where is indexing stored

[statethatiamin]

Originally Posted by niceboy

"So I gave up indexing and marked "read only" flag on index.db"

I am not a very technical person. Can you tell me a simple way to un-enable indexing? I plan to load many books on my Kindle 3 but I do not want them indexed.

Sure, I found a way! All you have to do is change a flag in the file start.sh...
https://www.mobileread.com/forums/sho....php?p=1341280

[DarkScribe]

Does anyone know what sort order is used by the Kindle indexing process? When looking at "Not Yet Indexed" list it is sorted by Title, but that does not appear to be the actual method. I am trying to find an easy way to discover which files are choking the sort process when indexing.

[statethatiamin]

Sorry, I haven't a clue as to why the Kindle chooses to index books preferentially out of order

[vitalidon]

Originally Posted by OakIris

I don't believe there is anyway to disable indexing on the Kindle, but if there is, someone here is sure to chime in.

Holly

What about deleting index.db file and create instead a folder with exactly the same name Index.db (case sensitive, as I think Linux treats caps and lower case differently) and also just in case mark it as read-only, hidden and also set the Archive attribute. If kindle will not be able to delete the folder, then maybe it will not be able to recreate another index.db file, and therefore it will not be able to start the index process.

[eschwartz]

As NiLuJe said in post #7, there are search commands that can disable indexing. See: Kindle_Touch_Hacking#Search_Bar_Shortcuts
Many of these commands work across several Kindle models, including the ;di (indexerDisable) command.

[susan_cassidy]

The Kindle processes will run as root, therefore, they can overwrite any file they want. You can't protect root from itself.

[eschwartz]

Originally Posted by susan_cassidy

The Kindle processes will run as root, therefore, they can overwrite any file they want. You can't protect root from itself.

Have you ever tried saving a file on a linux machine in the same location where there is already a folder? Clearly not...
You'd have to know the folder might exist and make sure to clean up first.

But hey, try this out and see if you get a different result:

[eschwartz@arch /tmp]$ mkdir somedir
[eschwartz@arch /tmp]$ sudo bash -c "echo blah > somedir"
bash: somedir: Is a directory
[eschwartz@arch /tmp]$ sudo wget -O somedir www.google.com
somedir: Is a directory

I assure you, not only is it a very interesting PEBKAC-stlye error, it is also one currently used here to great effect in blocking OTA updates.


tl;dr
no one ever suggested the goal was stopping the root user from overwriting a file.

[Andy_D]

Originally Posted by eschwartz

Have you ever tried saving a file on a linux machine in the same location where there is already a folder? Clearly not...
no one ever suggested the goal was stopping the root user from overwriting a file.

Vitalidon did in a post above when trying to disable indexing:

"What about deleting index.db file and create instead a folder with exactly the same name Index.db [...] If kindle will not be able to delete the folder, then maybe it will not be able to recreate another index.db file, and therefore it will not be able to start the index process."

This doesn't work.

I tried something similar when I first got my Kindle with "special offers." I noticed that if the folder, .assets, in the system folder were deleted, no ads would be displayed, however within a day if one went online for even a minute the .assets folder would be re-created and ads would reappear as the folder filled with image files. Then I tried making a .assets file in the system folder on the same theory as the one espoused by the poster who wanted to disable indexing. However, no joy. The .assets file got deleted to make way for the .assets folder.

This is similar to Windows behavior. Using this method, I have successfuly stopped the creation of unwanted files created by various third party programs, but if one tries to prevent the creation of a system file, Windows has no trouble creating that file. Apparently Linux works the same way.

[eschwartz]

Is this the right time to go on a lecture about the differences between:

• the linux commands `rm` vs. `rmdir`
• and the ubiquitous Windows command `del`?

The assets folder presumably got deleted in recursive mode, which is natural enough and catches a file too. Usually people programming with a single file in mind do not bother to use recursive mode, hence the verifiably working update.bin.tmp.partial/ update trick which you can read about in our Wiki: Kindle_Touch_Hacking#Blocking_Official_Updates
And hence why I specifically referenced the manipulation of a "PEBKAC" error of Amazon's. Which may or may not be present in their indexer handling as well.


I have no idea why or how Windows does system files, but it is irrelevant since the name "system" on the Kindle, or indeed any linux OS, means nothing. The developers could have decided to name it "donkeys", and every last aspect of the Kindle would work identically in every respect. (Not entirely true -- the folder would no longer be invisible when plugged into a Windows computer.)

Stop speculating and start testing.