Free Software invariably gets into the Enterprise as a skunkworks project. The managers you are talking to have a budget for a business portal. They want the project to succeed, so that they look good, and they aren't really interested in having money left over in the budget when they are done. They are shopping around for a solution, not a project.

If you really want to get Free Software into your business the proper way to do so is talk the manager in charge of the project into spending most of his money on a proprietary product that won't actually work. There are plenty of commercial offerings out there that are likely to be a bad fit for your business. Talk the manager in question into purchasing one of those, but make sure that he takes all of the credit. It shouldn't be hard if you spent the first part of the purchasing process pushing for Free Software.

Watch the portal project crash and burn.

Now fire up a basic portal on the Free Software platform of your choice. If possible pre-populate it with data and tie it into your existing authorization and authentication mechanisms. The idea is to have a working demo of most of the functionality that the executives wanted.

The downside of this method is that, if you do it enough, you eventually end up being forced into management yourself.

Drupal case studies.
http://drupal.org/cases [drupal.org]

And for Liferay
http://www.liferay.com/web//guest/products/portal/stories [liferay.com]

Next question?

Linux got accepted because some big vendors like IBM started supporting it. Until you can get some big trusted vendors to start supporting these apps, they won't see large-scale deployment in the enterprise.

I just used Symphony today for the first time and I must say the polish on it is really impressive. It was extremely easy to use and I didn't have any compatibility issues with my old MS Office created documents.

I did notice however that in Symphony Documents, my options for creating fields were all missing! A minor nuisance to be sure, but fields are nice...

In any case, the lack of an OpenOffice database equivalent made me switch back to OpenOffice. I kind of get the feeling Lotus Symphony was designed for corporate desktops, where OOo Base wouldn't be all that useful.

That's perfectly doable on Linux, and SunRay systems have been doing it for years...
There are all kinds of ways to do this... LDAP, Kerberos, SSH keys and client certs (if you've authenticated to your user account and got access to your homedir then all your user specific keys/certs are there)..

On the other hand, having a single password to access anything is not the most secure option, it's a case of convenience over security.

JBoss has been pretty good at penetrating the corporate data center. I think Glassfish will do well also since it's backed by a company that already has a presence in many corporate data centers.

Since Liferay is a J2EE app, it should be a little easier since most corporate customers are already using the J2EE stack. Liferay also offers "enterprise support" if that means anything.

This might be a good time to call a Sun rep and give them your requirements and tell them you want an open source solution.

There was talk of Java Enterprise System being open sourced but I don't think that ever happened. If that's a more palatable solution for management, it might be cheaper.

Sun isn't very popular on here but they're good at getting open source into the enterprise... with support.

First of all, I would like to thank you for a series of excellent posts. Seriously, very well done.

To a certain extent my responses have been tongue in cheek. I have always liked my direct report managers. In fact, I have never worked for someone that I didn't feel had my best interests in mind. Now that I have some managerial experience myself it was clear that my previous bosses had a great deal of skill and knowledge. In fact, I would consider most of my bosses to be more intelligent than I am. I'm fairly good at gluing software together, but that's no big trick.

However, when I used to work for big business there were always plenty of cases where different departments came together to pick software. On more than one occasion the group I was working for was over-ruled and some incredibly Byzantine software was chosen. In one case in particular the manager of my group decided to basically set up a competing project that was ostensibly just for our division. He used Free Software, mostly because it fit into his budget, but also because he had technical people that he trusted that told him they could make it work.

The big budget project crashed horribly, and my manager got promoted when they picked up his project for the whole company. What he did was a bit of a gamble, but not too big of a gamble. After all, if the big budget project had worked he could have simply buried the skunkworks project.

This lead to a complete reversal on the use of Free Software within the company. It went from being strictly forbidden (including crazy things like the GNU tools for Solaris) to being fairly widely accepted.

There are plenty of examples of web services running on Open Source for 'enterprise' use - groupware, CRM, accounting, the works. Some of these packages are very good.

Its hard to be specific/determine what you're trying to do without knowing more specifics as to what you're looking for. Of the groupware projects I'm aware of, I know the following have a fair amount of support/use:

* Plone CMS [plone.org]
* OBM [obm.org]
* eGroupWare [egroupware.org]
* Drupal [drupal.org]
* Typo3 [typo3.org]

Of these, I know that Plone, Drupal, and Typo3 are all "platforms" for developing, managing, and extending content. I seem to recall either eGroupWare or OpenGroupWare extend/integrate with MS Office products. No, it's not going to be the level of integration that Sharepoint stuff offers, but it's something to mention, at any rate (and isn't going to have the massive licensing costs + perpetual lock-in that a MS solution has*).

Plone, in particular, has a lot of support and corporate/"enterprise" use. From their site:

That alone is impressive enough; but also consider some of the notable companies which utilize Plone [plone.net] in/for a variety of purposes:

Akamai (yeah, that Akamai - the guys who load balance Microsoft web servers)

Nokia (QT Software stuff)

MyCity ("real time monitoring system for Cities, Towns, Districts or utilities. It makes use of the GPRS service offered by the various GSM network operators")

Discover Magazine

Novell, Inc. (for enterprise services)

NASAScience (public site for NASA's Science Mission Directorate)

FSF (yeah, those hippies)

universities, university science/it departments, hospitals, public/government sites... the list goes on.

Those are notable company names, and at least in the case of Akamai, Novell and Nokia, everyone in IT should know about them. They're also some fairly diverse (and expansive) implementations using the same central CMS - and they're not shackled to a single software backend, able to run on any OS and server combination they could imagine.

* The cost factor associated with MS solution lock-in is a big consideration, bigger than just a simple argument of something like "OpenOffice vs. MS Office". With a web-based, top-level technology like this, it's much, much more important to keep the technologies used "open" - because it is the top-level interface to all your data. You can not move away from a closed package on the backend without moving the entire system, at once, to something open (more often than not, with MS). You're basically stuck with that stack unless you want to start over; there's no ability to independently consider parts of the stack and replace them, as there often is with open systems.

    Hehe.

    I don't have an MCSE, because I don't want one. :) I was talking to some folks who are experienced with the test side of that. From what I've been told, you're drilled on the test until there's no way you couldn't know the answers, and then you take the test. So, you're exactly right. Unless the real world problem arises from a test based problem, it's very likely they will have to call someone for help.

    I won't say that's true of everyone though. Say I did do something silly like get an MCSE. I already have the real world experience to get myself through most problems, and the knowledge base (in my head, not the MS KB) to work through the rest.

    I'm just particularly annoyed by MS servers in general. Lets take two recent examples.

    In both examples, the hardware failed on the servers. One was a Win2k Server running MSSQL. The second was a Linux ingress mail filter.

    On both, after determining the hardware failed, we were given the option of moving to another server. In both cases, we had similiar but not absolutely identical hardware available in house. Both were in production for quite a while, so there was no good option for obtaining identical hardware.

    On the Win2k server, we moved the drive, and rebooted. The drive controller wasn't identical, so Windows would panic at boot time. The solution? An in-place upgrade of the OS with the original media, and then do all the updates to bring it back to current again. This took hours. And yes, I consulted the MS KB. I already knew the answer, I just was hoping there was a better way.

    On the Linux server, I wasn't even present for it. I gave instructions to the site over the phone. "Move the drive to the other machine, and turn it on. Besides the time of physically moving the drive, it was up in a matter of minutes.

    On both, I've been playing with the hardware to diagnose it down to the part since then. In both cases, it was the motherboard. The Win2k server is staying in it's new host, because we don't want to do another in place upgrade. The Linux server will be moved back to it's original machine when the new motherboard comes in. That will account for maybe 10 minutes of downtime.

    Because of their duty, we have different windows to work in on each. The Win2k server, being a SQL server, has to be available. The Linux server as a mail ingress filter, can be down for a few minutes and people don't really notice. The mail will still be delivered, just with a bit of a pause for new inbound mail from outside. After hours, people will still be hitting the web sites that require the SQL server, but people won't notice that it took an extra 10 (or even 30) minutes to get their mail delivered.

    Diagnosing is a lot different with both. The Windows event viewer doesn't give much useable information most of the time. The Linux system logs give a lot of information. In the specific case of the machine above, it never got to init, so syslogd never started, but I could see what it tripped up at during boot with the kernel messages.