Stress testing how Sync works

I am testing syncing so I can understand what it does and doesn’t do.
So far:

  1. first sync from devonthink to the ipad version was joyously successful

  2. new documents created either on the mac or ipad sync very easily and swiftly. This is so good (and so much easier than itunes) that it actually transforms putting data onto and getting it off the ipad.

now for the more stressful tests

  1. if having synced a file you edit it on both mac and ipad and then sync (ie 2 docs same name but now different) how does it resolve the differences? In this case it seems to duplicate the docs. My mac ended up with 2 different docs with the same name; my ipad had 2 docs with same name but they were both the same.

  2. If you edit a doc created on the ipad on a mac it does not seem to update the ipad copy. However edits on the ipad original seem to update fine on the mac.

There seems to be a master/slave concept behind syncing which makes sense. But it needs DOCUMENTING as it will have impacts on workflow.

That’s as far as I have got so far. What have other people found? Years syncing files across ftp has made me cautious about sync and I need to know how it works…

Good points, @buybeckett. Whenever syncing machines is involved I find it safest to follow this general workflow pattern:

[work on machine 1] [work on machine 2] [work on machine 1] …

Relying on some developer somewhere to get bi-directional asyncrhonous syncing right is too fraught.

I’m speculating that you are encountering something that I saw when beta testing. When stress testing the sync function, it will be helpful to let some time elapse (perhaps 2-3 minutes) between edits and also between syncing on each device. The sync routine looks at modification date and time to determine which documents need synced, and the Mac and the iOS device will rarely, if ever, have perfect time sync.