Responsiveness in folders that have a lot of file (70k+)

[o7blue]

When navigating into a folder that has a lot of files (my current large folders range between 70k to 800k files) it can take a good while before xyplorer becomes responsive. It also becomes responsive when I press Esc (ending some processing that's happening in the background I imagine).

My question: is there something I can do to improve responsiveness in folders with a lot of files? Or have xyplorer give up on fully looking at the files if there's more than x amount of them, and just do the bare minimum?

[highend]

And that does happen in a fresh instance?

[o7blue]

huh, the fresh instance is way more responsive (800k folder become responsive in 2-3 seconds. I gave up measuring my normal instance at over a minute). It's probably something I set in the configuration that's making it behave this way. Guess I need to comb through it a see what I did. I mean, it's unlikely that it's behaving like this just because I've been updating without a fresh/new install for the past 3 years, no?

[highend]

Check your view mode & the columns shown and custom event actions...

[o7blue]

I think I found the culprit, I had Special Property: Aspect Ratio enabled when I looked in Show Columns. I guess that's not something that's cached and needs to be re-calculated every time?

Thanks for pointing out the likely locations to look in, I'm glad I didn't have to scrap my configuration and redo everything from scratch. And of course the entire issue was self-inflicted haha.

[highend]

If you want such things “cached”, tag the files with that info…

[o7blue]

That's definitely my plan haha. I keep metadata on each image file in a json file, and it looks something like this:

Code: Select all

{
    "id": 380,
    "created_at": "Mon Jul 16 00:25:09 -0500 2007",
    "score": 149,
    "width": 640,
    "height": 480
 }

Is there a straightforward way to read a json in an xy script, or do I need to use regex or something like that?

[highend]

Regex or create a report or loop over them… Many ways lead to Rome

[jupe]

You could also use this via runret if you are so inclined:

https://jqlang.github.io/jq/

If there is no exclusive info in the json that you can't just query (with xyscripting) directly from the image file though, it is probably easier to just query the file directly.