CMS trends for 2014 | Hacker News Reader

CMS trends for 2014

despreneur.com

34 pointsaudriusarj39 comments

I'm slighly confused by the article's premise ("2014 will be the Content’s year") and it mentioning tools such as Wix, Squarespace, or even Macaw, who can't be considered as CMS but rather complete website builders.

Apart from its templating system and its vibrant community, one of WordPress' strengths is sticking to an admin interface that is actually focused on content: simple form elements and a decent WYSIWYG.

Drag & drop and in-place editing are technically impressive but don't belong in a CMS: I don't see editors and website owners using it on daily basis to update and manage their content.

I think the biggest hurdle we still face when it comes to online publishing is the clash between WYSIWYG and structured markup. Editors, in my experience, don't like working with structured content. But the web is not WYSIWYG. Tools like markdown attempt a compromise, but are still not suitable for many non-technical editors. Do we need a mind-shift, or a new piece of technology to solve this?

Yeah this is the main problem but I don't think it's really fixable in the near future.

There's 3 types of content:

- Arbitrary written content

This is where markdown shines and IMO completely solves content creation.

- Arbitrary written content with mixed in assets

Nothing works here. Ever try floating some text to the right of an image when you don't have access to CSS or the raw html because we have to assume the content creator has no technical knowledge.

Even wordpress fails horribly at this but it's such a common thing you want to do.

- Structured content that has clear properties

This is pretty easy to design a nice UI for because it's no longer "anything". For example a todo item or a photo that belongs to a gallery. All you have to do in this case is create a form field for each property. It's something a non-technical person could do.

  - Arbitrary written content with mixed in assets

  Nothing works here.

You're right. Offering a simple way for a non-technical user to manage (locally) his layout is the challenge of most WYSIWYG.

I don't use Medium but was quite impressed by this quick demo: http://www.youtube.com/watch?v=iO8hysXMfvQ

Check out around 2:00 how images are inserted within the text. The interface works because it offers only 3-4 possibilities: zoomed in/normal size, full width/left float. Also, there are only 2 levels of heading.

This limited interface is a great compromise between a content-focused experience, and some layout flexibility.

Yes, limiting the options is a great way to get good, consistent results. The problem crops up when that one user or client wants something different, then want to start making everything unique, then it's back to the same problem. Content management to me should be about editing the content with styling and layout done automatically, but in the real world I guess people want to be creative even when they are not designers.

The demo looks good but look how much fiddling he had to do to get the images the way he wanted them at certain points. There are so many little quirks that you can't really solve intuitively, you end up having to brute force a solution every time.

He even dealt with a number of non-image quirks and he probably coded the editor. Like when the paragraph and header were both bolded and he had to go back and fix it.

Medium is also highly targeted to blog posts too. A lot of non-technical users somehow want the ability to edit the content on the pages of their web site that might have fairly complicated markup/css to get the desired result.

You can solve the fiddling by replacing the whole UI with markdown. Oh, how will he push images to his liking then? Easy: {{image: cat2.jpg, left}} Ok. and howto make it fit the page and baseline without stretching out of the aspect ratio? {{image: cat2.jpg, fit}} Make a button for that like the Left,Centered,Right aligned text buttons and you're done. Even better remove the button and make the whole shit context sensitive, so that after dragging a picture you can do actions on it without having to mess with drag&drop. People don't get drag&drop, really, just trust me here. Most people aren't as aware as you about drag&drop and it doesn't always work too. Buttons help.

But yeah TinyMCE is DEAD.

Yeah I definitely agree with the drag/drop method not being popular for a lot of people. I'm not sure why, I guess it's because most people were brought up with "open the file dialog and pick the file".

If it were that simple there would be more good editors around. You should try to prototype what you're talking about and open source it, maybe it will gain traction if it's implemented well.

Thanks for the motivation, I'll have some more thought on that and try to implement it.

I always thought the C stood for compromise because that's almost always what you have to do when you use a CMS.

I don't think the future is drag/drop codeless systems. Good luck when your clients ask for very specific work flows that you can't bundle up in a pretty little draggable box!

Then get sad when you realize you have to dive deep into the code except now you're dealing with a 100,000+ line codebase that you didn't write or even worse a large plugin that some random dude wrote which has the worst coding practices possible.

The article doesn't mention API based CMSes, which are IMO a better approach to content management.

http://prismic.io seems promising.

Yeah I think the main omission was the whole trend towards markdown/git/api based "CMS" which basically means that most tech people have junked the whole idea of a traditional CMS, few of which have an API though there is CMIS.

Wrong. Personally for me a dead end criteria is a CMS without an API. But you mean most people don't need it right? I also think that's not right, because many companies want Mobile Apps, Intranet tools etc. and integrating their crap with the CMS through an API makes a whole lot sense.

I looked at http://prismic.io, but I haven't found anything that's opensource and comes close to it. Do you know something like that?

No not saying most people dont need it, they are a reducing pool, the people who do need it have moved away already.

>> […] people who do need it have moved away already.

hmm, what do you mean and to which platform?

I was thinking of enterprise users and companies who mostly start with a cheap cms, then add features here and there then an app, then an intranet thingy and at the end you have a spaghetti architecture, because everything has it's own front-end without a unifying API. It's rare that developers write their own solutions, you can say, hey nobody would do that, but the reality is far from that. Most devs simply use what's there and try to integrate different projects "by their own way".

I found some things that can be done better at prismic.io and think there is some wasted potential there.

- It requires writing your own controller that needs to implement caching or you have a 500ms delay for API requests

- No multilanguage from the beginning (gotta bake your own solution, that's not so fun)

- Slugs for objects in collections are ugly (see blog example)

Here are all REST API based "CMS", unfortunately, none of them is actually perfect. http://www.programmableweb.com/apitag/CMS?protocol=REST

I tend to agree. Any open source projects (to your knowledge) aligned with this approach?

WordPress might integrate a JSON rest API, you can follow it here: https://github.com/WP-API the developer is great but I wish they would throw more resources at it, it's not production ready.

Have you seen this? http://jetpack.me/support/json-api/

I'll look into it. Might be all I need for a few projects, thank you!

I'd personally include the upswing on static site generators. Not exactly a traditional CMS, but still a system that manages content.

In particular, I'd site NPR's work on their static site process:

http://blog.apps.npr.org/2013/02/14/app-template-redux.html

NPR stuff looks good, but their focus (from an initial skim) appears to be we apps - correct? That might be running before we can walk - how about getting all content-driven sites (blogs, portfolios, etc.) onto SSGs first - I think that's where the really big wins are.

Yes, I'd say that their sites fit into more of an "app" category. More reading, for the curious:

http://blog.apps.npr.org/2013/09/13/using-a-static-site-to-c...

Whenever I hear of these static content generators, I think of Fog Creek's City Desk [1]. Time for it to make a come back?

[1]http://www.fogcreek.com/citydesk/

A friend just asked me about doing a website for her bicycle repair business, and after thinking about it I thought that what I would like is a docker container that runs a lightweight httpd for serving static pages, a git server to allow her to push updates, and probably a simple web interface for deploying them. That way after the initial deployment my involvement should be pretty minimal. Maybe something like it already exists?

Edit: http://hyde.github.io/ looks like it might be a good starting point for something like the above, with bonus static site generation, markdown etc.

This article is horribly flat, once again bundling a blanket term "CMS" to apply to multiple needs and applications, and then using a great title like "Trends"....In essence this post is blogspam.

> Next year WordPress will still be a leader.

> 2014 will be yet another year of doing more while writing less code...Webydo, Froont or Macaw

Yeah right. As someone who'd consider themselves a WordPress expert and has been building clients sites with it (and other OSS offerings like Magento) for a long time there is no way in hell someone is going to use a WYSIWYG builder to create WordPress themes/sites.

If WordPress is the leader, maybe the argument that there is less code writing going on is a general one that means folks are using the CMS with plugins instead of rolling their own solutions -- where clicks replace code, in a sense. But you or someone else still needs to build the templates, customize the admin, register functions and nav and widgets, etc.

I suppose if you're really hell-bent on building a site with WordPress without coding you could like at a DMS add-on like PageLines but really, for most work with WordPress, you're writing code at some point.

"From writing code we will move to dragging and dropping, having everything visual and instant."

Sure, if he says so...

I had to do a double-take on Wix. (Even went to look at the numbers myself.) I had no idea that it had 1 Billion + in market cap. That's crazy impressive for a company in the crowded website builder space.

Edit: Back in August (most recent numbers I found), they had 679,536 paid subscriptions. That's insane.

I am not sure about the future of these "website builders". They have been around a long time, however for anything serious they almost/always disappoint.

His examples are exactly what's wrong with content management systems. They're too complex. I'm trying to build an insanely simple system at http://markdawn.com/ which emphasis content in a very strong way — content and only content.

I'm really working to make the interface as invisible as possible and mobile (iPhone) first. I hope I can get an API out in 2014, then it can be integrate with all Markdown editors and apps out there. It's cloud hosting for your Markdown files.

I am a little tiny bit dumber for having read this.

One static site generator that I recently discovered and plan to make frequent use of next year is Middleman (Ruby), which is like Jekyll, except less focused on blogging and more app-like conveniences...for example, being able to use erb rather than going through Liquid:

http://middlemanapp.com/

Overall, I'd like to see static-site generators become more widespread as an option. I've seen a few great sites/products recently in which the developer felt they needed to deploy a Wordpress instance even though their site was little more than a landing page and a few about pages. The less friction there is to getting content up, the easier it is to be motivated to create content.

I'm a big fan of Middleman too. Means I can use Haml, Sass, CoffeeScript etc. along with various Ruby view helpers to keep the templates tidy, then build the whole thing as part of the deployment process. Blog posts are in Markdown, which is even better, and the whole thing has live reloading. Very slick development process.

I wonder if it would be possible to build a CMS-alike on top of Middleman - something with a web interface to manage the content, edit templates and scripts etc., but using Middleman to generate a static site when changes are pushed. I guess it's another thing to add to the list…

This is not inherently related to Middleman, but I'm hugely grateful to it because through Middleman, I stumbled on to Slim, which is a nicer-looking Haml alternative that even allows Markdown. I've started using it in my Rails app and that alone was a game-changing improvement.

The way that Middleman can be so easily extended with plugins and gems the way that a Sinatra app can, while still maintaining the simplicity of a static app...It's really a breakthrough for me.

I think what's missing for static site generators to gain traction is a consumer-friendly admin interface. Kirby comes to mind here: http://getkirby.com/

Agreed...but what I think improvement has to come from both sides, not just in the software, but in the content creators.

Is it so hard for people to learn Markdown? Not just geeks, but anyone? How hard is it compared to learning the myriad of word processing tools that you're expected to know by high school? How much time does it take to learn Markdown compared to the minutes per day that you spend trying to debug the hidden markup of Rich Text/WYSIWYG editors (e.g., why is this text have the same font-weight as the headings now)?

Once the average person learns how to disassociate content from presentation and how markup is the key thing, then static site generators, and the real power of simplicity, become obvious.

I think to your average person markdown looks like a step down from WYSiWYG, especially when there's a massively successful, user friendly alternative called Wordpress.

I think that's a battle already lost.

> I think to your average person markdown looks like a step down from WYSiWYG

I agree. When I worked for a company that mostly sold custom websites to smallish companies, our sales guy would show his marks the TinyMCE interface and they loved it. I reckon we sold a lot of websites to people who believed we were giving them the ability to manage their own content easily.

Even better, we were able to generate more revenue when those people invariably messed up their content and needed us to fix it. So really it's win-win.

And all that lovely pasted MS word code ;) It's easy to end up with formatting nightmares. The roll-back button is a must.

I hate those editors with a passion, and though I like tools like Markdown (and Git), they are still problematic for people to use.

In many ways I don't see why the editor can't be right there as a first class citizen in the browser like it used to be. Linking and embedding needs to be simple.

Average users don't really want to get to grips with html or some other markup language. But it would be great if at least they could grok linking.

Reply on news.ycombinator.com