Christian Heilmann

First Decoded Chat of the year: Paul Bakaus on AMP

Today on the Decoded Blog I published the first ever Decoded Chat I recorded, where I grilled Paul Baukaus in detail about AMP.

This is an hour long Skype call and different to the newer ones – I was still finding the format :). There are quite a few changes that happened to AMP since then and soon there will be an AMP Summit to look forward to. All in all, I do hope though that this will give you some insight into what AMP is and what it could be if the focus were to go away from “Google only” with it.

These are the questions we covered:

  1. What is AMP to you?
  2. The main focus of AMP seems to be mobile, is that fair to say?
  3. Was AMP an answer to Facebooks’ and Apple’s news formats? Does it rely on Google technology and – if so – will it be open to other providers?
  4. It seems that the cache infrastructure of AMP is big and expensive. How can we ensure it will not just go away as an open system as many other open APIs vanished?
  5. Do large corporations have a problem finding contributors to open source projects? Are they too intimidating?
  6. Is there a historical issue of large corporations re-inventing open source solutions to “production quality code”? Is this changing?
  7. Whilst it is easy to get an AMP version of your site with plugins to CMS, some of the content relies on JavaScript. Will this change?
  8. AMP isn’t forgiving. One mistake in the markup and the page won’t show up. Isn’t that XHTML reinvented – which we agreed was a mistake.
  9. AMP seems to be RSS based on best practices in mobile performance. How do we prevent publishers to exclusively create AMP content instead of fixing their broken and slow main sites?
  10. It seems to me that AMP is a solution focused on CMS providers. Is that fair, and how do we reach those to allow people to create AMP without needing to code?
  11. A lot of “best practice” content shown at specialist events seems to be created for those. How can we tell others about this?
  12. AMP seems to be designed to be limiting. For example, images need a height and width, right?
  13. In terms of responsive design, does the AMP cache create differently sized versions of my images?
  14. Are most of the benefits of AMP limited to Chrome on Android or does it have benefits for other browsers, too?
  15. Do the polyfills needed for other browsers slow down AMP?
  16. How backwards compatible is AMP?
  17. One big worry about publishing in AMP is that people are afraid of being fully dependent on Google. Is that so?
  18. Are there any limitations to meta information in AMP pages? Can I add – for example – Twitter specific meta information?
  19. Do AMP compatible devices automatically load that version and – if not – can I force that?
  20. How can I invalidate the AMP cache? How can I quickly remove content that is wrong or dangerous?
  21. Right now you can’t use third party JavaScript in an AMP page. Are you considering white-listing commonly used libraries?
  22. It seems AMP is catered to documents, while most people talk about making everything an App. Is this separation really needed?
  23. What’s the sandbox of AMP and how is this now extended to the larger web as a standard proposal?