What's your point? If you disable all remote and interactive stuff you get a fast page render? That's exactly what i criticize here. The page should be considered loaded when all is fetched and usable. And not "loaded" when the main parts from his server are transferred.
But yes, the Disqus takes the longest loading time starting at 4s. Scripts start at around 1 second to load. So yeah, you COULD argue the page is loaded after 1 second, but the point is, that it is not _RENDERED_ and functional and not all parts have been fetched yet.
Ofc your mileage may vary, i don't block disqus but use uBlock and the page is most likely hosted on another continent which makes latency another issue.
To make things fair i took to https://tools.pingdom.com/ and selected american and german servers respectively, which show 140ms for my page (Frankfurt) and still 800ms for his (from NA) which i think is mostly due to the use of http1.1. and ignores all remotely added scripts and tracking.
> Ofc your mileage may vary, i don't block disqus but use uBlock and the page is most likely hosted on another continent which makes latency another issue.
The page is indeed slower if I route it through a German VPN, but all content including the widget still stops visibly changing in under 5s on a hard reload, 1.5s on a soft reload: everything past that is various tracking nonsense which presumably has no impact on functionality.
My bigger point is, no one but a bot is going to be using the Disqus widget to check the comments within 5 seconds of visiting a page on a blog they've never seen before. All substantial content above it stops moving within a second, from Germany or from the U.S. (Unless your RTT is especially bad where you live?)
I just don't see what your prescription would be here, except to tear out the widget entirely, just for the sake of improving a metric with no direct importance. Unless you advise that the author implement his own commenting system, which is a huge can of worms filled with spam, legally-sensitive user data, and so on. I certainly wouldn't take it as an indictment on his choice of frontend or backend.
What's your point? If you disable all remote and interactive stuff you get a fast page render? That's exactly what i criticize here. The page should be considered loaded when all is fetched and usable. And not "loaded" when the main parts from his server are transferred.
But yes, the Disqus takes the longest loading time starting at 4s. Scripts start at around 1 second to load. So yeah, you COULD argue the page is loaded after 1 second, but the point is, that it is not _RENDERED_ and functional and not all parts have been fetched yet.
Ofc your mileage may vary, i don't block disqus but use uBlock and the page is most likely hosted on another continent which makes latency another issue.
To make things fair i took to https://tools.pingdom.com/ and selected american and german servers respectively, which show 140ms for my page (Frankfurt) and still 800ms for his (from NA) which i think is mostly due to the use of http1.1. and ignores all remotely added scripts and tracking.