Originally Posted by
alex67500
- Marking posts as read as you scroll used to work but stopped a few weeks back (maybe 3 or 4). The workaround at the moment is, if there are multiple pages to read, to refresh the page when you get to the last post.
I have noticed this frequently as well in recent weeks, feels like maybe around the time of the last update. It *may* be more likely to occur if I navigate away from the page quickly (I am a fast reader
) -- maybe whatever code it is that updates the "last read" pointer hasn't had enough time to be triggered or something. It feels like if I let it sit for a while before closing the page then I have this problem less. But also it could just be that I'm refreshing the page and this is placebo effect.
BTW although I feel like the issue I reported here:
long page load times
has gotten *slightly* better, but only slightly -- for example on a thread of let's say 100 pages, it will randomly load page 30 (say), and you can briefly see those old posts (often years old) but then a second or two later it will jump to the actual first new post that might be on page 99 or 100, which IMO observationally supports my blackbox testing hunch here:
Originally Posted by
Zorak
It almost feels like the app server is loading *all* of the posts for the thread (and possibly transmitting them to the browser client) before displaying just the slice that's on the visible page (which needless to say, if that's what's actually happening, is needlessly inefficient).
I mention it here because there may be a common thread of, what all the heck is the page load code doing, and might it be over-bloated and time for some optimization...