[inthewoods]

Yeah I've been thinking about this kind of an approach for a while - the idea of making a content management system that is just a content store that is accessed via API. Like it.

The largest obstacle for me to adopt such a system is that it requires me to bet my company's content on your system being around. I've looked at, for example, Webpop which is different but has some similar ideas - but I can't get away from the fact that Webpop may or may not be around in a year, so betting on it isn't something I would want to do.

That's what continues to push me to Wordpress and the like, although recently I've been spending more time with static web generators like Harp and DocPad. Those have a similar issue - my boss saying to me "Well, I know we can find Wordpress developers, but will wevbe able to find a Harp developer?"

[einrealist]

It exists: Apache Jackrabbit + Apache Sling

I work with Adobe Experience Manager (former Day CQ) and the way the content is managed is the best, I have experience with. Of course, it does not mean, the product is perfect. But the concept feels right.

[ccleve]

Yes, it looks like the same idea as Jackrabbit / JCR. I hope they've done their homework and understood the advantages and limitations of that API. I always found it a pain to work with, and that's probably the reason it never took off.

I really, really hope they succeed, but they've bitten off a difficult problem. It's not hard to design a simple content management API. What's hard is making it general enough to handle many situations without getting too complex.

[einrealist]

Yes, the JCR spec is not easy to implement. But with a mature implementation, you can create your CMS convention driven. Those conventions can be as narrow as you need for a simple CMS. And when you want offer more, you are not limited by the content repository but only by your conventions, which you can widen at any time. I hope, more people will look to JCR as an option, because it solves the content repository problem very well.

[bobfunk]

Thanks for considering Webpop! I'm the founder.

We've been around for more than 3 years now, have clients with hundreds of published sites and we're profitable, so we're not going anywhere.

That said, the lock-in is a problem and I wish there was a really great way to address the lock-in. Open-sourcing doesn't really solve it when the software is built to scale up to handle thousands of multi-tenant sites on a distributed infrastructure, but not to be simple enough to install and run that it's viable for a relatively small number websites.

[phpnode]

I'm working on a product in a similar space to prismic, and as a developer myself, this would also be a concern for me. The way we plan to approach it is offering a self hosted version of the product, and committing to an open source warranty so that if something went wrong with the company the product could live on. Would these kind of measures make you feel more confident?

[inthewoods]

There are two concerns - one is that the company goes away. The second is that there aren't any developers that know your product, therefore we can't hire anyone without training. So your measures would deal with the first but not the second. Have you considered open sourcing some part of it?

[rudyrigot]

Hey, indeed, part of the easy learning curve is that all of the kits (i.e. the part that developers actually use in their projects) are open-source on GitHub: https://github.com/prismicio Also, learning how to develop with it could really hardly be easier, check out the 10-minute video on the developer's portal: https://developers.prismic.io/

[jordanlev]

Agreed. I think the way openphoto/trovebox does it with photos might be a solution to this problem... an open specification with the choice to store it on your own server or your own AWS acccount... then the company provides pay services for hosting themselves or for providing the pre-built back-end and front-end tools.

[oneeyedpigeon]

Experience with a specific product is way overrated because a) products don't last forever, programming does b) any decent programmer can learn how to work with harp in about the same time it would take a Wordpress expert to get up to speed with your particular Wordpress setup

[inthewoods]

Somewhat agree/somewhat disagree. Products don't last forever, but the large open source products like Drupal and Wordpress have been around a fairly long time - long enough to instill confidence in my boss that they will, more than likely, continue to exist in the future. I do agree with your second point. But my point was not necessarily about my personal opinion - my point was that organizations are often concerned about this (e.g. my boss) - and that makes it difficult to commit to a smaller, closed-source or hosted vendor.

[haraball]

I share the same concerns as you. This looks like a great product, but it needs a big amount of trust and commitment to get any value from it. Usually these kind of services launches with a boom, start giving huge discounts shortly after and then shuts down silently.

[sadache]

Is your problem related to availability or continuity of the service? Things that could help their, a usable export for the first case. And an insight into the API architecture for achieving scalability and availability for the second.

[sadache]

if you're interested, here is a blog post talking about the API availability https://blog.prismic.io/UlMRYknM0-kFs5Cy/how-prismicio-api-s...