The WP REST API team and WordPress core contributors met tonight to decide whether to merge the content endpoints in 4.7. After the merge proposal was published a week ago, several core developers expressed concern regarding the brokered authentication scheme and the team has since decided to remove it from the proposal in favor of focusing on it for the 4.8 development cycle.
Discussion at the meeting tonight centered around six topics: security, performance, user feedback, if merging will negatively impact API development, whether content endpoints will benefit core development, and whether those endpoints belong on every WordPress site. The team also discussed possible ways to measure the success of the project once it has been merged into core.
Contributors agreed on a conditional approval for merging the endpoints, provided that the team address outstanding questions on object meta and that others outside of the REST API team provide a proof of concept for how WordPress core can use the endpoints. These conditions must be met before the enhancement deadline next Wednesday.
“Making something work in core does not mean committing it to trunk,” WordPress 4.7 release lead Helen Hou-Sandí said. “This is also a call to the greater dev community – there is a chance to do something that doesn’t necessarily carry the weight of being shipped in core that can serve as proof that you want something. A developer’s vote.”
Hou-Sandí said that requiring the proofs to be created by developers outside of the REST API team will demonstrate “how other people experience the development process.” It also frees up the project’s team to focus on other pre-merge tasks.
Multiple proofs of concept are encouraged and some of the features being considered include Press This, Quick Draft, infinite scroll on admin list tables, and anything else anyone wants to try. Adam Silverstein volunteered to take a crack at Press This and by the conclusion of the meeting said, “I’ve got Press This creating new posts already actually, that was pretty straightforward to switch over.” He plans to include his work in a new ticket on trac.
“I have no blocking objection but I am pretty wary and want to make sure that conditions for keeping in 4.7 are hammered out in the next 24 hours and conditions for beyond worked on before beta,” Hou-Sandí said. Matt Mullenweg, who has been one of the most vocal critics of the project’s readiness so far, agreed with her statement but also said he isn’t yet satisfied with how the team plans to measure the success of the project.
“I also feel like the measure of success is still woefully undefined, and there’s still a lot of fuzziness in the core arguments of ‘if this is in [core] people will use it more,’” Mullenweg said.
Contributors seem very motivated in the final stretch and are working towards producing the necessary proofs of concept prior to the enhancements deadline. If any of the proofs are solid enough to be merged, the WP REST API content endpoints will ship in WordPress 4.7 alongside a core feature that is using the API.
It was a delightful meeting to attend and watch wonderful arguments/conversations. Glad the REST-API finally getting in WordPress core for a many of us to enjoy and work with the awesomeness in the future.
A zillion thanks to the team and everyone who worked on to get it this far.