Speccing A New CMS

[oharab]

We are just starting to think about looking at CMS at work.
What sort of features should we be looking for?
What questions should we be asking potential suppliers?
What makes a good CMS and what makes one CMS better than another?

Ben

----------------------------------------------
Ben O'Hara

http://www.RobotParade.co.uk

"Where are all the stupid people from...
...And how'd they get so dumb?"

NoFX-The Decline
----------------------------------------------

 

[ganast]

My opinion of a good CMS:

1) content/design abstraction. This is paramount to having a good looking site. Most (90%+) sites built on a particular CMS look like all the rest of the sites built on that CMS...

2) Ease of use for the content creator. (Between this point and the above point, I would be hard pressed to decide which is more important... its like arguing about marketing vs. R&D)

3) Integration of several concepts that masquerade as "CMS"... those being: Workflow (Process) management, Document Management, and (if you can get it) Groupware. Many times a company that has one of the above will call it a CMS... its not, a CMS should do all the above (maybe not Groupware), should have good design abstraction, and should validate to w3c standards.

***So, how can you tell? Look at sites made with a particular CMS... are they good looking and lack content? They are good at #1 and not so good at #2. Are they all the same but produce tons of content? Better at #2 than #1... and probably are limited to one part of #3...

Here are a few good articles:

http://www.alistapart.com/articles/cms1/

http://cms.filsa.net/archives/cms-list/

- reading these archives should really give you all the info you need. The main list is "off air" right now, I have not been keeping up with it, so I don't know why...

http://lists.cms-forum.org/pipermail/cms/

- more archives.

...and see my posts in the "Best CMS thread..."

--gabe

 

[jimoblak]

My biggest concern is ease of use for authors. I have tested several systems and found them to be too complex for the average person. While all the bells and whistles of groupware-like systems are nice, they can be very intimidating to an author that quickly wants to place something on a site. The ideal CMS would offer a choice between a simple user- and an advanced user- authoring interface. Has anyone seen one like this?

- - picklefish - -
Why is everyone in this forum responding to me as picklefish?

 

[ganast]

Its hard to say what you would need without knowing what you have looked at: big CMS's like vignette?, little Free Software things like Nucleus BLOG portal?

How you intend to implement the design and maintenance of the software: are you a one person shop or a multi-person IS dept with graphic designers, programmers, and sysadmins?

Do you have a databse that already contains a large part of your work that the CMS will need to be able to interface with?

For instance, Nucleus is really easy to use. I have not found a writer yet that cannot manage Nucleus, however, the rest of the system is so limiting that it is useless for any environment with more than one writer... its BLOG software after all.

If you are looking specifically for a "CMS" for writers, you might look at Goose Quill as its interface is specifically tailored to writers and its "frontend" is designed for interaction with the writer's market. It does not have most of the features that you might expect in a "CMS" but it will take very little technical know-how to set up and manage, and ease of use will be very high.

On the other hand, many larger CMS's allow the admin to tailor the backend to limit what a particular user sees. As a matter of fact, many will (should) even allow you to do "front end editing" such that a writer never has to see more than his/her stories in a pasword protected part of the "web site," possibly with comments from the editor, and an interface to upload (cut/paste) new stories.

The latter will of course require more work from an IT department, or from a third party developer to customize your setup.

Ease of use IS paramount... I agree. The best way to evaluate a system is to see how someone else uses it. Only then you can judge actual productivity first hand. You will also see what parts of the system are actually used, and what parts are... really expensive gimmicks.

I am partial to Free Software as a concept, and I believe that any organization can get a better result from Free Software simply because they do not buy a canned product. They buy a consultant or a developer that will custom tailor a "canned" (but open source) product to the buyer's need. The consultant does not pay a annual API license fee to the software developer, and can in fact see all of the source code. This eliminates a significant cost overhead and "accessibility overhead" for the consultant. Usually the consultant is, like you, a small company, and is therefore much more flexible, innovative, and open to the possibilty of using software other than "their own."

All things considered: a great advantage in most cases.

--gabe