AMP lost all luster when I browsed to an AMP page on mobile and saw a popup at the bottom: “logged in as email@example.com.” Note that this was a random article, not anything that required me to login in any way.
I felt pretty naive to believe that G was doing this for the good of the web and not to amass more data on me.
Any non-AMP script can do that as well, so I’m not sure why that would change anyone’s opinion of AMP. Am I missing something?
AMP’s CDN is hosted by Google. Google is leveraging this (apparently via required JS) to amass yet more data. I don’t wish to support that.
There seem to be a lot of “Why AMP is Bad” posts lately, and I’ve been meaning to write a blog post from the other side. For now I’ll just say this:
I’m a software engineer at BBC News. We have a huge web performance problem, which kind of counter-intuitively is not a technical problem. Performance for us is an organisational problem, because we have dozens of other BBC product owners who want to shoehorn their content/widget/whatever into BBC News pages.
I have plenty of concerns about AMP, both technical and ethical. But when we joined the AMP trial, we immediately saw higher user engagement on our AMP pages. This ended up being a massive catalyst for a shift towards performance-focused culture across the whole of BBC News. People are curious why AMP has better engagement, and we’ve actually had conversations about how we can make the regular pages more like the AMP pages. It’s enabled me to speak openly without fear of annoying upper management about my opinion that BBC News is a content site, and we should build it to be content-first.
I know it’s easy to see AMP as Yet Another Evil Google Thing, but that completely ignores the fact that —like most big organisations— Google is quite fragmented. The people who run the AMP project genuinely care about web performance. Not a single one of them is thinking about Google’s ad revenue. All they care about is building a platform that gets content in front of users quickly, without the usual cruft that fills up most web sites.
I’ve only had AMPs come up in search results a few times, but it has literally never worked for me. I just get failure pages on Firefox Android + ublock o, so it’s become an indicator of something to avoid if I want to get to content.
It also doesn’t seem like something I’d work too hard to target as a content publisher, either… I’d rather just take responsibility for good mobile experiences.
same. AMP literally just doesn’t work for me at all. I have to go reload google in desktop mode for it to even be usable. i’m legitimately tempted to switch to Bing at this point to avoid it.
The problem with AMP is the walled garden nature of it and handing over control to a third party, but just like “My company website is just a Facebook profile” before it, it will take off and sky rocket because everyone else is doing it, and because prisoner’s dilemma.
Although that’s a huge potential issue in the future, the AMP community seems to be completely reasonable about submitting new components into the ecosystem thus far.
Is this meant to be ironic? All I see rendered is a blank white page with two URLs and a short orange line on it.
Author here. Could you provide your OS and browser? I’ll look into a fix.
Sure: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Firefox/45.0 on Debian Jessie with NoScript, but I whitelisted the site in NoScript.
Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Firefox/45.0
Should be fixed. Looks like it was a bug in the Fira Sans font itself.
The problem is the web-font used in the article. Just overwrite the font-family-tag of <article> and the text is readable. A shame you have to go through this to read this article…
A quick fix would be to run on the site:
Firefox lets you disable custom fonts for all sites via the UI settings. Chrome has an extension called ‘Disable Web Fonts’. Since I’ve disabled web fonts, I’ve found my web experience has generally improved.
That worked; thanks. Whether the irony was intentional or not, it greatly amused me.