Beta 3 Testing

Dirk Hohndel dirk at hohndel.org
Sun Oct 4 20:02:38 PDT 2015


On Mon, Oct 05, 2015 at 12:21:56AM +0000, John Smith wrote:
> Running Win7 SP1
> 
> Installed ok, no issues seen.

Well, except for the warning that this package is from an unknown
developer. A warning that is getting pretty annoying with Windows 10.
Not quite as bad as on a Mac, but still.

I will have to break down and spend yet more money on a signing
certificate at some point :-(

> I had done an uninstall prior to installing but some of the logs were
> kept from the last installation and loaded at startup.

Yes, we intentionally keep settings and data files. Do you think that's a
mistake?

> Deleted the logs and used the reset all settings button in preferences. Most of the settings went back to default except - 
> 
> 1. Web Server Service ID was kept

The userid can also be saved in your data file. I don't remember off the
top of my head why we did this - I know there seemed to be a good reason
for it at the time. So if you are opening a data file that may re-populate
that user id.

> 2. Font went to a strange setting - not the same as the install font

Interesting. Which font did you get? On Windows 7 you should get Segoe UI
as font (which is the recommended font to use for those OSs). I just
realize that the code to do this is broken for Windows 10 which will go
back to Calibri. Arg.

Which font did you have before? Which font do you have now?

> CLicked New Log, added a dive and then clicked 'save as"
> 
> Default settings in preferences  shows "C:\Users\fibre\AppData\Roaming\Subsurface\" as location for logs whist 'save as' takes you to "C:\Program Files\Subsurface" as default.

Hmm, that's definitely bogus as we shouldn't save into the program install
directory, period. Looking at the code in file_save_as we seem to just not
populate the default filename there. Strange this never got noticed.

I will try to fix this.

> Imported 19 dives from Divelogs.de ok.
> 
> Loaded 13 dives from Suunto D4i via USB ok.
> 
> The dive number seems to be based on order loaded rather than date - I personnally would prefer the numbering to reflect the date.

Yes, that would make more sense. But I'm somewhat surprised by that. Our
code does first sort the dive table and then try to number the dives.
Look at process_dives() in divelist.c. 

My divecomputers seem to give the dives to me in the right order so I
can't quite reproduce this here. So of you start with an empty dive file
and load the 13 dives from your D4i they are not numbered in chronological
order? Or is this an issue because you first imported 19 dives from
Divelogs.de and got numbers from there? But even then, it should only
create numbers for the dives loaded from the Suunto if they are ALL newer
than the ones you got from Divelogs.de. So I guess I'm a bit confused here
as to what exactly is happening.

> Loaded CCR dive using IMPORT LOG FILE.  Default location again is "C:\Program Files\Subsurface". File loaded ok but 
> had zero as the dive number.

See above, we only number the dives if they are all newer than the
existing dives. Was that the case here?

As for the directory - that's the "lastUsedDir()" as found in
qt-ui/mainwindows.cpp - I wonder if this got messed up when you earlier
did a Save as and got the install directory there.
Because otherwise import should really give you either the last directory
that you used or simply your home directory. I'll need to play with this
on my Windows VM I guess.

> Saved all data to cloud ok.
> 
> Restarted but this time getting data from cloud. If you dont have a passowrd in the network settings, it doesnt give you an error message, it just stops.

It does for you? I get a red error message at the bottom of the Subsurface
window asking me to configure user name and password.

> Added location. The first line (in location) is white text on light blue background - its very difficult to read.

Very easy to read here on Linux. Tried it in the Windows 10 VM and it
seems easy to read there as well. Then I tried in a Windows 7 VM and you
are correct - that's almost impossible to read. Yikes.

We get the colors from the OS - Tomaz, any idea how to fix that? Can we
make the text font black?

> Tried changing cylinder data but couldnt delete the first entry in table - ended up doing a manual gas change at T=0  to allow system to calculate SAC. When adding cylinders, the O2 % was different each dive.

So you cannot delete a cylinder that's in use. Which in general is the
first cylinder in the table unless the dive computer explicitly tells us
that another one is in use.

The rest of the sentence above I don't quite understand. "manual gas
change at T=0 to allow system to calculate SAC" - is the dive computer
reporting incorrect cylinders and pressure data? Can you explain a bit
more what's going wrong there?

And that last sentence. "When adding cylinders, the O2% was different each
dive" -- when I add a cylinder here, the O2% always defaults to nothing
(which means 21% or air). Am I misunderstanding what you are trying to do?



THANK YOU SO MUCH for this detailed test and all the feedback that you've
given. I think I need a little more information on some of the items above
but I will certainly try to fix what I can reasonably fix before the
release.

/D



More information about the subsurface mailing list