Page 2 of 2 (29 items) < Previous 1 2
This post has 28 Replies | 6 Followers

Posts 8940
Forum MVP
Bruce Dunning | Forum Activity | Replied: Sun, Mar 11 2018 11:21 AM

Mark Barnes:
You can't merge without indexing, but you can index without merging (by re-creating the whole index, rather than creating a supplement index and merging it into the main one).

That totally makes sense. I just viewed them as the same thing because I don't think of it unless it causes problems with my PC but, as you point out, they are not.

Mark Barnes:
In other words, there appears to be no improvement. That suggests the improvements are only in the 'create the index' stage. If so there will be no benefit to Logos users other than when they download/install for the first time — which means we're all getting over excited.

It sure does sound like we're getting overly excited about something that apparently will not help unless we in the creating stage. Maybe the merging stage will be improved in the future.

Thanks for the explanation Mark.

Using adventure and community to challenge young people to continually say "yes" to God

Posts 1997
Forum MVP
Reuben Helmuth | Forum Activity | Replied: Sun, Mar 11 2018 1:13 PM

Phil Gons (Faithlife):
We're also exploring selective download and cloud resources

This is very welcome news and I hope it'll get implemented sooner rather than later!

Phil Gons (Faithlife):
We're considering some improvements to resource hiding, too

The implementation I mentioned here would (in my mind) solve both issues at once. Create a 'master library' or 'manage resources' page on our faithlife.com/logos.com account. Having the solution web-based would allow a user to keep from ever downloading certain resources (e.g. after a bundle purchase, simply log-in and filter/find the resources in question and toggle the switch!).

It would be nice if a resource (or selected resources) would have 3 possible values for each desktop installation: hidden/online/local.

Posts 611
Dave Thawley | Forum Activity | Replied: Sat, Mar 17 2018 9:02 AM

Phil Gons (Faithlife):

Version 7.14, which is in beta right now, includes a significant improvement to indexing. My library of ~60,000 resources went from indexing in 5h 21m to 2h 12m, a 59% reduction or 2.45x faster. Not all computers will see the same improvements, but the average increase may be around 2x.

We're also exploring selective download and cloud resources, which would let you choose what to install locally (and index) and what to leave in the cloud and have access to only when online.

We're considering some improvements to resource hiding, too.

wow. I can't wait for this 

Posts 8006
LogosEmployee
Bradley Grainger (Faithlife) | Forum Activity | Replied: Sat, Mar 17 2018 12:17 PM

Mark Barnes:

Looking at a new log (indexing 2 new resources in 7.14 beta 2), it took 4 seconds to create the supplemental index, but 16 minutes to merge them.

In other words, there appears to be no improvement. That suggests the improvements are only in the 'create the index' stage. If so there will be no benefit to Logos users other than when they download/install for the first time — which means we're all getting over excited.

Measure it again in 7.14 Beta 3. 

Posts 13359
Forum MVP
Mark Barnes | Forum Activity | Replied: Sat, Mar 17 2018 2:59 PM

Bradley Grainger (Faithlife):
Measure it again in 7.14 Beta 3. 

  • beta 1: 18m 20.2s
  • beta 2: 17m 5.0s
  • beta 2: 18m 43.8s
  • beta 2: 16m 25.5s
  • beta 3: 13m 59.7s
  • beta 3: 15m 13.2s
  • beta 3: 11m 46.9s

Averaging that out:

  • Betas 1 & 2: 17m 38.6s (1,059s)
  • Beta 3: 13m 39.9s (820s)

So an improvement in beta 3 of 22%. Not to be sniffed at. (The times quoted are those given at the end of the indexer log, for the total time to create the supplemental index, merge into the main index, and build the term/document info index. I omitted the beta 2 indexing log that reindexed all the reverse-interlinears, as that would be an unfair comparison.)

Posts 153
Randy Lane | Forum Activity | Replied: Sun, Mar 18 2018 4:49 AM

Phil,

Is the indexing identical for everyone, or is some of indexing based on feature usage?

For example do user-added tags get indexed, so a user heavily tags resources will see indexing take longer than someone who does little or no user-tagging with the same resources.

If this is true could FL list features that based on the user having impacts on indexing?

Posts 13359
Forum MVP
Mark Barnes | Forum Activity | Replied: Sun, Mar 18 2018 11:32 AM

Randy Lane:
Is the indexing identical for everyone, or is some of indexing based on feature usage?

The indexing we're talking about here is pretty much identical for everyone. It occurs when books are added to your library, or updated by Logos. The time it takes is dependent on hardware spec, library size, and the number of resources being updated.

Randy Lane:
For example do user-added tags get indexed, so a user heavily tags resources will see indexing take longer than someone who does little or no user-tagging with the same resources.

I'm not sure what you mean by user-added tags.

If you're talking about tagging books in your library, then doesn't affect the indexing we're talking about here. It does affect the time taken to go through "Updating Library Catalog". That time is dependent on how many resources you have, and how many tags, ratings and collections you have.

If you're talking about tags added to highlighting/notes, or labels added to parts of resources, these are not stored in the main index but in a personal documents index. For most users this index is small, and is updated fairly quickly.

Posts 205
Marshall Harrison | Forum Activity | Replied: Sun, Mar 18 2018 5:29 PM

I was happy with my 2804 books but after reading this thread I feel sort of inadequate.

Posts 2401
David Ames | Forum Activity | Replied: Sun, Mar 18 2018 5:42 PM

Marshall Harrison:

I was happy with my 2804 books but after reading this thread I feel sort of inadequate.

Some of us feel that we should have stopped there.

Page 2 of 2 (29 items) < Previous 1 2 | RSS