This site only highlights all the bad things happening on those sites - and the marketing teams there have most likely already been told by their developers what including 1 million cookies & 50 million tracker APIs will do to the performance. They want this gunk in there so that revenue targets can be met.
So the best bet is to put this up on GitHub where folks like me could learn from the code :-D
Maybe open source the worker code, but the whole website?
I do not think it is 'not useful'. I was - not very clearly - posing a question on what your idea was to put it into 'practice'. I.e., whether you were hoping to turn this into a business or an open source project by posting here =D
Merely curious! Thank you for sharing!
I'm going to make a mild suggestion that the author double-check their implementation logic. Is it possible that the ad-encumbered page isn't loading before it times out? Or is it possible that without ads the page is loading more content?
Even showing something like a comparison screenshot at the end might help in odd cases like this.
My theory is, without the ads, there is more space in the webview to load video thumbnails, each item representing a video in the HTML document probably requires a handful of HTTP requests to load and pre-fetch their corresponding metadata. I would not be surprised if other websites react in the same way. I hope this is the case, but I am suspicious enough that I will investigate further how the ad-blocker affects these websites.
But consider this case:
https://webtest.app/?url=file:///etc/passwd
In short you should restrict URLs to protocols of `http`, `https`, and even then you should filter based on IP. You don't want people to view http://localhost/server-status, etc.
Finally you need to make sure you avoid recursion:
https://webtest.app/?url=https://webtest.app/?url=https://st...
I find it funny that the app reports even this file loads faster with uBlock.
Now don't get me wrong, I don't hate the greens party, but god damn, my eyes.
I don't think it lets you retest either, the first run is cached for a long time.
Also I never realized it was SO BAD without adblock
What striked me is JSHeapTotalSize, never really think the ADs are eating so much RAM / resources (make sense). In Guardian case this accounts for 50% more allocation.
Would be nice if there was a way to display/download all cached results. Would be a nice dataset for visualisation or a dashboard.
Without uBO, Forbes loads 151 cookies.
Did you consider computing the Speed Index? [1]
It would help assessing the performance impact these ads have.
[1] https://sites.google.com/a/webpagetest.org/docs/using-webpag...
