Where Did It All Go Wrong?

Good to see all the old familiar faces

You all make fantastic points. Like I said in the first post, the original pre-fork was about simplicity in site design and MODx in pre-Tattoo days furthered that notion to a degree that garnered widespread attention. When MODx dropped in edit in page editing, it was like a shockwave that made the world take notice of this new kid who was using AJAX while people were still trying to figure out what it stood for.

When Ryan mentioned about how he wanted to unveil MODx to the world at SXSW, it was all about (for lack of better term) the "Apple" initiative: Make it "just work"". Sometimes the best innovation is making whiz-bang features heretofore only the domain of those who speak in forked tongues accessible to those who could care less about how something works and only want to make sure it does work.

Instead it started to try to be something for everyone. I’d cite the cries for a better blog as a perfect example of this. MODx was a CMS first, not a "by installing this package will give you a CMS, blog, photo gallery, new kitten, a four year degree in Xenobiology and a hug" magic option for anyone with a shared host account.

In my humble opinion, there is a void for some API/Framework to bridge together the open source communities fantastic projects and to unify them more easily for the common user. Imagine a single login ID that gives you an ACL that allows you to author documents in a CMS like MODx, post threads in a community forum like SMF, and comment on blog posts in Wordpress.

It’d be a wonderful world if you could do all of this and have a design to match, but open source teams have always had trouble enough in their own teams let alone working together cross project.

The team should be commended for their commitment to producing and supporting a fantastic product and a future product that its interior community seems to want. My concern is more for the silent community that I feel represents a large chunk not being properly cared for by Drupal (too techy), Joomla (too much) or Wordpress (too bloggy).

If the direction of the project is to compete with platforms like Silverstripe and EE2, then I am sure the project is on target. But I always felt like MODx was not supposed to be the bridge between your designer and developer...it was supposed to BE the developer FOR the designer (if that makes sense).

Any project can be sufficiently complicated and still succeed if its layered properly. If you expose too much of the comprehension surface to the starting user, the project becomes too byzantine for them to grasp quickly and suddenly "learning" it becomes a chore rather than an adventure.

I hate to say it, but I tried Wordpress for a small little journal project and was blown away. The way they have integrated plugin and theme management and installation is like a Synaptic repository for your website. They have "custom fields" which was sort of the idea behind TVs in the first place, and all of this is presented in a way that feels ridiculously eerie to some Thrash guy’s mockups for what he wanted the back-end of MODx to evolve into.

Yet in the end, its not a CMS and shouldn’t be treated as such. The void therefore still exists.

Let’s look at who needs what from a website and how they go about getting it. Using charliez’s idea of five points, I present to you my five hypothetical demographics:

1. Developers looking to the web as a software platform.
2. Designers/Mild Developers looking to produce webpages for profit.
3. Bloggers looking to share content and aggregate their own "cloud" of personality.
4. Businesses looking to produce and manage their own websites without hiring specialized staff.
5. General public users looking to put a page or two up on a site.
(I know I left out others like community builders etc, but five was a good number!)

Based on this, who services what tier of potential client? The general public users are definitely satisfied by Google Pages, Apple’s services, etc. Bloggers have tools like Blogspot and Wordpress to cater to them. Developers have dozens of frameworks to latch onto so they don’t necessarily need to be represented by a CMS...they’d mostly build one themselves.

This leaves businesses and Designers. I don’t mean the law firm with an intern who can use Rapidweaver for a site on the receptionist’s computer. I mean companies and organizations looking to have Sales people edit the product pages, HR updating the careers page, and that ever present intern posting up crazy pics from the last company party. This also includes the designer in all of us that got their start doing a web brochure for pennies on the dollar (if that) and trying to explain to our god awful client that we can design your site, but this doesn’t mean I can write the breadth of content needed to actually sell your services online.

To me, this was the niche MODx filled. When I look at the Showcase on the MODx website, I realize that I am not the only one who thought this. Looks like the silent voice has a bit of a voice after all.

All in all, this is merely an observation. Everything I have seen about Revo is amazing and I can’t wait to get my hands a bit dirtier with it. But the kid grew up to be an adult and I’d be hard pressed to see the resemblance if you asked me to.

- John

PS: If this post seems to ramble or be disjointed, I was trying to cover a lot of ground and cover everyone who was kind of enough to respond to my first ramble

Evo is not going anywhere, by the way. In fact it’s getting more refined with every release. As Revo pioneers in new areas (which I would argue serves #1 in a way that’s not been done before too), those might be able to be back ported to Evo. Similarly, manager conventions can be migrated up the chain so to speak to Revo. Evo is very mature and has a lot of years behind it and people know their way around it and can make it do some very cool things. If you’ve not been around for a while and missed out on ManagerManager I suggest you take a peek there for an example.

It’s not about Revo or Evo, with the other being left behind. There’s two paths here. Evo is not that different than when you took a hiatus ... just a lot more refined and capable honestly. With ManagerManager it can be back in the stunningly simple category. ManagerManager is one of the key pieces I’d like to see developed for Revo in due course. First to get it out of beta, though!

Great discussion!

I want to challenge the idea the Revo is somehow more complex and difficult that Evo. I think people are being confused by the new teminology and the fact that Revo has extra features like contexts and property set editing grids. No one, however, has to use these features or even know about them to make a site. You can actually cut-and-paste most existing Evo sites into Revo, modify the tags to the new style, and it will work fine (assuming that the add-ons are there). I converted my wordsmatter.softville.com site to Revo (including a number of custom snippets and an external database) in an afternoon, without raising a sweat.

Once you’ve built a site in Revo, the UI is actually much more friendly, IMHO, than Evo’s. System Settings are easier to find, Property Sets are much easier to work with than snippet parameters, and the caching and parsing is much more reliable -- eliminating many of the frustrations of Evo (e.g. trying to make nested elements work properly by trial and error with permutations of the caching). Not to mention the Package Manager which literally revolutionizes the installation of add-ons.

Take SPForm as an example. To make a simple contact form with it in Evo, you have to create three documents, three chunks, and three separate snippets -- all by cutting and pasting code, being careful to get the spelling right on each one’s name. In Revo, you go to Package Management, select SPForm, select install, and you have a working contact form on your site.

If you want to modify a snippet’s behavior in Revo, you go to the snippet, open up the properties, and edit them in the grid -- no ampersands, no backticks, no spelling errors. This is not moving in the direction of catering to sophisticated developers -- just the opposite.

Similarly, if you want to change a prompt used in the manager (or by a snippet or plugin) in Revo, you go to Lexicon Management, find the string, and edit it in the grid. And you can find System Settings with a search string and edit them in their grid as well -- no more searching through the various Manager tabs to find the one you want.

Another huge productivity benefit in Revo is that while you are editing a document, chunk, or snippet in Revo, you can edit the content of any *other* document, snippet, or chunk in a popup window without leaving the editor.

I honestly believe that once the add-ons are in place, you’ll be able to build a typical site in Revo in about half the time it would take in Evo. My experience is that after you’ve used the Manager in Revo for a while, working in Evo is like trying to type with mittens on.

I hope we can get away from the notion that Evo is easy and Revo is hard because, in my experience, the opposite is true. Revo is more powerful and feature-rich, but I don’t think there’s anything inherently difficult about it. Snippets, chunks, templates, template variables, and plugins, all work the same way in both, they’re just easer to install and modify in Revo.

My only area of concern for Revo is the permission system. I haven’t been able to make any sense out of it at all and I fear that it will be a major stumbling block for new users. In fairness, there are no docs for it yet so my fears may be unfounded. I hope we can provide a simplified layer for naive users to work with them while still allowing power users to go deeper and have granular control of access to various capabilities.

Currently, in Revolution (SVN), after running the initial setup and such, I can build an entire microsite without reloading the page once.

I began writing a comment for this thread, but it evolved into something bigger, so I put my thoughts about "where did it all go wrong" into a new post:
http://modxcms.com/forums/index.php/topic,37128.0.html

I think that MODx is NOT difficult to use, but many have the perception that it is for lack of more organized, simple, and thorough documentation.