8 ms·
Each "page" in this multi-page article is literally 1 image, about 50 words of text, and 3 ads. The article content is fine, but there is no reason it has to be
by MicahWedemeyer 15y ago
Each "page" in this multi-page article is literally 1 image, about 50 words of text, and 3 ads. The article content is fine, but there is no reason it has to be spread across 7 different pages. Publishers chasing pageview and ad impression counts make the web a shitty place.
- tnuc 15y agoThere is an article you can read. You just need to sign up. And then they send you spam to make you think "I'm sure I ticked the no email box".
- dailo10 15y agoAnd if you don't want to register, you can use bugmenot to sign in.
- unicron 15y agoI viewed the comments just to write what you have. It's like reading CNet. Someone should write an app that turns shitty articles across lots of pages into one page!
- philwelch 15y agoSafari Reader has that feature, but I'm not sure how it works on slideshows as I use Chrome.
- fuzzix 15y agoYep, sorry about that, didn't realise the clickthrough article forced signup either. Delete?
- philwelch 15y agoIt's much better with Adblock Plus enabled, then you just get the big image and the text.
- alevans4 15y agoAgreed. I don't know why people run without Adblock.
- surement 15y agoBecause I don't use Firefox and the Chrome version says it will access any file on my computer if I install it, which I find unnecessary.
- jaredsohn 15y agoIt doesn't say that. It says it requires access to: " * Your data on all websites * Your tabs and browsing history " The reason for the first is that it needs to be able to modify every page you are looking at to block the ads. The reason for the second is that it needs access to the Tab API (to get the name/URL of a webpage you view, presumably to make use of white/black lists.)
- dredmorbius 15y agoWhich is a huge part of what makes various "privacy policy" terms so damned useless. There is a ton of software on my computer that "has access" to my "data" and "network traffic": the kernel, network stack, clients, filesystem drivers, etc. What they don't do is go tattling all this crap back to some mothership (or at least they'd better damned well not be). Where Android / iPhone apps and Chrome extensions start freaking the living crap out of me is where they say they "require" this access ... and I don't (short of setting up some network traffic instrumentation) know off-hand whether this is staying local to the app or getting broadcast to the world. Or some big brotherly subset of the world.
- jaredsohn 15y agoMy point is that what the extension requests access to is necessary for it to work, which you were disputing in your earlier message. Sure you have to trust it, and you are generally right to be paranoid about Chrome extensions, but on the positive side, the extensions are somewhat more limited in the bad things they can do compared to non-sandboxed native desktop apps, and the source code is always available (although at times obfuscated and can be updated automatically) since they are written in JavaScript.
- username3 15y agoSome people prefer pressing next instead of scrolling.
- monkeypizza 15y agoAutopager set up an infinite scroll for me, while I was reading the first one...