May 02, 2006
Charlie Wood has a nice summary of the Spring 2006 Gilbane Conference in San Francisco. I was at the conference too and I am sorry that I didn't get to meet Charlie because I really like his blog. One thing that Charlie described that I definitely felt this year was the sea change away from ECM as a monolithic solution and toward a model of inter-connected task specific content applications. I couldn't agree more. This is my fourth Gilbane conference and the trend has been remarkable. Back in November 2004, I noted how the analysts were decidedly split on whether anyone was delivering on the one size fits all ECM vision. Then in Fall 2005, there were fewer people willing to stand by the monolithic ECM vision. This year, I heard several people say that people had over-bought content management solutions and were having difficulty adopting them due to complexity and poor usability. This year, there was much more emphasis on lighter weight, targeted, viral technologies such as blogs and wikis that are actually doing what the big software vendors have been failing to do: enable communication and collaboration with as little impedance as possible.
May 02, 2006
If you have read my comments on Alfresco, you probably noticed me mention that Alfresco is not all open source. In fact, I have said the "E" of their ECM was closed source because the features that an enterprise would need (LDAP integration, group based permissioning, etc.) were only available in the for-pay version.
Alfresco's recent announcement shows a change in their licensing model which makes the entire code base open source (licensed under the MPL). The Enterprise and Community versions will have the same functionality but only the Enterprise version will be certified. The Community version will be more oriented toward research and development and undergo less QA. This is similar to the relationship between Fedora and Red Hat Linux Enterprise Linux.
It is unclear what the difference in quality and reliability between these two versions will be. Alfresco strongly urges anyone using the software for non-experimental purposes to use the Enterprise version but says the Community version is a great low risk way to try the software and see if it works within your enterprise. This sounds a little like the ShareWare model but it does give companies the choice to release themselves from paying the Enterprise subscription if they feel like they are not getting the value out of it. The absence of lock-in is certainly a big benefit. Who knows, maybe some other company will step up and specialize in supporting the Community version. Open source is all about choices.
Another potential benefit of this move is that it may give a developer community a chance to grow around Alfresco. Before the licensing change, there was little incentive for an independent developer to contribute code if it could wind up in the for pay version. If a developer community does take root, the community version could become a viable option for enterprises in much the same way that there are many production servers running Fedora and openSUSE.
I think that this was a necessary step for Alfresco and I am glad that they did it. The only challenge is how their marketing department will create messaging around the difference. How will they encourage people to buy the Enterprise version without derogating the Community version?
Apr 27, 2006
From InformationWeek via Jeff Potts, Boise Cascade recently announced their use of Alfresco as a system to manage digitized images of invoices. The business problem was the need for recognizable images of invoices to support billing disputes. The system needed to be reliable, scalable enough to handle terabytes of data, and run on hardware that is cheaper than their mainframe's storage options. Using Alfresco, Boise Cascade was able to solve this targeted problem at a fraction of the cost of using Documentum.
Apr 24, 2006
Google recently posted its 2006 Summer of Code List. Here are the the OS CMS projects that will be mentors:
If you use one of these applications, now is the time to contribute or lobby for enhancement ideas.
Apr 20, 2006
Over the past couple of years, I have been seeing this recurring theme of companies redesigning their websites (internal and external) according to aspirations based other websites with a lot more content than they have. During the design, stakeholders envision a user experience rich with information that is personalized and interrelated. The problem is not so great with static websites because identification and production of content happens when the website is developed. If the content does not exist, it is not in the design. The presence of a CMS shifts responsibility for this content away from the project team (often consultants) and to the business users whose eyes are often larger than their bandwidth. The time of this effort is pushed later - after the project budget has been spent. The capability to change content can turn into a burden if the expectation of fresh content is communicated to the audience but effort was not budgeted for.
Sometimes no matter how much these facts are communicated, there is a level of disappointment when the site does not look like the mockup because the content is just not there or because of the extra workload that the "labor saving" CMS starts to impose. At all costs, we need to avert this letdown. Otherwise the project will not be considered a success even if the implementation was executed flawlessly. Rather than talk at the abstract level, my recommendation is to point to very specific responsibilities and get commitment and ownership at design time. Here are some good questions to ask when you hear stakeholders over-committing...
-
"We want the site to have lots of graphs and charts."
Who in the organization has the tools, skills, and time to generate images? Give them an assignment to create a weeks worth of images for dummy data and see if it gets done or if other tasks are delayed.
-
"We need the text to be in this special (non-standard) font. It's our brand."
This means that someone without graphics production capabilities will not be able to update the site. You will now have the same bottleneck as you had with your static site.
-
"We only want to show relevant information to the visitor. Every page needs to be personalized."
What do you show a visitor if you don't have anything relevant to show them? How will content producers preview and QA the site if it will look different to everyone?
-
"We want the site to look different every day with new and interesting content"
Who is going to make the site look different? The potential to aggregate syndicated content will make this goal much easier to deliver on. However, you might ask the question why it needs to look different every day. It is unrealistic to want someone (who is not an employee) visit a corporate brochure site every day. Those sites are generally designed to educate, not be a news source.
-
"The reason why no one contributes content is that the system is so unusable"
That may be the case. One way to test this is to stand up a wiki (or if you have article based content, a blog or Drupal) as a temporary site to stage content for the new site. Do everything you can to get people to contribute. If something as simple as a wiki stands in the way of people contributing, maybe you have an organizational problem, not a technology problem. This is especially the case for intranet based sites which have an alarming failure rate. Very seldom are people rewarded for their intranet contributions like other job responsibilities.
The big point here is that, in addition to making sure that your CMS is usable, you also need to think about manageability issues. Is the design something that can be maintained? Is there enough value in the site to justify an ongoing investment in managing it? Each category of content on the site must have an owner who is motivated (either because it is his job or his passion) to be responsible for the quality and usefulness of the content. Otherwise no amount of design or technology is going to lead to success.
Apr 11, 2006
If you live in the Boston area and you want to learn more about working with Plone, join the Boston Plone Users Group on Thursday, April 27th at 7:00PM. Optaros is hosting the meeting at our Boston office on Canal Street.
This session will be a tutorial and discussion for applying a skin to Plone using the best practices and file system (rather than through the web) development. Here are the details:
Rob Baker - webmaster at Oxfam America (www.oxfamamerica.org) will demonstrate how to skin a Plone site including modifying the Plone CSS files, and do file system based development using DIYPloneStyle. The example skin he will be demonstrating is for the new bostonplone.org website. This will kick-off our group effort to build a community site for Plone enthusiasts in the Boston area.
Apr 11, 2006
Last week I spoke on a panel about real world experiences about the cost of open source. Unfortunately, the panel was not publicized and no one but the speakers knew about it. Despite the lack of attendance, I found the session (which turned into a sort of round table discussion) very interesting. The following people participated in the conversation:
-
Terry Barbounis, Chief Technology Officer, The Christian Science Monitor. They are changing the way they work with software and are building many of their new applications on open source. Terry also brought along his lead applications architect, Russ Danner who had several insightful comments.
-
Chris Chicoine, Director of IT, AthenaHealth. AthenaHealth has one of the most (if not THE most) sophisticated SugarCRM deployment.
-
Maurizio Ferconi, Managing Director, Financial Engineering, Putnam Investments. Putnam has seen a huge ROI using open source at the infrastructure layer.
-
Ron Bongo, CEO, CorraTech. Does a lot of systems integration work with SugarCRM
-
Matt Asay from Alfresco moderated.
-
Me
For anyone who missed it (and everyone did). Here are some of my notes:
-
Competitors are collaborating about technology selection. Maurizio said one of his most trusted information sources are his peers at other companies. The core technologies do not make the competitive advantage. It is how they are integrated and used. Other useful information: word of mouth, general buzz, web research, and Venture capitalists [although, personally speaking, I find that VCs are always trying to push their portfolio companies].
-
The Christian Science Monitor is very active in both the developer and user communities. Terry sees community interaction as one of the real benefits of using open source software. The exchange of information and ideas is extremely valuable.
-
AthenaHealth uses best of breed over suites. They selected SugarCRM because they needed to build an extremely custom system and selected Sugar as a base because of its rapid growth and community buzz. The Christian Science Monitor uses more of a best of solution approach where they prioritize fitness to a specific use over generic best of breed.
-
Putnam uses open source to get better leverage on commercial software vendors. The open source option forces commercial software vendors towards reasonable pricing, better standards support, and more innovative features.
-
All agreed that if you choose open source in order create a heavily customized unique application, go in with your eyes wide open because you are in for some software development. Of course, this is no different from commercial software. The only difference is that open source gives you more latitude to customize if you wish to.
-
Everyone is using open source in mission critical applications. Terry talked about the responsibility he bears in supporting journalists who endure great risk to post their stories each day. Reliability is a key requirement and the panel agreed that they are finding this reliability in open source software.
Apr 10, 2006
TYPO3's much anticipated Version 4 was recently released. In the words of project leader Kasper Skarhoj:
Version 4.0 represents a milestone in the development of TYPO3 that we have been working toward for a long time. It supports qualitatively superior enterprise modules. The introduction of the the long awaited versioning feature is not to be forgotten. With the management framework for versioning that we call ‘workspaces’, versioning will be brought to the user’s desktop. And that brings a transparency which even surpassed my own expectations.
Stay tuned for a review. I am looking forward to giving it a test drive.
Apr 02, 2006
On April 23-24, CM Professionals will be holding its Spring Summit in San Francisco. If you are planning on attending the Spring Gilbane conference, you should consider arriving a day earlier and join us. If you live in the Bay Area, you shouldn't miss it.
The theme of the summit is "Content Management and the Customer Experience" and the list of speakers is an international who's who of content management:
Special thanks to Astoria Software, a provider of XML based content management solutions, and my company, Optaros, for sponsoring the event.
If you have not been to a Summit before, this is what the organization is all about – getting people together to share experiences and build professional relationships. There is no sales presence or marketing bias, just people getting together and learning from each other.
Mar 27, 2006
While not quite as trendy and chic as Ruby on Rails, PHP is becoming widely accepted as an excellent platform for rapidly building web applications. PHP runs fast and lean on a server and is easy to work in. The PHP language supports object oriented programming and there are a number of frameworks that support patterns and functionality similar to some of the Java web application frameworks. Even IBM, which invested millions of dollars developing and promoting Java, has integrated PHP into their strategy and has made some bets on the technology. JSR 223 is a new Java specification for writing Java objects so that they can be accessed from scripting languages like PHP. Not surprisingly, many of the open source CMS are built in PHP.
One valid and accepted concern with PHP is security. Over the years, there have been several rather serious exploits to PHP based applications. Part of this is the technology itself and part of it is the style with which the applications may have been developed: fast and loose. This problem is not unique to open source or PHP. I have seen plenty of commercial software products with gaping holes. For example, it used to be when you installed Interwoven TeamSite, you got a wide open proxy server that anyone could use. If you were not careful, you could have half of China connecting through you to get to sites forbidden by the government. This information wasn't published anywhere obvious (certainly not on the sales literature), I found out when the bandwidth consumption was going through the roof.
With open source, you have a better chance of finding about about these problems because they are publicly discussed within the community. When selecting software, look at the security track record which you should have access to. This article by Harry Fuecks gives some useful pointers. Once you have deployed the application, stay on the developer list (many projects have security specific lists) and subscribe to web security alerts. With a little bit of attention, you can have piece of mind about the security of your application.