Web Sites and Archiving

1
Web Sites and Archiving

• Presented by
• Erik Oliver

2
Overview

• Guiding Principles
• Lessons from Running a Large Site

3
Guiding Principle

• Design of site
• lowest common denominator
• Focus on the content not on having the newest
  must have format
• Plan Ahead
• Or just
• K.I.S.S.

4
Guiding Principles Site Design

• Avoid automated HTML tools
• Especially Microsoft Word
• Follow Core Standards
• HTML 4.01 or better yet XHTML 1.1
• CSS Level (Formatting and styling)
• Best viewed on any browser(test with at least 2
  browsers)
• Avoid Flashy Gimmicks
• Flash
• JavaScript
• Java

5
Guiding Principles Learn It

• W3C Web Site (Standards Documents)
• http//www.w3c.org
• Barebones HTML 4.0
• http//werbach.com/barebones/
• CSS Help (aka avoiding ltfontgt)
• http//www.htmlhelp.com/reference/css/
• Usability
• http//www.useit.com

6
Guiding Principles Check It

• HTML and XHTML Checking
• http//validator.w3c.org/
• CSS Checking
• http//jigsaw.w3c.org/css-validator/
• Links
• http//validator.w3c.org/checklink
• Browser Safe Colors (Chart)
• http//www.valdemar.net/erik/colors.html

7
Guiding Principles Focus

• Content is king
• Usability
• Readability
• Search capability (find specific stories)
• Access (browse different ways, etc.)
• Avoid
• Flashy graphics, blinking, moving, etc.
• New fads (must haves)

8
Tips and Lessons

• Case Study Master Apprentice Archive
• Started May 1999 0 stories
• Over 1000 in lt9 months
• Now gt2500

9
Phase I Initial Conception

• Ad hoc design of core pages
• Posting of first stories manually
• Key design decisions
• Free (no banner ads either)

10
Phase II Initial Popularity

• Recognition of limits of manual maintenance
• Need to maintain too many pages if you want
  multiple ways to view data
• Further success of archive/genre would decrease
  ability of archive to handle
• Key design decision
• Simple automated system needed

11
Phase III Database

• Selection of a simple 2 script Perl-based
  database system
• Free
• Flexible
• Easy to modify for non-programmer
• Key design decisions
• Selecting Perl as the glue

12
Phase III Database (cont.)

• Heavy initial setup time/cost
• You may need some programmer help, but mostly
  good knowledge of HTML
• Lots of new features become available
• Recent Category lists Author lists
• Watch for 80-20 rule violations
• Remember to KISS

13
Phase III Database (cont.)

• Results
• Managed fast(est) growing archive
• Delegation to archivists possible
• High Usability ? High Bandwidth consumption
• Key design decisions
• There is no way MA could be what it is without
  having used a database

14
Phase IV Bandwidth

• Popularity can be a problem
• Fund raising
• Voluntary
• Use PayPal or other online mechanisms
• Concerted campaign raises more than general
  donation/tip link
• Find a good hosting provider
• We use http//www.lightspeedservers.net/
• MA server has extra capacity too, contact Nicole

15
Phase V CD Project

• Decision to make a CD revealed weaknesses in
  initial ad hoc design
• Many files were not valid HTML
• Database-file system synchronization
• Archivists and users will do things you dont
  expect (multipart stories problem)
• Database driven vs. static pages
• CPU Speed Etc.

16
Phase VI Redesign

• All valid XHTML 1.1 and CSS Level 2
• CD can be created by changing one word in style
  sheet
• Now possible to change look of all files with a
  few keystrokes
• Better support for archivists
• Ties file upload with database record creation
• Better documentation and interface

17
Phase VI Redesign (cont.)

• Consistent file layout and naming
• /folder/ltfirst lettergt/filename.html
• File naming restrictions
• Extension
• Lowercase
• No special characters
• Length (lt31 characters)
• Expose existing, but hidden functionality (quick
  search)

18
Listen to users

• Actively encourage and respond to user requests

19
Conclusion

• Keep it simple
• Plan ahead for growth
• Check out
• http//www.valdemar.net/erik/archiving/