Hi Bob
I started a new thread as there was a lot of information on this issue and I personally finding it hard to follow. Having watched and read both sides of the debate I feel that I now understand why you need to take a new approach in version 4 to manage the integrity of the Logos Application and the Data (our books) that it processes.
What I would like to understand is looking beyond 4.0 what options exist to make elements of this more user friendly to ensure that the update when delivered creates a fully working system with minimal disruption.
I appreciate that the Beta process is generating a lot of misleading impressions and that many of the processes will be very different when the product is finally released but based on the beta implementation of the update process and your comments about needing the application and the data to be synchronised to have a fully working system I have definite concerns about the current approach which at the moment I understand to be:
Step 1 - download and apply the program update
Step 2 - download and install any data updates
Step 3 - full or partial re-index of the data
My concern is that step 1 is typically a quick process as the download is measured in MB rather than GB but when complete if my understanding of what has been written elsewhere is correct it leaves me in an imperfect state until step 3 completes. I can use Logos 4 but there may be problems with the results. In this model I have an 'imperfect' system for the duration of steps 2 and 3.
My suggestion is that the order be changed, I appreciate step 3 by definition must be last but given the level of central control you will be exercising in future it seems to me that the order of steps 1 and 2 could be changed and the revised process be:
Step 1 - background download of data to a staging area
Step 2 - download and install of the program updates and install staged data updates
Step 3 - full or partial re-index
I believe that this approach would improve the overall integrity of Logos during the update process and also prevent problems that may arise when a user receives the program update but has a network problem that delays the final delivery of the data. This does not limit the control that you exercise over the process but would I believe improve the user experience.