Maybe on Windows, but this is macOS. Wouldn’t it be easier to just use Time Machine to look at the completion log from the currently running copy job, and then know exactly how long it took?
No, I haven’t said that. It doesn’t list every file being copied, since that’d be too much data.
Give it time, Michael. It’s not stuck. The reason this stuff is collapsed out is so people don’t overly obsess about it…remember that Classic didn’t list the file at all!
This is just a bad place to do support. The board is restricting my replies, etc. Write to support. Thanks.
1 Like
No, that would not be useful…at all. Time Machine is not doing remotely normal copies, you have no idea what selections it’s making, how the kernel is scheduling its I/O, how it’s interfacing with the APFS file system driver (let alone the others), etc. It’s no more useful than doing a “disk benchmark” and somehow deciding that’s a baseline. It’s not.
It was a joke. Time Machine, looking at the future.
1 Like
It was in response to your message, but the content was for general knowledge.
I have absolutely no doubt that you, and anyone else who has written estimated-time code understands these issues and/or has been bitten by them in the past.
What makes you think that the elapsed time for Time Machine’s most recent run is going to have anything at all to do with an in-progress backup made with different software to a different destination device?
On my own computer, I have three external hard drives. Two different model drives (Toshiba N300 and N300 Pro) in three different model enclosures.
For the two that I use for CCC backups, I see that the N300 Pro in one enclosure takes less than half the time of the N300 (non Pro) in one of the other enclosures (typically 16 minutes vs. 35 minutes). This is over the same USB bus (I literally move the cable from one drive to the other when I make my backups), and with the two backups being run sequentially (start the second a few minutes after finishing the first), so the data backed up is almost the same.
d’Oh! 
Serves me right. 
Anyway, I just posted to the blog about this: Estimates — Shirt Pocket Watch
2 Likes
Wow! You really did some work on your fever bars! Nice! But, as you point out, there are some places where it’s just not worth the effort. For example, parsing a text stream where you may randomly encounter an infrequent object of indeterminant size and complexity where you just simply have to parse the whole thing before you know how long it would take. In other words, fuhgeddaboutit 
What I did in that case was to put up a spinner to let the user know it was still working while it was running through the mysterious object and then return to updating the fever bar when life came back to normal.
Dave
1 Like