Closed hollowaykeanho closed 1 year ago
May have to revisit full directory recursive scan due to #30
Scan all directories method not possible and it will revert back to Issue #26. Basically, we're stuck.
Managed to develop a way to scan all pages only at root page while maintaining performance (commit: https://github.com/ZORALab/Hestia/commit/7a6003c5b09a8e6e06aa557c261a357590881979).
metric-stage-5.ods metric-stage-5.txt metric-stage-4.ods metric-stage-4.txt
Can leverage the scan loop to process all caching data.
Fixed in 265f8651a9d39e2195e3445d70c2e38040443117. Updating all existing documents...
All documentations updated.
Description
Originally, the PWA cache system was to walk through every directories'
__caches.toml
data file where it lists all the caching URLs for that particular contents. Due to this recursive scan, it caused extreme performance issue tracked under Issue #26 . Since, then it was removed in commit https://github.com/ZORALab/Hestia/commit/a90901d9b7535c1c9caeb0f28fa8fe909383b5cd and the setting was reverted to site levelHestia.PWA.Cache
table. With this compromise, Hugo is back to reusable mode again.While doing so gain in performance, consolidating the entire repository's caching URLs under a single file can be a nightmare for maintenance as that one particular file is kept changing. It's a guarenteed way to look for trouble.
Expected Behavior
A directory-oriented way to provide PWA cache input instead of risking consolidations to 1 repository.
Current Behavior
Using consolidated 1 toml file for the entire repository.
Attachment