My coworker Mike and I will be leading a discussion on #NearlyHeadlessDrupal at the Portland front end DUG meetup this Thursday. We'll be going in depth on the ups and downs of the process, our approach, and the module work I'm doing to make it all easier.
Event info here: https://groups.drupal.org/node/427533
For those that don't know who I am - I met several of you at DrupalCon Austin. I was the guy talking up doing headless Drupal now on Drupal 7 in the Headless Drupal BOF, the Core Conversations track with Carl, the Core Conversations with John Albin, and the small meeting we had in coder lounge. I'm currently working on a suite of modules to make this all easier to do on Drupal 7. I'll be posting more about them when they are more usable.
Comments
Recording?
Hey Jay -- any way this could be recorded or captured, say with Google Hangouts?
I second eatings' recording request
Jay, I talked to you briefly in Austin after John Albin's talk. Since then my team and I have been playing with Angularjs and Drupal with the goal of creating a headless site with Angularjs. We've dubbed our work so far as Spineless Drupal since we're still wrapping inside of Drupal's front-end. I would love to be able to see your talk but I'm in Boston.
Recording...
We're going to see what we can do to accommodate.
Glad to see interest in this. I use the term NearlyHeadlessDrupal because of SEO concerns - I plan to speak to that in detail at the talk. The basic idea revolves around setting up Drupal to handle Google/Bing's incoming hashbang / escaped_fragment logic. In a perfect world we would all use pushState, but I don't think it's a good fit for those trying to escape the Drupal theme layer. I'm using Drupal's CMS layer for data, and Drupal's theme layer for search engines.
Our DrupalGive page
Our DrupalGive page (http://liveaxle.com/drupalgive) now has links to the Audio / Slides / Code I've been working on. Any feedback / questions are welcome!
Pushstate and SEO
Just finished your slides — very cool, thanks for your work on them. I have a load of questions :)
Can you elaborate on this a little? Two things mainly:
I don't see much reason NOT to use pushstate — interested to hear from you if I'm missing something
how does Google crawling differ between /# and pushState?
Also, I gather the gist of what you're doing for SEO is serving Google a version of each page as straight HTML?
how exactly are you doing that?
is this really needed in the year 2014? Surely Google knows the web has had useful SEO data inside dynamically loaded content for about 10 years?
Finally, generating JSON data out of Drupal every time the front-end requests it begs the question of caching, for anonymous content at least. Have you thought of saving the JSON from NHD JSON to a file on disk only whenever the source entities change? The frontend would then pull the latest, fresh content (for anonymous) without even bootstrapping Drupal.