Indexing - What if...
From the recent new posts from the newbies concerning the issue of indexing, I would like to bring to a discussion whether it is possible to eliminate the whole ordeal of indexing. Instead, I wonder if it is possible to have resources that are pre-indexed before they are downloaded into our computers. Thus, cutting the time on our end, eliminating any worry or question, and having results in no time.
Comments
-
Unfortunately not as each (total of all books) index is, essentially, unique to each person's unique library. Also because of the nature of Logos (cross referencing multiple books) it would not be practical to have multiple indexes for each book, then you would need an index of indexes — back to square one.
The indexing is a 'necessary evil' which, if left alone (and for notebooks with mains power always attached) to finish cleanly is usually not problematic. It is also a once off task. Granted it is a bit of a pain, but it's unavoidable.
"I want to know all God's thoughts; the rest are just details." - Albert Einstein
0 -
The last version, Libronix, came with an index in each book file. Placing the indexing process on the local machine allows for the fast searching, returning to a per-resource, prebuilt index would probably slow us back down to Libronix speeds.
Prov. 15:23
0 -
Robin Lee said:
From the recent new posts from the newbies concerning the issue of indexing, I would like to bring to a discussion whether it is possible to eliminate the whole ordeal of indexing. Instead, I wonder if it is possible to have resources that are pre-indexed before they are downloaded into our computers. Thus, cutting the time on our end, eliminating any worry or question, and having results in no time.
You will swap indexing time for downloading time (not to mention limits!) even if you have a stock box Library. But let's dream a little ie. you have Scholar's + 100 other resources. You download and install the Scholar's resources + c. 2 GB of indexes and can be searching after a few minutes of discovery. Then L4 will download and index the other resources. Hopefully, they will index by themselves and there will be no rebuild!
The design changes to accomplish this are not trivial eg. suitably separating the two sets of resource downloads.
But it's nice to dream in a New Year!
Dave
===Windows 11 & Android 13
0 -
Dave Hooton said:
You download and install the Scholar's resources + c. 2 GB of indexes and can be searching after a few minutes of discovery.
Hi Dave — the only problem I'm seeing with doing that is that the index file will (most likely) have the full path to the resource, and that will also include the user unique code folder and the user's user account folder name.
Hard drive >> Users >> [Your user account] >> Library >> Application Support >> Logos4 >> Data >> [a unique code] >> LibraryIndex
so Logos cannot ship a generic set of indexes. I agree that indexing is a little tedious, but it's a necessary task given that we are looking at a program searching hundreds, or thousands, of books sometimes in complex query permutations. Touch wood, I have never had any issues with indexing, for the initial run I leave machine running all night and that's it. I am wondering how many people running MacBooks forget to make sure that there is mains power connected to the machine and power savings is temp. turned off. I would not be surprised that many (corrupted) indexing issues come down to this.
"I want to know all God's thoughts; the rest are just details." - Albert Einstein
0 -
Concur indexing tedious - currently not have viable technical alternative to offer for creating Bible Index and Catalog Index for all Bibles and Resources in a Logos Library (different for users) - plan to learn more about database design and implementation this year (ongoing resolution).
Thankful Logos 4 Search panel shows "Some search results may be missing or incorrect while your library is being indexed." along with Home Page message: "First-tIme indexing underway" - wonder what search(es) work while indexes being created ?
Keep Smiling [:)]
0 -
wonder what search(es) work while indexes being created ?
Probably those involving resources that have not yet been indexed.
0 -
What if Logos was to build "intermediate" indexes on the local machine (once) and then merge those indexes into a single master index. That way, when new resources are added or a few are updated the "intermediate" (previously computed) indexes don't have to be rebuilt for all resources in the entire library - only for new and updated resources - and then those can be "merged" into a master index...
Or is that the way Logos already works today?
0 -
Stuart Robertson said:
What if Logos was to build "intermediate" indexes on the local machine (once) and then merge those indexes into a single master index. That way, when new resources are added or a few are updated the "intermediate" (previously computed) indexes don't have to be rebuilt for all resources in the entire library - only for new and updated resources - and then those can be "merged" into a master index...
Or is that the way Logos already works today?
Essentially.
When you add new resources (new purchases or updated files) Logos builds a supplemental index. When that supplemental index reaches a certain size (approx 25% I believe) of the Main index the supplemental index is merged into the main index. After a certain number of merges (I don't remember the number) a complete rebuild is done in place a merge.
Prov. 15:23
0 -
My Libary index folder is >4GB, even if it was possible to have a standard index that you donload, do you you really want to spent maybe an aditional day downloading it comparred to the 8-12hrs it taakes for indexing, for me local indexing wins each time...
Never Deprive Anyone of Hope.. It Might Be ALL They Have
0 -
Prov. 15:23
0 -
Patrick S. said:
Hi Dave — the only problem I'm seeing with doing that is that the index file will (most likely) have the full path to the resource, and that will also include the user unique code folder and the user's user account folder name.
The [unique code] is a random User Id and Logos gets it from data in the Logos4\Users folder when you Sign-in ie. the destination for the Indexes is independent of this Id.
Dave
===Windows 11 & Android 13
0
said:
