Windows binary saga

[Sander M. Kleijwegt]

Whoops, it probably adds to the confusion that I named the files in the wrong order. Let me explain in detail.

WinMachine is a windows machine where I created the logs on (using subsurface -v -v). UbuntuMachine is a machine running ubuntu which I use to a) add dives to the cloud using subsurface, b) check the cloud git repo using the git cli tool and c) use chrome to check the cloud.subsurface-divelog.org web gui.

Scenario 1, which gave me the log called load_with_cache.txt, which is actually a clean load.
WinMachine: 
- remove C:\Users\<user>\AppData\Local\Subsurface\Subsurface
- run subsurface -v -v
See that all the dives that are in the cloud are shown

Scenario 2, which gave me the log called initial_load.txt, which is actually a load with a cache in C:\Users\<user>\AppData\Local\Subsurface\Subsurface
Given the WinMachine with subsurface still open from scenario 1.
UbuntuMachine:
- add a dive with some text in the notes field so I can recognize it easily
- press file -> save to cloud
- use CLI to git pull my cloned cloud-git-repo, verified a new commit(=dive) is coming in
- use chrome to check the webgui, verified a new dive is shown

WinMachine
- press file -> open cloud storage
- see that although the log shows things like fetch remote and sync with remote, the dive I added using UbuntuMachine is not shown.

After these 2 scenarios I verified that when I removed the cache from WinMachine and restarted subsurface, the cloud was loaded and the new dive was shown as expected.

I hope I am more clear now, it was early for me too this morning.

regards,
Sander
________________________________________
From: Dirk Hohndel [dirk at hohndel.org]
Sent: Friday, September 25, 2015 15:27
To: Sander M. Kleijwegt
Cc: Subsurface Mailing List
Subject: Re: Windows binary saga

On Fri, Sep 25, 2015 at 07:31:07AM +0000, Sander M. Kleijwegt wrote:
> Please find attached 2 log files
>
> initial_load.txt -> Opening subsurface without the cache directory present. It connects to the git server and downloads/shows all dives correctly

But the first one appears to do a git save, not a load; I'm confused.

> load_with_cache.txt -> After adding a dive from my ubuntu machine I pressed file->open cloud storage on my windows machine and didn't get the new dive. I was able to git pull on the cloud-git manually from my ubuntu machine to verify the new dive was there.

When you add a file on your Ubuntu machine and saved it to the cloud, what
did you do on the Windows side to see it? Restart subsurface and open the
cloud storage? The second log shows a git load as expected and loads 87
dives from it. It seems to access the remote correctly (it does a git
fetch).

I still don't quite understand the sequence of operations, to be honest.
Must be the lack of coffee.

This is the sequence that I think you are doing. Can you verify?

Machine 1: remove cache (so this simulates a new machines that connects)
           open cloud storage
Machine 2: open cloud storage, add dive, save to cloud storage
Verify via web interface that the new dive has arrived on the cloud server
Machine 1: restart Subsurface and open cloud storage
           new dive is not showing up

Is that the sequence?

/D