Loading querystring urls sequentially

Commit message

Loading querystring urls sequentially

Fixes issue with (every) four urls being loaded potentially colliding with the previous load. Reported by Jeffery P via email. Thank you!

• All input/iframe element creations were started at the same time, but with two separate delays.
  1. 1000 milliseconds for creating an input textbox and an iframe.
  2. 250 milliseconds to put the url in the input textbox and update the iframe with the url from the textbox.
• Both delays were spaced out by multiplying by the index (+1) of the url to load.
• Depending the index of the url to load, there would be collisions between the element creation delay and the url updating delay.
  • The first elements were created after 1000 milliseconds.
  • The fourth url was updated after 1000 milliseconds.
  • The result would be a mixup during the dynamic element lookups, which assumed that the last input element was the one to enter the current url into.
  • The pattern would presumably repeat every multiple of four urls.
• Previously, the querystring url parser was tacked onto the existing code by mimicing/automating/triggering user interface interactions (button click, textbox change).
  • Parser logic changed to create elements and update elements (more) directly.
• Querystring parsing is now loading urls in sequence.
  • Performed using recursive calls with a (single) delay between calls.
• Computer system performance, internet connection speeds, and the contents of the loaded urls may have affected the chance of collisions.

Loading querystring urls sequentially

Fixes issue with (every) four urls being loaded potentially colliding with the previous load. Reported by Jeffery P via email. Thank you!

- All input/iframe element creations were started at the same time, but with two separate delays.
    1. 1000 milliseconds for creating an input textbox and an iframe.
    2. 250 milliseconds to put the url in the input textbox and update the iframe with the url from the textbox.
- Both delays were spaced out by multiplying by the index (+1) of the url to load.
- Depending the index of the url to load, there would be collisions between the element creation delay and the url updating delay.
  - The first elements were created after 1000 milliseconds.
  - The fourth url was updated after 1000 milliseconds.
  - The result would be a mixup during the dynamic element lookups, which assumed that the last input element was the one to enter the current url into.
  - The pattern would presumably repeat every multiple of four urls.
- Previously, the querystring url parser was tacked onto the existing code by mimicing/automating/triggering user interface interactions (button click, textbox change).
  - Parser logic changed to create elements and update elements (more) directly.
- Querystring parsing is now loading urls in sequence.
  - Performed using recursive calls with a (single) delay between calls.
- Computer system performance, internet connection speeds, and the contents of the loaded urls may have affected the chance of collisions.