Time consumption for identifying the changes for a mirror update

@user42 Hi, I edited the title for you. If some minor edits need doing, after the timeout. Just flag the post/thread in question and Forum Staff can look at it :wink:

Well… finally I know better!
I had the idea that it might not be the difference between Win and Linux that causes the effect,
but the sync application I used in Win.
It is “a bit older” (15y+), but at that time it was recommended by our national cyber security agency,
so it couldn’t be so bad.
However now I compared it to a new application (FreeFileSync, FFS)…
and… surprise!.. it compares (file size and time) abt. 9,300 files/s (460,000 files at the whole).

So, that’s a blow for me!
Actually it was the 2 hrs of my “old” program that inspired me to write my “Time Machine” at all!
If it took only 50s (as FFS) there wouldn’t be a necessity.
Another question (although OT):
Why the hell is FFS (and rsync) so quick? That old program won’t have coded pauses etc.
There must be a more clever way to compare folders… (?).
Sadly both a closed source.
Another observation:
FFS/ Win is even faster then rsync/ Linux.
But that’s not surprising for me so much, as Win (and esp. Win7) doesn’t do so “clean” and
safe as Linux do. I’m pretty sure there will be several flaws if having a closer look.

OK, all-in-all I just wanted to correct my former assumption concerning slow syncing at Win.
However I will drop it finally completely as soon as I can. :slight_smile: