Version 126.96.36.199 is a collection of different enhancements (V.188.8.131.52 fixes a small bug only). The more significant new features appear in the Search Tab, some optimisations are also applied to the Document Details window.
It is now possible to search for Item IDs, using a comma-separated list. This is done through two dedicated new options in the main "(New Search\)Find Documents" drop down. They are: "with these internal IDs (comma separated)" and "with these imported IDs (comma separated)". Both will accept either a single value or a collection of values separated by a comma.
The first option will search against the Internal ID (the unique identifier used within EPPI-Reviewer), the second will search amongst the Imported IDs (usually the IDs that identified the records on the source database).
In the previous versions, it was possible to search for only one ID at the time, and this option was available as a secondary choice after selecting "containing this text" from the main "Find Documents" drop down.
Full text searches (the ones that use the "containing this text" option) will now record what fields were searched: Title/Abstract, Uploaded Documents, Year and so on…
Document Details Window:
When selecting the "auto advance" option, EPPI-Reviewer will now show a little reminder on the lower-left corner, after loading the next reference in the list. A small blue box will appear on the lower left, showing what code was added to the previous item. We believe this is useful to reassure users that they had indeed applied the code they intended. The reminder will disappear automatically after a few seconds.
When completing some coding (clicking on the "no entry" sign at the root of a "multiple user" code-set), the "completing" code has been optimised in different ways:
- Performance was optimised, so that the interface should reload the results in a much faster way. We expect this to be noticeable for users that work through a slow Internet connection.
- The same optimisation was also applies to the mechanisms that are activated when applying the first code to an item and when unselecting the only code in a set.
- The behaviour of warning messages has been improved to offer more information and consistency.
- The behaviour of the codes tree has been tweaked in subtle ways, we hope that this will make it slightly more user friendly.
Taken together, these small changes comprise also for a Bug Fix:
The "completed" status of a code-set (shown by the "no entry" or "green V" icons) was not updating properly when applying the first code to an item or when unselecting the only code in a set, this is now fixed.
- Also, to help avoiding confusion, the wording of the warning message that appears when unselecting a code has changed. It now is: "Are you sure you want to unselect this code?" (was "delete").
Other small enhancements:
- On the "Select Review" dialog that appears after logging on, the last logon time is now shown. This is useful for the users that have access to many reviews: it is now possible to order the list by the "last logon" column, so to place the most used reviews on top.
- Data exchange between the client and server has been improved at logon time. This may speed up the "loading data" phase a bit, especially over slow connections and/or when using slower computers.
- "My Info" tab: the list of reviews now includes the "last logon" date and time, as well as the name of the Review Owner. The review owner is generally the person that initially created the review. The "last logon" columns in both cases refer to the last access recorded for the current user.
- Duplicate checking was misbehaving if "getting new duplicates" lasted more than five minutes. This bug was introduced with version 184.108.40.206.
1) The "Sorry, the duplicate checking is still running…" window was showing up multiple times instead of only once.
2) When this happened, the final results of the "getting new duplicates" phase could be inconsistent.
Both problems are now solved and all inconsistencies have been sorted out.
- Also fixed: creating a new Outcome, under some rare circumstances, could generate a superfluous error.