The Road Ahead

I would like to take a few minutes to express my concerns on an issue that seems to be hovering over the future of MODx development. As you will see, my Subject line is relevent.

There appear to be two major schools of thought here. One is to make a system with a point-and-click configuration interface that will try to be all things to all people. The other is to make a core framework that will provide a base for being all things to all people.

Let me make an analagy here. A Windows machine pretty much limits you to whatever the designers and developers give you. It’s very easy to work with if you don’t want to do much (although the "average joe" user still can’t do a simple installation, let alone actually configure or change anything). But if you want to really get into it and tweak it, it’s a nightmare if it’s possible at all.

A Linux machine has at its core a directory full of text files. Granted, they all have different formats and use different syntax, but they are there for the dedicated and the determined to play with. On top of those are the X windows system with Gnome or KDE or whatever, or if you prefer you can administer your server with WebMin. But if you want to do something that these don’t have pretty GUI buttons for, you can get down and dirty with the text configuration files.

Same with IE or Safari and Firefox. Very difficult if not downright impossible to do anything with the first two beyond what you are given in Preferences. Firefox, on the other hand, not only has about:config from the address bar, but if you are really adventursome you can go play in the .js files in the Firefox directories.

Again, you have javascript/Ajax, and Prototype and Scriptaculous. What if all we had was Scriptaculous? If we wanted to do something that wasn’t provided in the library, we’d be stuck.

So my take on MODx is that I would really like to see a solid, spare core framework that works well; then anybody who wants to can build "wizards" and plugins and modules and autobuilders to their heart’s content. How many people do we get coming to the forums expressing their unhapiness with whatever system they were using because it didn’t let them do what they wanted? Let’s not limit MODx in the same way!

Susan, great post and I tend to want my cake and eat it too. But if I had to choose, I choose a solid framework that works for the big guys and that can be made to work for the little folk later via wizards and whatnot.

Quote from: sottwell at Aug 28, 2006, 11:06 PM

There appear to be two major schools of thought here. One is to make a system with a point-and-click configuration interface that will try to be all things to all people. The other is to make a core framework that will provide a base for being all things to all people.

I thought the core idea @MODx was the later, and that’s precisely why I enjoy MODx so much !

I think if we don’t keep this philosophy, we’ll end up loosing a lot of flexibility and ability to customize. Plus, we wouldn’t be a framework anymore, but Yet Another "One Click Install" Content Management System.

But I am not very concerned : unless I didn’t understand what I read, Jason is building an even more flexible and modular application with MODx 1.0 / Tattoo. It’s not built to be Yet Another "One Click Install" Content Management System.

Last but not least, if someone wants to build wizards, pre-installed chunks and TVs, basically an enhanced version of the default content currently shipping with MODx, as you say they can. It would not compromise the framework itself, as long as we don’t embed those pre-configurations into the application itself.

Quote from: sottwell

So my take on MODx is that I would really like to see a solid, spare core framework that works well; then anybody who wants to can build "wizards" and plugins and modules and autobuilders to their heart’s content. How many people do we get coming to the forums expressing their unhapiness with whatever system they were using because it didn’t let them do what they wanted? Let’s not limit MODx in the same way!

I fully agree with this !

One thing comes to mind : if the number of people asking MODx to be what it’s not meant to be increases, maybe we have a communication problem here... we have made clear time and again that MODx was not meant to work out of the box, that it required a least some skills and certainly entails a learning curve. Maybe we have to make sure this is very clear from day one someone registers on the forums...

And Susan : thanks for the analogy, very powerful way to explain this !

Linux
The Linux analogy is apt. However, people have made versions of Linux that install easily and come with a suite of applications to get people up and running quickly. I am sure that there are people who switch entire distros just for one with a different look, rather than changing the look themselves.

And for every web technology, there is someone that packages things up to create something greater than the sum of its parts - take cPanel / Plesk, for instance.

I wonder whether it is worthwhile encouraging 2 versions of MODx - MODx Core and MODx Turnkey Portal. In fact, as many Turnkey versions as people may develop - it may further demonstrate the flexibility of MODx if there were multiple Turnkey versions. Turnkey isn’t quite the same as "point-and-click configuration", but it is several steps towards being a newbie-friendly version and makes it clearer where the deeper waters lie.

I think there’s benefit in making the distinction - possibly even using a name for the turnkey that further distanced itself from the core, like "InkNeedle Turnkey Website powered by MODx". If MODx gets through to the next stage of the CMS award thingie and we want people to consider it a key part of the brand for judging purposes, we may need a more subtle distinction: "MODx CMS" as opposed to "MODx CMF".

Either way, this could have a separate support forum, and we would be better able to divide out those looking for a turnkey solution from those willing to work with a powerful core.

In essence, the default content would move from being tech demos to being a coherent website based on best-practices, just waiting for someone to replace the text. (Which admittedly is something that the current default content promises, but doesn’t particularly deliver.)

Further, by keeping the distributions separate, we could more regularly update Turnkey with the latest snippets, thus avoiding the issue of newbies finding problems with snippets that most of us stopped using months ago.

Hmmn. I think I just advocated going down both roads.

Microsoft
One crucial similarity that MODx has with Microsoft is that the core developers also create applications that work with the main framework/OS, and that there is a level of feedback which enables these snippet/application developers a head-start in working with new features, and a voice in core development decisions. Some other application developers do not get this insight.

Because of this inter-relatedness, and the use of the same brand across OS and application, there have been many people (particularly those that just turn up to work and use a machine someone else installed) that confuse one with the other. Back in the day, many Windows users would say they were on Windows 97 or Office 98 as they would confuse one version number with the other.

There are people who are confusing MODx with the snippets bundled with it, and discounting the advances made in the snippets because they expect things to be in the core. I suspect that many of the users who keep asking for a new release would be happier if they could download today Turnkey Website 1.4 even if it was just MODx 0.9.2.1 with a couple of basic bugfixes and the latest snippets.

Quote from: PaulGregory at Aug 29, 2006, 07:19 AM

Linux
The Linux analogy is apt. However, people have made versions of Linux that install easily and come with a suite of applications to get people up and running quickly. I am sure that there are people who switch entire distros just for one with a different look, rather than changing the look themselves.

Actually, that is my point exactly. People do not make "versions of Linux", they make "distributions" consisting of the latest know stable Linux kernal and a core set of well-known GNU utilities which are common to them all, then adding on their own selections of "extras", X.org or Xfree86 and windows managers to make up the difference in the distribution. I ran my first Linux installation on a 386sx notebook with 4Mb of ram, and I still could if I wanted to. Sure, not the latest X.org with Gnome or KDE, but what if I don’t want to? I’m perfectly happy on the command line, with vi... no, emacs... no, vi... no...

That always annoyed me with the Mandrake and other newer distributions, even if during the installation process I specifically cleared the checkboxes for the X environment, it got installed anyway.

I can see two or three Installation packages for MODx even with what we have now. I would love to be able to install a plain system with no RTE nor logging nor more than the most basic user validation system.

Well said Susan!

The other day I configured a machine to run both Linux and Windows at the same time. I did this because I realized that there were some things that Linux could do that Windows could not do. The flip side it also true.

I’m for a little of both worlds. I’m for something that has a one-click solution out of the box but also allows the developer to do more than just point-and-click.

Quote from: xwisdom at Sep 07, 2006, 11:53 PM

The other day I configured a machine to run both Linux and Windows at the same time. I did this because I realized that there were some things that Linux could do that Windows could not do. The flip side it also true.

I used to do that too, but I find it tiresome to switch from one to the other...

Quote from: xwisdom

I’m for a little of both worlds. I’m for something that has a one-click solution out of the box but also allows the developer to do more than just point-and-click.

I believe Mac OS X is exactly that, a great comprise between user friendliness and command line kind of stuff, and that’s also why I switched