Explore whether research/review can happen parallel to research

[timgdavies]

The current locking model prevents reviewers from working on a sheet at the same time as researchers.

However, if a reviewer is only leaving notes, is there a strong reason to prevent them from writing to the notes sheet whilst a research is also working on the results sheet?

Allowing some parallel working could allow a reviewer to start giving feedback on questions marked as complete (see #16 ) before the rest of the research is finished.

[timgdavies]

One option to consider is per-page locking. Currently, locking happens through a lock-string written whenever updates are made and the ability to 'take over' a whole survey for editing.

Locking is important because we do not want the case where:

• Researcher A loads the survey
• Researcher B loads the survey
• Researcher A edits Question 1
• Researcher B edits Question 2, over-writing researcher A's changes

If we moved to checking the lock on a per-page basis, then instead of the lock being checked when loading the survey, it would be checked when opening a given page.

This would then let two researchers work on different pages of the survey at the same time.

If they open a locked page, they would be prompted to access read-only, or to take over the page.

Implementing this may benefit from having the ability to load or refresh data for a page without a full reload of the tool.

This may also help with increasing the speed of the application, if we only load answers 'on demand' rather than loading all answer data when first opening the sheet.