1.0.3 Feedback: A Running List

Version 7

    This is a list of all bugs, issues, feature change requests, or documentation deficiencies reported in v1.0.3 of VizAlerts, for tracking purposes. I'll keep this up to date with newly reported issues, and add notes should they be fixed in a newer version. Alongside each is who found / suggested each issue / feature request:


    Items struck through have been fixed in v1.0.4. All other items will carry over.


    Bugs:

    1. Data containing non-Latin characters is not yet supported (matt c)
    2. Duplicate emails are sent in Advanced Alerts when exported data contains duplicate rows (despite the viz showing only one). (toby e)
    3. HTTP errors during data exports are not reported in failure emails. Also need to translate HTTP status codes to intelligible, actionable information (e.g., 403 = subscriber doesn't have permission, 406 = viz render failed; click this link to see error) (matt c)


    Functionality Changes:

    1. Do away with need to add a comment to Advanced Alerts (toby e)
    2. Add human-friendly view, subscriber, workbook, etc names to state file (toby e)
    3. Improve efficiency by forcing refresh for each unique view only the first time it's checked (matt c)
      1. Note--could not implement this and guarantee alert accuracy, so built the plumbing instead.
    4. Improve efficiency by rendering png for Advanced Alerts only when VIZ_IMAGE() is used (matt c)
    5. Support plaintext email in Advanced Alerts (toby e)
    6. In a single email for an Advanced Alert fails, do not stop processing remaining email (toby e)
    7. Provide an intuitive way for end users to trigger a given VizAlerts manually for testing (toby e)
    8. Support additional formats for embedded content in Advanced Alerts (csv, pdf, png, etc) (toby e)

    Documentation Changes:

    1. Add an "advanced setup" section to the install guide. (toby e)
    2. Explicitly state email count expected from Advanced Alerts example (Documentation suggestion.) (toby e)
    3. Expand the images in the documentation so they are more readable (toby e)