.
>
> A massive, MASSIVE thank you to Berthold for the countless hours that
> went into this. If you look at the past 90 commits or so, you can see
> just how much work this was.
> Also, a huge debt of gratitude to Willem who came up with much of the
> underlying design of what was implemented - the design of the charts,
> the UI logic, etc.
>
> So what, you wonder, does this get you? Things like this:
>
>
>
>
>
> You can truly go wild :-)
>
> Just creating these screen shots I already found the first two
> issue... it appears that the temperature binning only works for 2º
> increments, I can't switch to larger bins in the UI.
> And the trend line in the scatter plots is sometimes drawn behind the
> graph :-)
>
> I'm sure we'll find many more issues. That's why I'm asking all of you
> to please help us with testing.
>
> Happy New Year!
>
> /D
>
> _______________________________________________
> subsurface mailing list
> subsurface at subsurface-divelog.org
> http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210104/d09f5b9a/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen Shot 2021-01-03 at 2.21.55 PM.png
Type: image/png
Size: 160109 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210104/d09f5b9a/attachment-0004.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen Shot 2021-01-03 at 2.20.35 PM.png
Type: image/png
Size: 242028 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210104/d09f5b9a/attachment-0005.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen Shot 2021-01-03 at 2.19.59 PM.png
Type: image/png
Size: 259828 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210104/d09f5b9a/attachment-0006.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen Shot 2021-01-03 at 2.17.52 PM.png
Type: image/png
Size: 222427 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210104/d09f5b9a/attachment-0007.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210104/d09f5b9a/attachment-0001.sig>