[resolved] 6.3 Beta 1: What the Indexer should not do

Dave Hooton

It shouldn't be called for a fix of the "Bible" Index when the resource downloaded was a non-indexing dataset.

1512.Indexer6.3.zip

Two or three seconds may appear inconsequential, but this legacy usage will continue to confuse/upset users about indexing.

Andrew Batishko

Dave, in addition to the 3 main indexes, the indexer also updates the auto-completer index. This line from your log:
2015-04-16 19:27:12.2551 6 Info IndexerProgram (1722ms) Updating Auto Complete from Resource Manager.

The auto-completer index can collect values from "non-indexing" datasets. It would not be correct behavior at this time to skip running the indexer.

Can you help me understand what is confusing and upsetting to users about running the indexer for three seconds? Historically we have found that many users didn't even realize the indexer was running despite notification in the task bar and the application.

Dave Hooton

https://community.logos.com/discussion/comment/717277#Comment_717277

1. Any time it runs it is perceived as indexing resources (books) e.g. "indexing all the time"

2. If it runs for too short a time something is wrong e.g. "indexing should take longer than this"!

3. It is inefficient (process start/completion has to be monitored, it has to create various 'managers', it sets a tool tip that Software is Indexing, sets Status as Indexing library, sets a timer for Indexing)

4. Indexer was run "because the Bible index needs to be updated"!

EDIT:  3. is my reason for having the new Update process use the Indexer only for resource indexing

          4. seems to be the reason given whenever the Indexer is called!