The Program Officer for the Digital Library Federation in CLIR, Rachel Frick, has asked me to make a proposal for moving DLF activity to a WordPress site. The following is a proposal was approved by CLIR on 3 December 2010.

In 2008 the DLF began to explore the creation of a new site for http://www.diglib.org. I worked with the staff of the then stand-alone DLF to determine what services they wanted to facilitate on the new site. DLF then hired eCoreExperience to build a Drupal website to the specifications we had devised. That Drupal site was never rolled out live because by the time it would have been ready in mid-2009 the DLF was undergoing a major transition and being re-absorbed by CLIR. At the time it was thought that DLF would likely not have a separate website, so there was no point rolling out the new Drupal site.

Now that DLF if again embedded in CLIR and Rachel is on board to lead the program, it has become evident that a separate website would still be desirable. However, many of the assumptions made when building the Drupal site for DLF are no longer valid. The old DLF expected to have a staff member trained in Drupal administration available full time to manage the site and keep Drupal up to date, the new DLF does not have that kind of staffing available. The old DLF wanted to manage workgroup communities on its own website and built collaboration tools right into the Drupal site, the new DLF expects workgroups to use a variety of tools sponsored by participating institution and has no desire to manage their collaboration on its own site. The old DLF had administrative staff to help maintain the content of the website, the new DLF has only its program officer to maintain the site. So while DLF community members have impressed their need for a separate website on Rachel, she has decided that rolling out the Drupal site as designed in 2008/2009 will not be possible.

Instead, after some discussion, she asked that I write a proposal for rolling out the 2008/2009 website minus the collaboration tools as a WordPress site. This means keeping the essential structure of that site (and not redoing the brainstorming and community discussions that went into its development) and keeping the graphical look of the site (saving the expense of a redesign), but reimplementing all of that on the WordPress content management system. WordPress is much easier to maintain than Drupal. In fact, WordPress can easily be maintained by a single person with little knowledge of content management systems or WordPress in particular. Its content can also be updated easily both with its own administrative interface and a host of desktop tools for the Mac and Windows. WordPress can also be implemented on the same hosting service acquired for the Drupal website.

This proposal describes the deliverables, costs, and timeline for a move of DLF content to a new WordPress website.

Building the WordPress site

The new site would be first implemented at http://wp.diglib.org and once approved would be moved to the main http://www.diglib.org domain. The design of the Drupal site would be left intact by using most of its CSS and HTML, even though some of that was a bit awkward in implementation. Instead of questioning these choices, I would simply endeavor to make sure they work in the WordPress context. I already have access to the DreamHost environment in which the Drupal site is implemented and where we would also mount the WordPress site. The tasks would include:

• combining Drupal HTML with a basic WordPress theme to build a new DLF theme in WordPress
• developing WordPress equivalents of the menu navigation scheme found in the Drupal site
• developing tools for managing sidebar content on the site
• devising a way for the front page of the site to be special both in design and content, as it was in the Drupal implementation
• reimplementing a "community feed" for the homepage sidebar
• moving the content of the Drupal site over to initial WordPress pages, though this content is about a year out-of-date and will still require plenty of editing by Rachel before it is current and correct
• implementing a custom Google search for the site so that we can index more than just the WordPress content

This task would take no more than two weeks. Some of the choices would require consultation with Rachel.

Deliverable 1: A fully functional WordPress site, visually similar to the 2008/2009 site.

Cost: $2,600

Facilitating failover to old content

Much of DLF's history is still embedded in the old DLF site currently available at http://old.diglib.org. Rather than spend a lot of effort moving this content into the WordPress site, I would recommend we implement a graceful failover so that any request of http://www.diglib.org that does not exist on the new site automatically retrieves content from the old site instead. This way all existing links and bookmarks will still resolve to appropriate content. Tasks include:

• developing code in WordPress to recognize the lack of content and redirect to the old site
• developing a "404 file not found" message on the old site that would be appropriate even in the context of a request being made on the new site

This task could be accomplished during the same weeks assigned to the first deliverable.

Deliverable 2: Graceful failover from new to old site so that no links break.

Cost: $500

Simplifying the backend

Right now the management of the DLF website at DreamHost requires occasional changes to the DNS information stored at NetworkSolutions. This would be particularly urgent should DreamHost ever change the underlying machines hosting the site. It would be much simpler to move DNS management to DreamHost so that most of these changes would be managed automatically by DreamHost itself. This would leave the domain renewal and whois information managed at NetworkSolutions, with DNS and webhosting managed at DreamHost. Tasks include:

• setting up equivalent MX (mail) records so that diglib mail keeps flowing
• changing namesever entries at NetworkSolutions
• testing new setup to be sure changes are in effect and not hindering service

Deliverable 3: DNS managed at DreamHost.

Cost: $200

This task could be accomplished during the same weeks assigned to the first deliverable.

Availability for support

WordPress is very easy to manage and has been getting simpler to work with as the content management system matures. However, the DLF theme will be a bit of a kluge since it will be adapted from Drupal with some odd choices still intact. For the first few months or year DLF may want to have assistance available to tweek the theme and help with advice and instruction. I believe it will be possible to run the new site without any support, so this is an optional item that may just increase the comfort level with the new system. Tasks include:

• limited theme changes to accomodate evolving practice
• availability to answer website related questions by phone and email
• occasional troubleshooting of content issues (fixing HTML in posts or pages)

I would propose a retainer equivalent to roughly one hour of my time per month, for no more than a year.

Deliverable 4: A support resource for Rachel and other CLIR staff managing the DLF site.

Cost: $150 per month

Putting the picture together

As a consultant, I do not charge by the hour. Given my availability, we should be able to accomplish this within a month of approval of this proposal. The costs above will be charged as a flat rate. These costs total $3,300. I will submit an invoice for $3,300 one month after approval. If a support retainer is desired, I would submit an invoice for $300 every other month after that until DLF decides it no longer requires the support.