Improve on the "prefer downloaded" dive computer model

It used to be that when you checked the "Prefer downloaded" checkmark,
we'd throw away *any* old dive computer data.  That was good, because it
allowed us to start from a clean slate when you had some old subsurface
data with questionable dive computer data.

However, it was a bit extreme, and it's really not what you want if you
already have (good) dive computer data from other dive computers.

So this modifies the logic a bit.  Instead of throwing away all old dive
computer data, the "Prefer downloaded" checkmark now means:

 - the newly downloaded data becomes the "primary" dive computer data
   (ie the first in the list)

 - if there was old dive computer data that *could* have been from this
   new dive computer (ie it didn't have model information, or it had a
   matching model but no device ID data), we throw that away

 - but any existing dive computer data from other dive computers is left.

This seems to be much closer to what we really would want for a new
"preferred" download.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Dirk Hohndel <dirk@hohndel.org>

0 files changed, 0 insertions, 0 deletions