areese801

Hi All - 

I know that using Dropbox to sync Alfred preferences between machines is the 'blessed' approach, but assuming that that isn't an option for me, my thought was that I could use git and 'manually' (git commit / git push / git pull) sync Alfred 4 between my machines.  High level it would look like this:

Set up a private remote repo on github

Via the 'Advanced' menu on each machine, change the Alfred sync folder to ~/alfred_4 (or something similar)

Add ~/alfred_4/Alfred.alfredpreferences to version control

Periodically, issue necessary commands to push/pull to/from remote repository (Or set up some automation...  Or get really 'meta' and set up an Alfred workflow that does this..).  Rinse / Repeat

So my questions are:

• Has anybody tried this?  If so, is there anything you'd like to tell me?
• Is this a terrible idea?  Besides not being officially 'blessed' methodology.  Will I get many merge conflicts or other thorns in my side?
• Most importantly:  Should I be concerned about security?  Specifically, how is sensitive data (e.g. Credentials use to log into some service / api that are set up for a new workflow, clipboard history, browser history, etc.) stored?  ..From what I can tell, these are probably kept in *alfdb files back under ~/Library/Application Support/Alfred/Databases, and thus not under source control.  Is this always going to be the case (say, with custom workflows from Packal?) Any insight / documentation about the underlying data model that can be shared?

Thanks in advance for any thoughts!