My history
Our Problem

1. My Problem
   
   1. My Problem
      
      1. NOAA wants more 'products' and 'services'
         
      2. different users want to use the data
         
      3. NOMADS: bare bones for the user who know what they want
         
      4. expert users download full data -"If I publishing, I will store a copy'
         
      5. As it is the data is only for the expert user is a narrow field
         
      6. we have requests to have the data transformed by policy people
         
      7. "what about the high school children?!"
         
      8. Mary Jo vs. Roberta: who do we search?
         
      9. Is there a prefect way to find a needle in a hay stack?
         
      10. we have expert users who won't use certain web tools due to complexity and bugs
          
2. No Perfect Answer: more than one path...
   
   1. Multi-path interfaces and multiple interfaces
      
      1. There is more than one way to enlightenment!!
         
      2. Multiple ways to find the same answer: like the town library, like Amazon.com
         
      3. However, different groups view the same data from different view points and different concerns
         
      4. My survey of 'expert' users found very different views of what 'useful' tools to use on the data
         
      5. Accidents in discovery are important
         
      6. Our expert users want quick and easy to use
         
      7. Our users have very different ideas about what is a useful 'thumbnail' of the data
         
      8. The 'perfect' tool/site seems to be a reflection of how each person thinks about there data!
         
      9. (it seems that the data is viewed as static, not as streams)
         
      10. No one system can be all things to all people
          
3. Parts is Parts
   
   1. Component based architecture
      
      1. But we can find common tasks
         
      2. dream: building blocks
         
      3. same tools: different 'skins' with different goals
         
      4. different chunks to handle very specific tasks
         
      5. Each site can string together groups of tools to do 'useful' things: developers win! by watching users create new uses
         
      6. Under the hood (bonnet)
         
      7. Common interfaces? data definitions?
         
      8. Can it be simple enough for a bored grad student to write a chunk?
         
      9. How would chunk to chunk work?
         
      10. translation interface frameworks?
          
      11. data caching for chunk to chunk hand offs?
          
      12. One core tool set - many different web sites
          
4. My Two Cents - the summary
   
   1. What I think regarding interfaces - ten commandments, no ten requests
      
      1. choose 'soft' meta-data carefully: it will limit, err, definite what the user can see or can stumble upon
         
      2. Have multiple paths to find the same information
         
      3. People trying to remember a previous finding will think different than the first-time users: respect this
         
      4. Have tutorials, examples, and the like
         
      5. Follow conventions of the web carefully: don't invent
         
      6. There should be multiple interfaces allow web folk to target smaller groups of users better: this will help developers later on as users find new uses
         
      7. Allow the users to succeed with only fragments of information
         
      8. Allow users to succeed with unclear goals
         
      9. Allow users to play, to make mistakes and recover easily
         
      10. Allow users to repeat complex tasks easily
          

Back to Top