| SitePoint Sponsor |
BeeStar:
I've looked at MODx, but I wasn't all that excited, but maybe I missed something. Care to elaborate a bit more on what you like/don't like with it? Is it something that is intuitive enough to give the customer access to? I think my main problem when I tested it was that pages were just one big block of content. I'd like a CMS that builds pages out of smaller blocks, making it possible to share smaller content blocks between pages, and also move the blocks to different places on a page.
It is easy enough to give to customers. This was one of my must-haves. The customer can log in on the front-end and can change content: so called front-end editing. The customer can also log in in the back-end and add pages etc. Both accesses require some training, but not much.Originally Posted by wysiwyg
The main content on a page is in one block, editable with TinyMCE. If you want the main content to be in different blocks, you have to try Typo3. That's the only CMS that has that, AFAIK.
The rest of the page is definitely not one block, but several, like navigation, header, footer, side bar, etc.
The so called Template Variables (TV) are very powerful, but do require some effort to get the hang of it.
Bee
thanks for the tip. MODx looks pretty good, but I think I'll still spend the 1000+ hours building my own (already spent a few hundred, what's several hundred more?), mainly for the learning experience. I can get some ideas from MODx though, they did a few interesting things that I could see right off the bat. I never thought about using a pop-up for editing a page... would be quite nice for previewing the actual page.
Yeah, I've looked at Typo3 to, and while it seems powerful, it's huge. There's quite a big learning curve, and I feel like it more or less demands that my customer has someone who works with their website on a regular basis. If they just edit a few pages every now and then, Typo3 is to big and messy, and way to technical for the ordinary user who just wants to write some content.
I couldn't agree more.
me too. typo3 looks really messy. my UI is so much prettier and cleaner so far. BTW, joomla allows you to put content in different places on the page. they do only have 1 main content block, but you can add additional "modules" to other places that can include articles, random articles, etc, though it's a lot of work if you don't want the same thing on the side of every page... i intend to fix that.
You might consider selling your CMS when it's done. Or open source it...
Bee
I would like to sell it, but I don't know if it would be worth the trouble (i know nothing about marketing, sales, a business, etc). I'll probably release it under GPL3 and hope I can get something out of it somehow, some day.
Wow, that's great.
Do you have a name yet for your CMS? If so, it might be wise to claim the domain name.
When your release is stable for production use, let opensourcecms.com/ know that your CMS exists. They will put up a demo so everyone can easily test drive your CMS.
Good luck with everything and remember to have fun while coding!
Bee
yeah, I have the domain name already, just in case
I'll definitely list it on opensourcecms.com if/when it's mature, stable, and open source. keep an eye out for something that starts with a z in about... a year.
The problem with the popular Open-source CMS systems is that they are too bloated. They are extremely difficult to use. I would focus on usability and flexibility. Also, is your focus making something for developers to install or making something for everyday people to install?
I agree eruna, I absolutely despise bloated software, so I'm streamlining my CMS as much as possible, thinking about allowing users to download it with only the features they want, but having an easy interface to install new features. I'm also focusing on usability, trying to make it as clear as possible and as easy and quick to use as possible. I was appalled by how difficult it was to use Joomla. I thought a CMS was supposed to make things easier, but by the time i finally had my entire site up and running correctly, I felt like I could have written my own CMS. So now I'm doing just that
I really want to focus on both. I think a good CMS needs a lot of plugins because everyone has a specific need, but I certainly can't make them all, so I need to focus on making it developer friendly, but I will definitely create an easy to use installer for the every day person as well!
Thanks for your input!
Here's an idea I've snagged from a recent post on hacker news by paul graham and slightly modified, which is to make a basic cms that can be tailored in a specific way to a specific type of site. The bloat found in most cms' is likely a result of trying to be flexible, and trying to be a solution to every problem.
It might be useful to step back and look at what the needs of a flexible CMS are rather than making a slew of flexible components, else you might risk locking people into a limited range of flexibility.
Back to the original idea of making it easy to tailor a site to a purpose. I think it would be interesting if your CMS included a fair amount of code and database generation utilities. Imagine instead of directly managing content, you manage data (tables and their rows) and relationships between sets of data. Give the admin the ability to create new forms (db tables) of data (rows in the table) without having them plug in SQL queries.
The next challenge is presenting that data. There are several ways you might go about doing this, some of which are complex. The presentation could be automated, but would require a degree of customization.
I think the final challenge is adding a degree of awareness to data. It's not enough to just present data; you wouldn't display the md5 hash of a password because it has private uses. Also, some data is predicated on the existance of other data or states. Some of these relationships are highly complex.
Your core cms would come with the standard features: providing a way to create pages, display them, manage users, etc. These features, of course, would be built around and with the same code that the code generator creates (although they would likely be sufficiently customized, ie: smart, to do what they need to do.
So, this is my idea of what the next content management system can do, it may or may not agree with you.
Although it seems unintuitive, I think offering a CMS that isn't a 100% solution but an 80% solution with the tools to get you that extra 20% would be appealing.
@peter
That is an interesting idea, but that sounds a lot like Lotus Notes to me (and actually a bit like Ruby on Rails too). If you haven't developed for Lotus Notes, it's essentially what you are describing, you create forms and views to display and your data (using formulas and code to manipulate your data if you so desire) and Lotus Notes automatically generates and manages the database and queries for you.
While this can allow you to make applications very quickly, you sacrifice a lot by letting the system decide how to store and access your data. You're locked in to the idea of only using forms and tables and lists and lose a lot of freedom. Also, to give the user the ability to create forms and such would require a tremendous amount of work to be flexible enough to just be usable, so you make it more flexible, but now you're back to the idea of it being too flexible and bulky.
In my opinion, this is unnecessary complexity. By the time someone learned how to use this amazing new interface, a PHP developer could have developed a plug in to do the same thing 3 or 4 times over, plus the plug in would have much less overhead and could be optimized much easier.
Now that is on the right track. My CMS is not going to be a complete solution for every type of website you could imagine, instead it's going to be a basic solution, for an average, non-specific site. However, it will also provide all the tools necessary to easily develop plug ins to extend it to your exact needs.
I'm also looking after a CMS who creates links and there you can delete or add the link or even edit.
thanks
OP: I don't mean to discourage you but if you need to ask about what specific features a CMS should have, then I don't think your venture will result in anything, other than a learning experience for yourself.
Drupal (although bulky and such) is probably best I've seen in terms of code organization, best practices, etc. TypoLite is another nice CMS and has a fairly clean code base.
Both are bulky, because that is where most users demand.
If you build something to simple, you will lose a lot of users. If you make it complex, you will frustrate potential users.
The most important part of implementing any software system is finding out who your target audiance is. Personally I would never use Drupal or Typo for any clients as they are simply to complex for most basic web sites.
The only constant in software is change itself
I beg to differ. A simple, well-done CMS would be refreshing.
A very interesting discussion! I am just starting a freelance web service company and am thinking about doing the exact same (sort of) to allow customers that wish to - edit their own pages once I get further into my building of clients. This is sort of to help ease some burden off me without having to outsource.
I appreciate the advice here as it is a good spectrum of thoughts on how to distribute and develop a simple CMS.
Rexibit Web Services
Don't just build it - CSS it
A good CMS in my book contains the following:
- clean and lean code
- intuitive and logical administration panel
- easy installation
- must be flexible and easy to extend, either via modules, plugins, or extensions
Functionality:
- have a logical templating system
- must have no design/information architecture restrictions of any kind
- must have the ability to create static content
- must have blog features
- must have the ability to easily turn off/disable features one does not want to use, or features that may cause overhead
- must have sections which allow for separate content sections and areas
- must have a good category and tag system
- must have custom fields (please not like WordPress)
- pages I create must allow for retrieving any combination of content parts from any given category, section, or custom field
- relational content display, e.g. section x is related to section y, therefore one can set conditions by which any content is displayed
- must have good file uploading management
That's all I can think of right now. ExpressionEngine does most of this which is why it is my absolute favourite tool of choice.
Ok, maybe I shouldn't have used the word "need." I don't NEED to ask, I know what a CMS needs, but unfortunately my knowledge is biased towards what I want. That's why I'm seeking advice from my target audience, you and your clients! Right now I'm developing the administration and backend to fit my needs, so I'm focusing on what other website developers want so I can plan for the future.
I will definitely keep complexity in mind for the average website "owner." I know how difficult it is to teach a non-savvy user to use complicated software. I hope I can balance complexity and simplicity correctly. It will be flexible, but not at the expense of usability. It will be simple, but not so simple that you can't easily adapt it to your needs. at least.. I hope so.
I don't know about no design or architecture restrictions... because that would no longer be simple... but I'm definitely with you on the reduction of overhead. I hate overhead.
What do you mean by "sections which allow for separate content sections and areas" and "custom fields" that sounds a little vague to me. Do you mean when building a template, to have custom sections for the template that you can assign whatever content or modules to it you want?
I also don't understand the last two bullets I quoted from you... elaborate?
You should develop the CMS to meet YOUR clients needs.Right now I'm developing the administration and backend to fit my needs, so I'm focusing on what other website developers want so I can plan for the future
Simple and Easy are very subjective words. I've been using computers since I was about 4 years old. When I can't figure something out real quick, I personally classify that software as "could be better". Someone that just started using computers might have a totally different perspective.
CMS software is usually way to complex for your average business man. The second you introduce "modules" or extensibility you've gone to far for most. Inserting special codes into a WYSIWYG like most CMS requires is confusing.
You can't expect a business person to learn what all those tags do. Some CMS try and work around this by building an interface to make it graphical -- this makes matters even worse. While it's certainly graphical it's usually complemented by 4-5 complex steps and insane amount of configuration. Honestly just using the tags is easier.Code:{sitemap: id=2, depth=all}
The best (although least dynamic) is to insert "modules" HTML resulting code directly into the WYSIWYG and have it visually displayed. The problem is, then the modules lose that dynamic and interactive effect of more traditional CMS software.
For instance, you could insert a realty listing module into a web page but those listings will not update or be searchable or interactive.
CMS is best when it's kept to just simlpe pages. Right now I'm working on a CDE (Content Delivery Engine) which basically pulls it data from a XML file and renders the page.
A content editor would be trivial to implement.
I developed a similar system a few years back:
http://www.sourceforge.net/projects/texocms
It's a simple editor that works directly on the script pages generated, each of which uses a template so mass switching is easy. It's an simple solution for basic editing needs.
The only constant in software is change itself
Sorry, maybe I should learn to be overbearingly specific.. I agree, I need to focus on my client's needs. That's what I meant by "my needs:" what I need for my clients. I am a web developer, I'm not building a complex CMS just for my own site... (that would be a silly waste of time!).
I also agree that you can't expect a business person to remember how to use tags, or even to learn what a tag is/does. An alternative is to create a UI, but the more flexible the UI is, the more complicated it gets, and the harder to use it becomes. I do know that from experience. I believe I have a good grasp on what a good UI is, and I'm doing my best to make a good UI that appears simple enough to the non-tech-savvy, but has advanced features that can easily be found and understood by the more technically inclined. I don't expect my clients to know or even learn how to create their own templates, change their site layout, or write their own modules or even install pre-made modules. That's what they pay me for, they just need to change the content and I need an easy way to make complicated changes to their site.
My CMS sounds similar to your CDE, only I'm using a database instead XML files to store the data and content. I guess you could say my CMS is actually just a controller for my web application framework I built first. All the CMS does is give you a nice interface to design and layout the site and edit static content and manage users. Anything beyond that is delegated to the appropriate module that does it's own thing and spews out HTML for the appropriate sections of the template.
I know I can't build the best CMS for everyone, but I'm just trying to reach as many people as I can. I want to make their lives, or at least their work a little easier.
Posting Permissions
Bookmarks