Comment by homosaur | Hacker News Reader

Comment by homosaur | Hacker News Reader

I'm so weary of people pitching the solution to WordPress and then tossing a proprietary cloud hosted thing at me. That's a complete non-starter.

All the time I hear from fellow devs "oh don't advertise you use WordPress, it's irrelevant to clients." Tell that to my bank account and all the jobs I've stolen from agencies who wanted to use their BS proprietary CMS.

I sell WordPress like this: when you want a rebuild in 10 years and I'm in Thailand smoking drugs with your money pile you just gave me, you'll be able to find someone to work on it no problem.

There's plenty to dislike about the WP codebase, but we all have to remember that WP was made in an era when coding reasonable software in PHP was nearly impossible. Nothing backwards-compatible can happen there that isn't going to destroy all the millions of existing sites. It's still pretty freaking easy to build a site with, especially when you compare it to other CMSes of that era.

Moreover, their development roadmap is actually really solid. I like where WP is going. I unfortunately would no longer recommend it for basic blogging due to its complexity but as a CMS, there's really not a lot of free software options that can function on that level without extensive database programming and security knowledge.

This is so true. Open source and hosted commercial products fill totally different niches. Sure, there are some projects where either one would suffice, but many where that's not true.

I'm working on an open-source Wordpress replacement right now. (No link yet. I want to get a little farther before I put it out there.) The basic concept is:

1) Use Rails, to ease development of plugins and per-site functionality.

2) Don't save content or site config in the database. Save it to disk.

3) Git keeps track of your content and config, just like a regular Rails project. The user-friendly admin interface is actually just a wrapper around Git plus a text editor and some menus. Developers can work in a text editor and a terminal while the client works in the GUI.

4) Because content and config are Git-controlled text files and not database records, you have a sane workflow for developing locally then deploying.

1) Use Rails

Wordpress works on all manners of shared and dedicated web hosting because they all (or at least the majority) support PHP

There's no doubt whatsoever in my mind that WP's popularity is joined at the hip with the availability of cheapo LAMP hosting.

Indeed. This is something I considered. But Rails support is growing more ubiquitous, particularly with the rise of VPSs. And even if a hosting provider doesn't explicitly support Rails, it's often possible to install Rails nonetheless.

The biggest missing ingredient is a one-liner installer for all the dependencies. Installing Rails, the MySQL gem, and Passenger on a fresh server is not terribly hard for an experienced Unix user, but it is quite intimidating for casual users. So one of my goals for the project is to create a "worse is better" Rails installer. I.e. a really quick-and-dirty script, bundled with the CMS, which gives you a minimal Rails/Passenger/MySQL environment. It would be optimized for issue-free installation on a variety of hosts, rather than for bleeding edge features or maximum performance.

I'd actually rather see a hardcore nice CMS in something like Laravel, which is fairly similar to Rails. I don't think Rails is a non-starter but I'd like to see something that's LAMP friendly due to the ease and availability to beginners, students, and the generally broke.

I certainly considered PHP, but PHP is a worse-is-better compromise. I want a CMS where I don't feel like I'm accepting inferior tools just to support legacy hosting stacks. We already have plenty of PHP CMSs for that purpose. I'd rather meet an unmet need.

The choice of language in particular is a place where I don't want to compromise, because it's such a fundamental component, and it sets the tone for everything else. It's not just the language itself, but the community you're importing when you choose a language. There are plenty of good programmers using PHP, but the language also attracts a high percentage of the clueless. (If you're reading this and you code PHP, chances are you're in the first camp, so this isn't a stab against you.)

So the advantage of Wordpress isn't anything you could find by looking at Wordpress itself... it's the popularity and huge amount of support that brings? Makes sense.

So the "next Wordpress" won't be able to be the next Wordpress until it's popular. It'll have to be good for other reasons until then.

Yes. People who say they are creating a Wordpress replacement need to make clear what they understand Wordpress to be and what exactly they are replacing.

You got it! The same Catch-22 applies to many other products like social media (facebook), online auctions (ebay) and the email protocol.

Well it'll definitely have to be self-hosted and open source. It's a nonstarter to have to work in a particular ecosystem off the bat for me. I don't know what setup I'll want to be running in 10 years.

Not to mention - if the problem you're trying to address is the admin area / interface, and not the codebase, then a plugin is all you need.

Reply on news.ycombinator.com