When ajax requests a new article, the article data is contained in a complete html file. For example: https://domain.ltd/i/?a=normal&rid=xxx&next=xxx&state=3&ajax=1

But this html will contain a lot of useless parts(html nodes) other than the actual data of the article. Especially when I subscribe to a lot of feeds, The <ul id="sidebar" class="tree"> tree can get particularly large. While Gzip can compress this part to a few tens of KiB, users frequently swipe the screen to load new article, and these useless tens of KiB are repeatedly transferred.

And to generate these useless parts adds a little bit of performance consumption, I guess.