Would like an "Auto-Empty" feature Bewertet mit 5 von 5 Sternen

Will brief summarize suggested feature and why. Based on testing in a porting lab:

Automatic empty of ENTIRE cache at a predefined cache threshold. You already have reload on success, so it can integrate nicely. Nobody has this feature AFAIK, so you would have something unique. User choice or developer preset. 75mb, 100mb. Better would be 50% of total, 90% of total, etc.

1) In the old days disks had slower I/O but the head had smaller platters to manage. Today faster I/O, but larger platters. Its a wash. Firefox considers the default size of the cache based on free space (they may change that, but see #2). Past 100mb, disk thrashing begins. By 200mb, thrashing is bad. On larger platters, not good.

2) FF spends more time pruning cache to make room when cache hits max capacity (a read, a write/delete, then another write with new cache entry). How does the pruning work? Look at google/youtube entries in about:cache. Why I capitalized ENTIRE cache. If its expiration some of the cache may never get removed.

Example(s) Auto clean at 90% of max, there would never be a hit for pruning. Auto clean at 50% of max, less disk thrashing, no pruning hit.

Diese Bewertung wurde für eine vorherige Version des Add-ons (2.0.1-signed.1-signed) abgegeben. 

Good points

You have some good points in there.

Auto-cleaning has been suggested before so I plan to look into it. No promises that it will ever become a feature.

Works perfect Bewertet mit 5 von 5 Sternen

Does what its supposed to do. Would like to get in touch with developer, have some additional options ideas/requests based on FF cache size vs. system/disk performance testing.

Diese Bewertung wurde für eine vorherige Version des Add-ons (2.0.1-signed.1-signed) abgegeben. 

Just post the suggestions here

Feel free to use this comment section to submit your suggestions. I don't see a reason to get in touch for these. Every single thing about this add-on is open (source, statistics, ...) so I think suggestions should be, too.

Thanks for your input.