I found that even if there is no need to generate thumbnails, the original image will be downloaded, which is unnecessary and may be increases page loading time? Or something I set wrong?
In addition, is it possible to adjust the s3 api request after the page html display? Because when loading a page with many images such as Special:ListFiles, it is easy to load timeout (502 or 504 error). But this should be difficult, mediawiki do it during the parsing process.
I found that even if there is no need to generate thumbnails, the original image will be downloaded, which is unnecessary and may be increases page loading time? Or something I set wrong?
In addition, is it possible to adjust the s3 api request after the page html display? Because when loading a page with many images such as Special:ListFiles, it is easy to load timeout (502 or 504 error). But this should be difficult, mediawiki do it during the parsing process.