Creative Services

Web Leadership Team Meeting Minutes

Web Leadership Team Meeting
August 13, 2007
2:00 p.m.
Cooper Library, Brown Room

Attendees: Dave Dryden, Carol Usher, Terri Vaughan, Carrie DuPre, Matt Bundrick, Dave Sharpe, Meghan Boyle, Elizabeth Anderson, Jamie Brown, Brian Cass, Gordon Cochran, Debbie Dalhouse, Jeannie Davis, Pam Draper, John Gouch, Tharon Howard, Chris Jones, Gary Malstrom, Walker Massey, Kelly McDavid, Patrick McGee, Tenneil Moody, Sandi Piazza, Peter Togel, Barbara Weaver, Tim Whims, Kathy Wright


Unable to attend: Tim Bowen, Pam Bowen, Amanda Brock, Neill Cameron, Billy Cook, Robin Denny, Ron Grant, Joel Greenstein, Barry Johnson, Teresa Morgan, Rich Pak, Bruce Rafert, Pam West, Sean Williams


Minutes:
Introductions by attendees.
Agenda items were reviewed:
1.    Web Site Guidelines
a.    Web Leadership Team members will have their names and emails listed on the OWS Web site.
b.    We will recommend to webmasters to use standardized URLs beginning with www.clemson.edu/XXXX.  (see Note 1)
c.    The standard for page titles will generally go from local to broad, left to right. OWS will continue to investigate the ramifications of changing this naming structure within Cascade, and search engine impact.

2.    WLT Mission will be reviewed for approval at the next meeting.

3.    Proposed Adjustments
a.    Cascade upgrade to V5 is due to be released at the end of September. Individuals who wish to attend the Hannon Hill Cascade user conference in Atlanta on September 30 need to register online by August 24. (See Notes)
b.    “Quick Links” will be changed to “University Links” and will be an optional element on pages using Cascade.
c.    The gateway “Future Students” will be changed to “Prospective Students.”
d.    CCIT is creating an “Email” link for the top banner to go to a page listing all University email applications, replacing the current “Webmail” link.
e.    “Student Life” will be added to the categories navigation on the left side of main pages, with content provided by Student Affairs and maintained by OWS.
f.    The “Computing and Libraries” category link in the left-side navigation will be split up to bypass the middle page and go directly to each site. The PSA is reviewing to determine if their link will also bypass its current middle page and go directly to its site.
g.    The “Visitors” gateway will soon be “Visitors and Parents,” wth a section specifically addressing parent needs.

4.    New Opportunities
a.    University Video Initiative: Kathy Wright presented CCIT’s plan for expanding video capabilities, particularly for academic needs on campus. Under Kathy Hollen’s group, a procedure is being built from the ground up to work within the University infrastructure. No implementation date is anticipated.
b.    Video Hosting by BrightCove: Dave Sharpe presented an interim solution to campus video needs, through vendor BrightCove. This vendor will allow for a Web TV network that can be branded “Clemson“ and use a Clemson URL. Costs would be covered by Advancement. Who would control content has not yet been determined.  (See Notes)
c.    iTunesU: Barbara Weaver presented updates on iTunesU’s pilot year at Clemson. Right now, faculty and staff can receive training from Barbara, covering copyright, pedagogy and audio casts. Group discussion decided that there would not be a free for all; general content will need to go through each department’s approval process.

5.    Other Issues/Group Discussion
a.    Colleges are not easily found on the current site. Discussed making the page more graphic and listing single majors in multiple ways (like listing both “Art” and “Visual Art”).
b.    Mentioned need for future focus groups to see how audiences use the site.


Next Steps:
Next meeting will be September 17 at 2:00 p.m. in the Cooper Library, Brown Room. To add an item to the meeting agenda, email Carol Usher at cryan@clemson.edu.


Notes:
•    Naming should use all lower-case.
•    Multiple word should be separated using a hyphen or underscore, a hyphen is easier to see. It is easier to read and remember urls this way. Separated names, by hyphen or underscore also work better with search engines.

•    Never use a space in a file name, web page or folder name.


Here are some Sample pages
http://www.clemson.edu/sample/
http://www.clemson.edu/sample/department-rotc.html
http://www.clemson.edu/sample/department-civil.html
http://www.clemson.edu/sample/department-arts.html
http://www.clemson.edu/sample/department-students.html

And the sample headers

    

Notes from Hannon Hill Evaluation

Roles

How can we resolve Roles issues with regard to users who have multiple roles for accessing multiple sites?

We have many users who contribute to more than just one site. They may be part of a group with the Publisher role for their own site, and a member of a group with the Contributor role for a different site. This has become a major problem for us since, as I understand it, if the user is a Publisher for one site and they are granted write access to another site as a member of a Contributor group, their Publisher role takes precedence and they can therefore publish out the site for which they should only be a Contributor. How can we configure roles appropriately to allow only the access a user needs for each site for which that user has some responsibility?

Hannon Hill’s Response:

Roles were designed with the assumption that if a user held a manager or publisher role, then they could be trusted to handle the access appropriately for sites they neither manage nor can publish from. We disagreed. While there is no solution to this within Cascade, CCIT’s Michelle Wood had an idea that would not only address the Roles issue, but would also automate the user/group creation process. Michelle will develop an automation process that will allow users, at the login screen, to choose from a list of sites to which they have access, which site they want to work on. Once the user chooses their preferred site, the process will create a unique login name, invisible to the user, based on the role the user has for that particular site. If the user logs in and chooses the sites they can access as contributors, Michelle’s program will customize their userid, once they have logged in, and automatically allow them access ONLY to the sites for which they are contributors. If they choose, at login, to access the sites for which they are publishers, then once logged in they will only be able to see those sites.

This user name would automatically become a member of the appropriate group and grant only the group access level for the user, regardless of the highest role the user actually possesses.

Ex: mwood, group = web_services_manager - has mgr access to web_services
mwood, group ccit_publisher - becomes mwood.ccit.publisher - has publisher access to ccit site but no access to web_services site

NOTE: This is just an example of how the user names MAY look. The actual protocol has not yet been determined.

Michelle will use existing Cascade web services api along with ldap, and write a php script to auto sync ldap entries for multiple account creation.


Metadata Sets


What do the Metadata start/end dates really do?

Hannon Hill Response:
Metadata start/end dates are used to determine when an asset is available for publishing and when it should be unpublished and archived. Once the asset is published, it become available to search engines. Once the asset is unpublished, it is no longer available to search engines and, once the search engine database is updated, the asset will no longer be found because it is no longer available on the web server.

Inline Metadata vs. Data Definition
Clemson's default asset factories were created using inline metadata instead of data definitions. When a new asset as created, based on the appropriate asset factory, the fields in the content wizard don't make sense (ie Author = Department Name) like they would if a Data Def was created (where we could make a field called Department Name and have it equal the department name. Is there an easy way to create a data definition that requests the same information as the metadata, make it more intuitive and apply it to existing assets without losing anything?

Hannon Hill Response

It is doable but would be very time consuming to go back and change the wizard to request a Department name block instead of an Author block. OWS would have to:
Create new region
Create current page block
Create XSLT to grab metadata
Apply both at template level

Once finished, OWS would have to go back through every asset that had been created, and update it to reflect the new Department Name block.

The Author block will now be explained as the block where the name of the department authoring the site is to be entered.

The difference between Metadata sets and Data Definitions was explained. If we are reusing content through other areas of the site - use metadata sets. Use Data Definitions for Main content areas.

Duplication of Config Sets, Asset Factories, Folders
How can we avoid having to duplicate the config sets, asset factories and folders for each site and subsite? Are we doing this as efficiently as possible?

Hannon Hill Response

The way Clemson University is using Cascade, with a need for unique footer, left nav and right nav  content, makes our method of duplication the most efficient way for now. It was suggested that OWS might want to look at having indexed blocks that will automatically search for unique block content and, if none is found, will move up the folder structure until an appropriate block is found. A new xslt will have to be created to accommodate this.

Block and stylesheet applied to a specific region.
Ex: Footer - instead of having a static block, have an indexed block that is pulling from indexed footer blocks

Migration to Cascade 5.0

Will our existing sites migrate seamlessly into Cascade 5.0 the way we currently have them set up?

Hannon Hill Response

Absolutely. PHP version should not be affected by upgrade.

Workflows

Another issue we run up against a good bit is workflow-related. The ccit_training/workflows folder has a workflow assigned to it. We have created a Full Page with Workflow page asset with the workflows folder set as the default parent folder so any time that asset is created, it initiates the workflow. This issue is related to folder vs. asset factory workflow, as well as the None, Do Not Inherit, Inherit from Parent options.


Hannon Hill Response

Upon reviewing the Full Page with Workflow asset and the asset factory, it was discovered that the workflow was assigned to the asset factory but the asset itself was set to have a  folder-controlled workflow. This was a result of changing the workflow to a folder-based workflow instead of an asset factory workflow. Upon testing, the issue seems to be resolved and we have a clearer understanding of how workflows function.


Video on BrightCove
Dave Sharpe has set up a Free Account  while Advancement continues to work up specifications/requirement that can be used to go out for bid. We will ask you to review before the bid process to be sure all requirements are included.