Do a google search for ‘FileNet Reviews’ and you quickly discover that information about their systems, I mean real information is more than just anemic. Why is it that this company who, as all the press statements I’ve found about them, state: “Established in 1982, FileNet practically invented the field of Enterprise Office Document Imaging Systems.” If you’ve been searching for information I’m sure you’ve seen this regurgitated garbage on more than a few websites.

But this is 2006, why in the hell is there no more information out on the Net about a company which is apparently so damn monolithic and powerful. Well useful information at least. In comparison if you do a search for Oracle you will find MILLIONS of useful hits on the topic of how their product works, tips on how to use, when it’s a good solution for you, etc.. But FileNet seems to be lacking all of these things.

It’s as if there is some sort of global conspiracy to keep this information out of your hands, either that or all the people running this fucking system have no idea how to setup a website using their Web Content Management System (or any other system for that matter).

Now, I did manage to find a book, with emphasis on the singularity of that statement. Even in the publishers comments for this book apparently there is a ‘dearth’ of information on the product. Well I’m going to attempt not to mitigate this lack of information as much as I’m going to despell actually provide some sort of user opinion on FileNet. Something, anything, useful for people actually looking to learn about the products this company serves up for more than a pretty penny.

Now a little background. I am a systems engineer. “Wow”, you say, “but who the fuck cares.” You don’t have to, so you can stop reading. What is important is the side of FileNet I have seen, the side of maintaining and running the system. I have gone through the experience of installing it, fixing it when it is broken, working with their support, and even developing extensions to the system. My aim here is only to shed light on these aspects of FileNet.

A few things which caught my eye about this entire FileNet project from the beginning, these all represent red flags for me, and may be something you should take into consideration if you are looking at their system.

• System requirements don’t line up: You are always struggling with a way to decode their documentation just to understand what needs to be installed, and how many systems are required to do it.
• Services do not co-habitate: Many features of the P8 platform require multiple systems to impliment because they ‘do not play well together’
• FileNet’s support structure: It appears as if their entire business model is based on you purchasing support ‘time’. This means that in order to get anything done, installing a major service pack, adding customization (in the form of JSP code), continued debugging; All requires you to pre-purchase support hours to do so. It’s not cheap in the order of tens of thousands of dollars for their support time.
• Sales staff persistence: They sales reps have the definate yes men attitude, anything can be done with the system. While I’d say this is pretty standard, FileNet goes further in that simple features which really SHOULD be in the system they claim can be done easily, cannot.
• Don’t eat their own dogfood: I know from an internal Engineer that FileNet themselves don’t even use the product for their internal use. This certainly shows when attempting to get support, they are constantly asking for the same information over and over again. Certainly not representitive of a company which specializes in document management and workflow.

To begin I must state that I’m not qualified to actually install FileNet. This is from their standpoint. I literally am not certified to install the system, so they REQUIRE a systems technician to come out and install it with me. At first I baulked at what as stupid idea this was. This isn’t an AS400 system, or a Sun data cluster. This is just an installation of a software package on a Windows Server. Until we did the install, then I understood why they needed to send out a technician.

Essentially the experience goes like this; I setup an Active Directory domain, we begin the installation procedure. Some incredibly convoluted error message (java errors) comes up. Technician calls the engineering team in Mounds View, CA. they spend one hour on the phone trying to figure out what happened. Rinse, repeat.

Going on like this the installation actually took 3 days. I’m not kidding. They didn’t even tell me what they fixed when they did, though from observation it usually involved a manual edit of some config file and a retry. Basically they were working out ways to override the errors they were encountering, not determining why the error was there in the first place; Just how to get passed it. The actual CPU time for the install, this includes actively filling in configuration info, clicking through the menus, etc. was about .5-.75 hrs of work. The remainder of the 3 days was the diagnosing the inevitable error which would follow each step.

I could keep going into the details, but I won’t. I’ll move on. Now that we have the system installed, I can maintain it though the myriad of convoluted and very old, clunky management interfaces. My Application log keeps filling with errors generated from FileNet. Engineering support says this is normal (WTF?). Occasionally parts of the system freeze up and simply stop working, they also do not notify me of this fact, the workflow through the product just stops. This requires me to restart various portions of the application. Including the three different management interfaces, several services across the two/three separate servers responsible for the install. Eventually it gets moving again.

Now this is something I found to be interesting about this application. We were using their e-Forms product, as you would expect we are developing forms for use online by external users. The fun begins here. If you are a random website visitor, you MUST log into the system in order to fill out a form. Fine you say, except for the fact that there is no automated way to generate an account for you to fill out the form, you have to have an account within the Active Directory system underneath.

The usefulness of the eForm just went out the window. Of course, there is a hack, well there are many hacks to this system. Most of which I ended up having to develop on my own with a little help from some FileNet engineers. I won’t go into details, but I will explain that a week into the install I was already writing my own JSPs to handle log in events with no help from them whatsoever.

Next roadblock to eForms. Now that I have people logged into the system, I would like to send them an email confirming we’ve recieved the form. If you read that correctly you probably already anticipate what I’m about to say. It’s not happening, until I receive some other hacked together code from FileNet and then back to writing my own JSP to extend it to send emails to external users (external from the Active Directory system).

Essentially most functionality you would expect from this system needs to be developed separately from it. I’ve already written so much JSP into this system that I may as well just keep going and build an alternative to it entirely. The argument from our management has always been “FileNet is SOOO powerful, but we just don’t know how to use it yet”. This is a double edged sword and in my opinion anytime you get near it your going to nick your juggular. First of all, it’s not powerful at all if it can’t handle the basic functionality that ANY modern enterprise system does.

If it can’t replicate an online form a 12 year old can create in PHP than what use does it have if I have to extend their system to do so? The only thing powerful about this system is it’s workflow and document management capabilites. Of course the flip side of this is it’s expensive, closed source, takes up WAY too many servers, used Java applets liberally is incredibly slow and overcomplicated. I have a hard time finding ANYTHING good about it.

For the money and time invested I would rather have used Plone, spent the immense amount of time getting over the Plone learning curve with regards to it’s workflow and I would have had as system more capable, faster, looks better in far less time. Not to mention the ability to avoid Java applets, which in and of itself is a support nightmare. This is all more than my 2 cents, it’s more of a change jar, but due to the lack of information out there about this whole system I needed to put something online for you people.