sfeather

So this is a ‘new’ behavior and I blame LSCache. New being, not sure when it came into being.

We noticed it on a testing server here. It did not appear in pre-production or in production because those environments have locked down file permissions, and LSCache couldn’t meddle with the Redis Object Cache file (wp-content/object-cache.php).

@saveatrain as mentioned by the author, I stick to Redis Object Cache and disable object cache in LSCache.

This, while appearing to be a minor inconvenience, adds yet another unneeded delay to the WordPress admin area. I don’t show the news feed in my dashboard, so why is it being requested at all?

This goes for the admin side of so many plugins.

I wrote an ugly bash script to take a file name (abcde.jpg) and create a/b/abcde.jpg in order to work around the 999 object limit. I just started to modify the script to handle more than a 2 level deep subdirectory system.

https://gist.github.com/stephenfeather/3aa4a9dadf0d8f516c0dabade13390fd

I also changed to putting new uploads into Y/m/d instead of just Y/m as we can potentially upload hundreds of product photos a day and could reach that 999 per month limit.

Thing is, in the source code it looks as though there was an attempt by the MediaCloud dev to address the object limits, but it doesn’t seem to ever process the second…. pages.