The committee building the 2008 revision of the main Cornell Library web site chose Drupal after evaluating several other options for source code control systems. There is not much documentation of the early stages of the decision making process. It takes a long time to evaluate these systems, and there wasn't much of it. Several other options were considered early on, but there was only time to concentrate on a few with high potential.

Here are some notes from the CUL Webvision Committee's decision process: 

Minutes of Nov 16, 2007 meeting


Topic of CS vs. Dreamweaver/Contribute was brought up and Matt clarified that it may have been simplifying things to think about this process as a choice between these two products, but that in reality, many other options exist

  • Response: This may be true, but there was a good deal of effort that went into selecting CS in the first place (3 years ago) and we're not in a position to initiate another search for a CMS at this time.
  • This opinion was not shared by everyone and by the end of the meeting it was suggested that perhaps we should look into other CMS solutions in-depth since CS was chosen by a larger group for different reasons
  • possibility that it's not the best option for this project

Dec 5th meeting

Interview with Dave DeMello re. ILR use of Commonspot - pros and cons.

Does well:

Templates, security issues, enforce initial design, story types, manage and propagate change with lots of granularity

Does not do well: Custom elements (e.g., no support for collection data types), has API but does not offer enough benefits, can't share elements easily among CS sites, can only send 'recipe' and others have to recreate it.


ILR is a CS success story because CS does for them what they wanted done (page driven site with discrete content added organically, incrementally, hierarchical site, collaborative authorship)There are a thousand content management systems out there - evaluation is a huge deal - some might be better for Cornell than others but he would pick CS again for ILR - he has peace of mind about succession planning - there is a supportive community on campus around CS - if in 5 years CU wants to move to another system we should all do it together.









There was no mention of platforms again in minutes again until May 22 meeting when decision had already been made.

Technical Subcommitte 

Technical Evaluation by the Design Subgroup

CUL Vision Report - Web Site Development Recommendations

Platforms finally evaluated

  • CommonSpot 5
  • Contribute/Dreamweaver
  • Drupal

Contribute/Dreamweaver was eliminated quickly

  • produces a static site that is hard to maintain
  • no CUL system integration or content re-use 

Commonspot vs. Drupal

  • Open source provides options for flexibility, debugging, testing that proprietary code and database definitions does not.
  • DLIT has very little ColdFusion expertise, lots of php experience.
  • Commonspot 5 required upgrading servers to ColdFusion 7 and upgrading all 16 sites from previous version.
  • Commonspot costs $5500 per year, upgrade to CF7 would be around $2500. It also requires a dedicated server.
  • Drupal would be implemented on existing servers without interfering with their current uses.

Conclusions of the report


Recap and the Big Picture

Both CommonSpot and Drupal can be modified to offer relatively similar
functionality to the end‐user. However, we believe there are significant differences
between the two products for developers and content contributors. As developers,
we believe that Drupal's accessible, modular architecture will allow for more rapid
development of more sophisticated features.

For content contributors, we believe that CommonSpot may have the upper hand
out of the box due to its page centric approach, but we are confident that Drupal
offers greater potential for creation and maintenance of content and will require a
concerted effort to ensure a positive experience for the contributors. In addition, it
will allow contributors to take greater advantage of current Web technologies like
tagging, RSS feeds, blogging and embedded media.

A key point to take away here is that no matter the platform, a significant amount of
development is required and necessary for this project to succeed. It is important
that we acknowledge this fact and make a conscious effort and commitment to
create the best possible Web experience for our patrons. Make no mistake, we could
move forward with Drupal and end up with frustrated contributors once again, a
repeat of our CommonSpot experience, if we do not appropriate the resources that
this project deserves and requires.

We believe that at this point in time, Drupal is the right content management
software for the CUL Web site. Drupal offers flexibility, a vast array of custom
modules, ease in the implementation of expected Web 2.0 features and out‐of‐thebox
interactivity with a variety of web services.

In addition, Drupal's infrastructure will allow us to leverage existing programming
expertise and make possible continued development and maintenance within CUL.
Finally, Drupal offers flexible, forward‐thinking functionality that is not tied to the
agenda of an outside entity but instead offers the chance for CUL to create its own
path in developing user‐friendly information landscapes.

  • No labels