On lundi 23 novembre 2020 12:16:35 CET GianLuca Sarto wrote:
> Hello All,
> 
> it's not really an issue, just wondering...
> 
> I load my RAWs in one folder which has now about 10K images.
> 
> Each time I add a bunch of 30-50 images, it reads all folder just to add
> the new ones.
> 
> Does it has to be so?
> 
> I mean, rsync has the ability to filter files that are already on
> destination, could the same feature be implemented in DT?
Rsync still has to check some data for each file as well. Which means the 
whole directory has to be read in (not each individual file, in first 
instance). And rsync "just" compares raw directory info.

As you add the images to an existing folder, dt has to check each file against 
its *database* to see if the file is present, and if there have been changes 
since last startup. Database queries have more overhead than simple directory 
comparisons, so will be slower. 

How long does dt take to add those 30-50 files?

Another question is whether storing all the raws in one flat directory is the 
fastest in normal use. Unless you always use a filter, scrolling through the 
files is going to be slow. Using sub-directories probably speeds up using dt. 
(and once a directory is imported, the data is retained in the database, so 
e.g. searches will see all the imported images) 

Remco


____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]

Reply via email to