Source? Or at least some specific details on how I could reproduce myself this?
In case the number of network requests was used to measure this (assuming Opera reports the number of network requests blocked), one must be careful with this specific measurement.[1]
I am not sure how to react. On the one hand, it seems like a good thing at the first sight. On the other, with ad-blocking becoming more and more mainstream, and the discussion looking more like clashes and agreements between companies, wouldn't advertisements find new ways of seeking attention, eventually evolving into something that we cannot escape?
What if they find a way to ban ad-blocking? What if web becomes completely binary, somehow making ad-blockers obsolete?[1][2]
That's highly speculative and phrased in a way impossible to answer:
> eventually evolving into something that we cannot escape?
What I think: if they find a way to block adblockers, we'll find a way to block the adblock blockers! It looks like a fight that will keep going on for some time. Until they make ads so good that they are indistinguishable from good content, then we'd won [1]
I now block the blockers naturally, just not following those pesky links. I would love some statistics for instance, from the times Forbes and others are in the top HN page before vs after they completely started blocking adblock.
I think the "something that we cannot escape" the OP is describing is already here in the form of "native ads", which are advertisements presented as articles. Buzzfeed does a lot of them, and there's a non-zero cognitive load in differentiating (since they look so much like articles).
I'm really not sure they're better than ads. They're certainly more deceiving.
I don't think it's possible. At the end of the day, the browser (user) determines if and how any given piece of markup will be rendered. There are server side hacks of course, but those are still relying on the user's agent to reply to them a certain way.
As long as the web is based mostly on HTML, the blockers win.
There's always the possibility of moving to another protocol, another ecosystem. Especially if it can be bidirectionally hyperlinked with the existing ecosystem. It might be something a bit more like USENET or Gopher, just adapted for the 21st century. With a huge Butlerian-Jihad-like opposition to active content - it would almost certainly be constrained to text, pictures, and video.
The other "ecosystem" is mobile apps, and it is how most younger people use the internet. The less revenue from a traditional website there is, the more incentive there is to make certain features app-only.
Young people use social apps, they don't read articles in mobile apps just like no one does that. The social apps are just a replacement for talking on the phone and text messaging, it's not like they are doing it instead of things they used to do on the web.
> Butlerian-Jihad-like opposition to active content
Genius line.
Individuals are already trying to improve things within their own spheres, witness the static site movement. Many people using static sites are trying to output some form of "reduced HTML" (without popups etc.) but it's very informal. It would be nice if there was actually a RestrictedHTML document format they could target, if you're trying to start a movement it's important to have words for what you're doing.
Note that while you'd be dropping huge amounts of features from the current HTML+CSS+JS stack, there are some features you'd want to add as well. For instance, the source of a webpage not being included with the page itself is not really optimal. (Think of how few pages are written in HTML directly -- they're written in Markdown on LaTeX or something else and then compiled to HTML). It would be nice to have native support for "ctrl+click on part of the HTML and then be taken to the original source".
> Accelerated Mobile Pages are just like any other HTML page, but with a limited set of allowed technical functionality that is defined and governed by the open source AMP spec.
How would they be able to ban adblocking? Browsers can be made and distributed anywhere in the world - even after a decade the pirate bay is still active.
I have looked at webassembly. Let them do whatever they want there, ultimately they still have to write the text in the clear either in the dom or in a canvas (and a canvas is really slow to write, especially on a mobile phone) and in either case as long as your code can run before theirs in the same javascript, you can redefine a few key built in functions (e.g document.createElement, document.getElementsByTagName) you can force their code to call your implementation of whatever functions they want to write the information with - and you can write their text elsewhere, or filter images or whatever you want. Webassembly makes their code more difficult to debug, but they still have to call the same javascript functions straight javascript does.
So adblockers won't become obsolete, but it is possible that it will no longer be enough just to filter the dom.
Hmm... I wonder, if they put everything (both navigation and ads) on the same canvas (maybe pixel-by-pixel even), is there any way to block ads inside?
If webgl will be adopted widely enough, this might even look plausible for some.
Pixel by pixel would be unbeleavably horribly slow in WebGL - OpenGL is all about loading graphics and moving verticies.
I think I am getting what you are saying though, but functionally that would be the same as just rendering it server side and serve it as an image (if you write the pixels randomly you still have to use the correct coordinates) which is possible today, yet nobody is talking about doing that.
So as long as they are still going to be rendering text client-side we can filter it. The moment they are not, we will have to start analysing the image, which should be fairly straight forward (ads are going to be in a rectangular area at least 20x20 pixels; text is going to a lot smaller and in ugly shapes).
This is an upheaval that is a long time coming. So many websites we browse rely 100% on ad revenue. And with adblocking slowly entering the mainstream, it's going to get ugly. Literally anything could happen. I'm excited to see what happens. This will be a massive culling of shit and junk on the Internet that rightly deserves to go under.
Brave is no major browser. Then again, if you look at the market share and depending on where you draw the line, Opera isn't a major browser either (1.68% according to netmarketshare.com).
Plus, seeing how small Opera's desktop user share is today, there have been various small browsers that have also had built in ad blockers, like Konqueror (whose code supplies the base of most user agents today).
iCab has had built-in ad blocking since the very beginning, in the late 90's. But by including the vague qualifier "major", one could technically argue that Opera was the first.
I think this is the case on older Windows versions but is negligible now on most systems (and apparently fine on Linux, etc., going by replies to a similar post I made a while back)
Faster, but less comprehensive. Many web sites serve content and ads off the same domain. Personally, I like uBlock Origin in combination with uMatrix in the browser, with privoxy and dnsmasq in use system-wide.
/etc/hosts doesn't work for a lot of ads, so no matter how fast it is, it's not a complete solution unless there's a browser based (or very intelligent proxy server) on top of it.
While this may be nice for the end user, the fact that Opera is in the process of being bought out by a Chinese consortium for $1.2B makes me think Opera will soon block more than ads, and will also track users.
I wonder if blocking ad-blocking people from viewing content is viable in the long term.
People producing content want to maximize the probability for their content to go viral don't they? Blocking anybody (ad-blocking or not) seems counter productive.
Well that's happening anyway with some sites if you're using ad-blocking. So the situation is no different. I imagine those users will just switch to a different source rather than give up the ad-blocking.
Biggest problem with with ads is tracking (privacy). Performance is an issue, but I'm not surprised that opera is making it sound as that's really the problem. Privacy concerns are mentioned briefly. Mozilla aporoach to ad blocking is truly user/consumer oriented, not just a lame attempt to still control and revenu from the whitelisting market.
That might be the case, but I don't think it is. In my experience, people dislike adverts on the web because they are obtrusive, obnoxious, and slow. Tracking is fairly low down the list of complaints.
> Biggest problem with with ads is tracking (privacy)
There are a lot of problems with ads. And if you ask 5 people what the biggest one is, your going to get 5 different answers. Rather than point out which one is the "biggest", we should point out which reason we are focusing on.
> opera.com is whitelisted by default, but you're free to remove it from the whitelist. There are no hard-coded exceptions to allow Opera or Opera Mediaworks ads through.
On a brand new install, there are a bunch of whitelisted sites (news and otherwise) I had to remove manually. I'm curious why they are allowing ads on places like NYTimes by default?
Cause people need to make money? Most people don't have a problem with small ads here and there, what people have a problem with is 75 ads on a page with 2 paragraphs of useful content and 17 download buttons that are ads and only one small link with the actual download.
Yes, but I should be the one to choose, not Opera. They already have ad blocking turned off by default why make people turn it on twice (switch on, then remove site from whitelist) and why decide FOR me which sites I'm supporting by whitelisting?
(Not that I actually do that, but hey some people are nice enough to whitelist so I'm making the argument for them. I block all ads and am fine with businesses going under because of it)
I think people are underestimating how big this is. When Brave did it you could excuse it as the action of a startup looking for attention; Opera now legitimizes ad-blocking as being a feature that should be on by default. This is similar to what went down with pop-up blockers, it quickly became a must-have feature.
I wouldn't be surprised if Safari was next and to ship their own content blocker in iOS and have it on by default.
The question is, how much will this trend cost Google?
Anti tracking and ad blocking are not the same thing. I don't want to see ads period, not just tell the websites to not track my preferences so they can gauge what ads to show.
Firefox's Tracking Protection is not the same as Do Not Track (which "tells websites to not track my preferences so they can gauge what ads to show"). DNT is simply an advisory header and it has no effect because very few websites honor it.
Firefox's built-in Tracking Protection uses Disconnect's blocklist to block content that tracks you, preventing the tracking from happening in the first place. As a side effect, it blocks some, but not all, online ads. The thought was that online advertising is still an integral part of the web's sustainability, and so blocking it entirely would be too much - but advertisements that nonconsensually track users are a privacy and security risk, and should be blocked.
Until we have a reasonable alternative to advertising for funding the web, I personally consider Firefox's Tracking Protection to be a reasonable middle ground that hopefully sends a strong signal to advertisers - online tracking is creepy, unwanted, and puts millions of people's privacy at risk.
Opera versions 12 and lower had built-in native ad-blocking. You could right-click on any element on a page and block it by exact URL or a partial URL. This feature was only just now reimplemented for the new Opera that's based on Chromium.
The question is how long will it take for sites to switch to fully run on header bidding and full server-side incorporation of ads into native placements in a largely unblockable manner.
If sites did server-side incorporation it would probably be nearly as fast as the site itself and therefore solve the problem Opera says they are trying to solve (web speed).
Hmmm...Isn't Opera in the process of being taken over by China's Kunlun and Qihoo? With their history, I doubt this will end up being in the users best interest.
Ad blocking is great but we are forgetting that most people will not pay to use a website. Advertisers are going to figure out a away around these things and we'll still end up with ads but at the same time we are going to see many ad supported websites disappear.
Yeah... many great sites are already just labours of love as it is...
I mean, I hate crappy clickbait sites as much as the next person... but in our effort to thwart those we're really going to put the squeeze on more honest sites as a side effect.
The advertisers have only themselves to blame for this mess. They've been abusing users for two decades now: pop-ups, pop-unders, flashing ads, auto-play videos, and now serving outright malware. They've utterly refused to take responsibility for their actions, and only blame users for blocking their ads, despite these ads making it nearly impossible to actually view websites (like when closing a pop-up window causes dozens more pop-up windows to appear) and trashing peoples' systems. Now they want us to feel sorry for them? No.
So we should just give up, sit back, and take whatever the ad networks see fit to cram down our throats? No thanks.
Also, I wouldn't put so much faith in the ability of advertising companies to completely stop adblocking. Similar things have been tried in the past and failed. (See the Betamax case[1] where home TV recording was challenged).
This seems like you're assuming the conclusion. When you say, "readers won't pay for sites [and sites cost money]", there are only a few solutions, and they all involve non-readers paying for the reader<->site interaction.
As a result, your only solution is hand-wringing about ad blockers, which probably won't result in "advertisers are going to figure out a away around these things", because browsers have complete control over what is displayed.
If you don't do this, things get more interesting:
- "People won't pay for sites" -> make sites that they will pay for.
- "non-readers paying for the reader<->site interaction" -> how can this be consensual, positive? (there are approximately seventy-three people, total, who enjoy web ads. Superbowl doesn't count.)
- "[ad supported websites are a good thing]" -> can they be supported publicly?
- "ad supported websites disappear" -> maybe in some cases that isn't bad.
- "ad supported websites disappear" -> "most people will not pay to use a website" stops being true.
Some small websites won't survive, unless they create free content. Newspaper sites will die, save one or two.
People don't want to have to subscribe to 10 different sites, so I suspect it will go the way of a single pay network that owns several pay websites. It will end up being Comcast of the internet, who controls which websites you can access for $99/month.
I didn't minds banner ads and the one or two side bar adds.
But then came the between content adds and the clickbait.
Also the tracking etc.
Sorry, but when adds look more like virusses and some are then I don't want to take the risk.
When a website however gives me a warning about my addblocker then I disable it to see how it looks like without addblocker on that website.
(I once had a site ask me this and then I got horrible adds with sound and autoplay and loops, that site went back to being blocked very fast)
Personally I will always be quite uncomfortable when the blocking of ads/tracking/data mining is provided by those with (not necessarily direct) links to the ads/tracking/data mining industry. My view on this is that this is the least worse solution for them to deal with the reality of more and more people using blockers, the worst being the solutions for which the ads/tracking/data mining industry has no say at all -- rather keep some control over none at all.
Opera 37 dev version is available for linux, so I installed it to perform my own measurements.
The example given in Opera's blog post is not very useful, they tested a lightweight page with only one network request blocked. That page loads fast even without a blocker. I preferred testing high-traffic, heavy front pages.
Chrome 50 + uBlock Origin 1.6.4 with default settings.
Opera 37 in ad-blocking mode, I did not touch the built-in whitelist.
For both browsers, click-to-play was enabled. I disabled the blocking of 3rd-party cookies for both browsers, this is necessary for the page load speed tool to make benchmarked web pages render properly sometimes.
For each site benchmarked, I opened a new tab for the page load speed tool.
I also tested a web page with nothing to block at all, to test for a case where the CPU cost of the blocking engine can't be offset by the gain of not loading blocked resources.
I did not interact with the pages once they loaded, to be sure I benchmarked exactly the same thing. Sometimes merely hovering the mouse over a page can cause javascript code to execute, etc. I didn't want this to happen. From what I could observe, pages looked similar in both browsers, except that there were Taboola ads on Business Insider with Opera, not present with Chrome+uBO (taboola.com is blocked by Peter Lowe's filter list, which is selected by default in uBO).
- Chrome 50: Average (valid): 258.51 ms (11/14 iterations)
- Opera 37: Average (valid): 281.82 ms (11/12 iterations)
Note that if you run your own benchmarks using above tool, you can't compare results with someone else's results, as your connection speed may vary with each computer. The results are comparable for the same computer+site with different browsers, or same browser but different blockers, or same browser same blocker but different blocker settings, etc.
For whatever the default exception list is quite large and clearing it required pressing X for each list entry and the list shrinking in height didn't help because one had to move the mouse cursor around after each delete. Otherwise it seemed to work well. Good on you, Opera!
What does give a website the right to include adware or malware and the like into the legitimate UseCase of viwing it? Even if it is just the Ad Network isn't it neglect on the websites part?
111 comments
[ 4.0 ms ] story [ 205 ms ] thread> We don't prioritize in any way our ads network. The most important is speed and user experience. We are not in the extortion business model as ABP.
Requires a touch more technical knowledge, but works across all browsers and applications.
I still get blocked by Forbes, but it's more funny (in a tragic sense) than anything else.
Source? Or at least some specific details on how I could reproduce myself this?
In case the number of network requests was used to measure this (assuming Opera reports the number of network requests blocked), one must be careful with this specific measurement.[1]
[1] https://github.com/gorhill/uBlock/wiki/About-%22This-other-e...
What if they find a way to ban ad-blocking? What if web becomes completely binary, somehow making ad-blockers obsolete?[1][2]
What do you think?
[1]: https://hacks.mozilla.org/2015/12/compiling-to-webassembly-i...
[2]: http://techcrunch.com/2015/06/17/google-microsoft-mozilla-an...
> eventually evolving into something that we cannot escape?
What I think: if they find a way to block adblockers, we'll find a way to block the adblock blockers! It looks like a fight that will keep going on for some time. Until they make ads so good that they are indistinguishable from good content, then we'd won [1]
I now block the blockers naturally, just not following those pesky links. I would love some statistics for instance, from the times Forbes and others are in the top HN page before vs after they completely started blocking adblock.
[1] https://xkcd.com/810/
I'm really not sure they're better than ads. They're certainly more deceiving.
As long as the web is based mostly on HTML, the blockers win.
Genius line.
Individuals are already trying to improve things within their own spheres, witness the static site movement. Many people using static sites are trying to output some form of "reduced HTML" (without popups etc.) but it's very informal. It would be nice if there was actually a RestrictedHTML document format they could target, if you're trying to start a movement it's important to have words for what you're doing.
Note that while you'd be dropping huge amounts of features from the current HTML+CSS+JS stack, there are some features you'd want to add as well. For instance, the source of a webpage not being included with the page itself is not really optimal. (Think of how few pages are written in HTML directly -- they're written in Markdown on LaTeX or something else and then compiled to HTML). It would be nice to have native support for "ctrl+click on part of the HTML and then be taken to the original source".
> Accelerated Mobile Pages are just like any other HTML page, but with a limited set of allowed technical functionality that is defined and governed by the open source AMP spec.
Sweeeet. Thanks for the tip!
Advertisers make more money on profiling and tracking users than showing you ads.
If you aren't paying for your content, your content needs to pay for itself.
I have looked at webassembly. Let them do whatever they want there, ultimately they still have to write the text in the clear either in the dom or in a canvas (and a canvas is really slow to write, especially on a mobile phone) and in either case as long as your code can run before theirs in the same javascript, you can redefine a few key built in functions (e.g document.createElement, document.getElementsByTagName) you can force their code to call your implementation of whatever functions they want to write the information with - and you can write their text elsewhere, or filter images or whatever you want. Webassembly makes their code more difficult to debug, but they still have to call the same javascript functions straight javascript does.
So adblockers won't become obsolete, but it is possible that it will no longer be enough just to filter the dom.
If webgl will be adopted widely enough, this might even look plausible for some.
I think I am getting what you are saying though, but functionally that would be the same as just rendering it server side and serve it as an image (if you write the pixels randomly you still have to use the correct coordinates) which is possible today, yet nobody is talking about doing that.
So as long as they are still going to be rendering text client-side we can filter it. The moment they are not, we will have to start analysing the image, which should be fairly straight forward (ads are going to be in a rectangular area at least 20x20 pixels; text is going to a lot smaller and in ugly shapes).
Firefox had it first. Enabling it's tracking protection makes almost all ads go away (because ads track you).
They're also forgetting Brave.
Brave is not a major browser yet. Although technically neither is Opera at this point.
[0] http://www.operasoftware.com/press/releases/desktop/widgets-...
> Opera’s ad-blocking feature is deactivated by default.
Not so smart, it seems.
People producing content want to maximize the probability for their content to go viral don't they? Blocking anybody (ad-blocking or not) seems counter productive.
There are a lot of problems with ads. And if you ask 5 people what the biggest one is, your going to get 5 different answers. Rather than point out which one is the "biggest", we should point out which reason we are focusing on.
[1] http://operamediaworks.com/
> opera.com is whitelisted by default, but you're free to remove it from the whitelist. There are no hard-coded exceptions to allow Opera or Opera Mediaworks ads through.
https://www.opera.com/blogs/desktop/2016/03/native-ad-blocki...
(Not that I actually do that, but hey some people are nice enough to whitelist so I'm making the argument for them. I block all ads and am fine with businesses going under because of it)
I wouldn't be surprised if Safari was next and to ship their own content blocker in iOS and have it on by default.
The question is, how much will this trend cost Google?
"Opera’s ad-blocking feature is deactivated by default."
Firefox's built-in Tracking Protection uses Disconnect's blocklist to block content that tracks you, preventing the tracking from happening in the first place. As a side effect, it blocks some, but not all, online ads. The thought was that online advertising is still an integral part of the web's sustainability, and so blocking it entirely would be too much - but advertisements that nonconsensually track users are a privacy and security risk, and should be blocked.
Until we have a reasonable alternative to advertising for funding the web, I personally consider Firefox's Tracking Protection to be a reasonable middle ground that hopefully sends a strong signal to advertisers - online tracking is creepy, unwanted, and puts millions of people's privacy at risk.
Another part of it involves mouse-gestures, which I use heavily.
I mean, I hate crappy clickbait sites as much as the next person... but in our effort to thwart those we're really going to put the squeeze on more honest sites as a side effect.
Also, I wouldn't put so much faith in the ability of advertising companies to completely stop adblocking. Similar things have been tried in the past and failed. (See the Betamax case[1] where home TV recording was challenged).
[1]https://en.wikipedia.org/wiki/Sony_Corp._of_America_v._Unive....
As a result, your only solution is hand-wringing about ad blockers, which probably won't result in "advertisers are going to figure out a away around these things", because browsers have complete control over what is displayed.
If you don't do this, things get more interesting:
- "People won't pay for sites" -> make sites that they will pay for.
- "non-readers paying for the reader<->site interaction" -> how can this be consensual, positive? (there are approximately seventy-three people, total, who enjoy web ads. Superbowl doesn't count.)
- "[ad supported websites are a good thing]" -> can they be supported publicly?
- "ad supported websites disappear" -> maybe in some cases that isn't bad.
- "ad supported websites disappear" -> "most people will not pay to use a website" stops being true.
People don't want to have to subscribe to 10 different sites, so I suspect it will go the way of a single pay network that owns several pay websites. It will end up being Comcast of the internet, who controls which websites you can access for $99/month.
https://news.ycombinator.com/item?id=11258292
The example given in Opera's blog post is not very useful, they tested a lightweight page with only one network request blocked. That page loads fast even without a blocker. I preferred testing high-traffic, heavy front pages.
The tool to benchmark page load speed is the one I wrote, found at: http://www.raymondhill.net/ublock/pageloadspeed.html (available on GitHub[1]).
Settings:
Chrome 50 + uBlock Origin 1.6.4 with default settings.
Opera 37 in ad-blocking mode, I did not touch the built-in whitelist.
For both browsers, click-to-play was enabled. I disabled the blocking of 3rd-party cookies for both browsers, this is necessary for the page load speed tool to make benchmarked web pages render properly sometimes.
For each site benchmarked, I opened a new tab for the page load speed tool.
I also tested a web page with nothing to block at all, to test for a case where the CPU cost of the blocking engine can't be offset by the gain of not loading blocked resources.
I did not interact with the pages once they loaded, to be sure I benchmarked exactly the same thing. Sometimes merely hovering the mouse over a page can cause javascript code to execute, etc. I didn't want this to happen. From what I could observe, pages looked similar in both browsers, except that there were Taboola ads on Business Insider with Opera, not present with Chrome+uBO (taboola.com is blocked by Peter Lowe's filter list, which is selected by default in uBO).
Results:
http://cnn.com/
- Chrome 50: Average (valid): 351.65 ms (11/17 iterations)
- Opera 37: Average (valid): 409.94 ms (11/15 iterations)
http://www.washingtonpost.com/
- Chrome 50: Average (valid): 933.18 ms (11/14 iterations)
- Opera 37: Average (valid): 1115.15 ms (11/14 iterations)
http://www.businessinsider.com/
- Chrome 50: Average (valid): 1188.54 ms (11/14 iterations)
- Opera 37: Average (valid): 1494.69 ms (11/17 iterations) (had clickbait Taboola ads)
https://en.wikipedia.org/wiki/Main_Page
- Chrome 50: Average (valid): 258.51 ms (11/14 iterations)
- Opera 37: Average (valid): 281.82 ms (11/12 iterations)
Note that if you run your own benchmarks using above tool, you can't compare results with someone else's results, as your connection speed may vary with each computer. The results are comparable for the same computer+site with different browsers, or same browser but different blockers, or same browser same blocker but different blocker settings, etc.
[1] https://github.com/gorhill/pageloadspeed
http://cnn.com/
- Opera 37 + uBO: Average (valid): 2280.89 ms (11/16 iterations)
- Opera 37 + built-in: Average (valid): 2113.47 ms (11/12 iterations)
http://www.washingtonpost.com/
- Opera 37 + uBO: Average (valid): 1135.04 ms (11/15 iterations)
- Opera 37 + built-in: Average (valid): 1072.35 ms (11/12 iterations)
http://www.businessinsider.com/
- Opera 37 + uBO: Average (valid): 1517.48 ms (11/15 iterations)
- Opera 37 + built-in: Average (valid): 1489.73 ms (11/17 iterations)
https://en.wikipedia.org/wiki/Main_Page
- Opera 37 + uBO: Average (valid): 304.43 ms (11/13 iterations)
- Opera 37 + built-in: Average (valid): 274.84 ms (11/16 iterations)