JBPatch

[ixtab]

@knc1: Maybe. Maybe they took an actual device screenshot and "built it" into a picture of the device. Anyway, I'm still positive that this is a real screenshot. Why? just compare the cover view on the image to the ones that the coverview patch produces.

[sandman7920]

I have interest to test this patch method on K4NT.
If i copy jbpatch.jar to /opt/amazon/ebook/lib and add -istart jbpatch.jar in init.xarg is there a risk for brick.
I accept any advices.

P.S.
Sorry for my English.

[ixtab]

Originally Posted by sandman7920

I have interest to test this patch method on K4NT.
If i copy jbpatch.jar to /opt/amazon/ebook/lib and add -istart jbpatch.jar in init.xarg is there a risk for brick.

Most definitely: yes!

If you are trying to port this to a different device, then yes, you should be prepared for the entire thing horribly failing, having to debrick the device, having to carefully monitor log files etc.

However, if you do know how to debrick the device, then all it needs (at least initially) is a little patience and a bit of fiddling around to find out whether the same method as on the K5 works for bootstrapping, or which other method can be used.

[kindleloh]

Hi ixtab,

thank you for the update. I have checked it but it has no effect regarding the bug. If I swipe to the next chapter then it works, the progress bar states 0%. If I then go a page back and for I remain on the first page of the new chapter, but the progress bar states now 100%. May be the solution is to set the progress bar to the next chapter with 0% even when the current position is 5 Positions before the next chapter. May be the chapter index is not as accurately as aspected.

[ixtab]

Originally Posted by kindleloh

Hi ixtab,

thank you for the update. I have checked it but it has no effect regarding the bug. If I swipe to the next chapter then it works, the progress bar states 0%. If I then go a page back and for I remain on the first page of the new chapter, but the progress bar states now 100%. May be the solution is to set the progress bar to the next chapter with 0% even when the current position is 5 Positions before the next chapter. May be the chapter index is not as accurately as aspected.

Hmm... can't reproduce (and just to make sure, please double-check that the patch version is indeed 2012-09-05.) I have tried with several books now. For instance, "Krieg und Frieden" contains chapters which start on page boundaries, and it works exactly as expected. Can you give an example of a free book from Amazon where it behaves erratically?

[kindleloh]

Hi ixtab,

you are right, I have checked with different books and found the described behavior only on books with a TOC created with means of calibre. So I will check with the way I insert TOCs into books.

[ixtab]

Originally Posted by kindleloh

Hi ixtab,

you are right, I have checked with different books and found the described behavior only on books with a TOC created with means of calibre. So I will check with the way I insert TOCs into books.

Even so... if you can, please let me try to reproduce it. If you have a free book where this is happening, please attach it here and give exact steps to reproduce (preferred). Otherwise, you can also send me a PM.

[kindleloh]

Hi ixtab,

I have append here a simple ebook from Wikipedia. Book created in odt format and then converted to mobi via calibre.
If you open the book and goto a chapter (e.g. Prelude) then you see a 0% bar. Go one page back (you get about 40%) and go one page forwards so you are at the begin of the chapter Prelude. Now the progress bar is not set to 0% instead it is set to 100%. The same occurs with when you normally read a book and go only forward. The first page in a chapter is always set to 100%.

Thank you very much for your investigation.

[ixtab]

Originally Posted by kindleloh

Hi ixtab,

I have append here a simple ebook from Wikipedia. Book created in odt format and then converted to mobi via calibre.
If you open the book and goto a chapter (e.g. Prelude) then you see a 0% bar. Go one page back (you get about 40%) and go one page forwards so you are at the begin of the chapter Prelude. Now the progress bar is not set to 0% instead it is set to 100%. The same occurs with when you normally read a book and go only forward. The first page in a chapter is always set to 100%.

Thank you very much for your investigation.

Ok, so I tried it with the book that you attached and a "debug" version of the patch. Here is the (commented) output:

# ctx means "context", i.e. the positions which are used for determining the
# start and end positions for the chapter; they are taken straight from the
# book's TOC; the numbers are in the internal format of the TOC.
# Unless ctxEnd == bookEnd, the following invariant always holds:
# ctxStart <= position < ctxEnd
# (if and only if position == bookEnd, the "<" becomes a "<=")

# 0. Open book, goto beginning, turn on chapter progress bar:
position: 0
ctx start: 0
ctx end: 2924
# 1. Goto "Prelude" chapter
position: 8282
ctx start: 8254
ctx end: 17950
# 2. Page back
position: 5968
ctx start: 2924
ctx end: 8254
# 3. Page forward
position: 8238
ctx start: 2924
ctx end: 8254
# 4. Goto "Prelude" chapter again
position: 8282
ctx start: 8254
ctx end: 17950
# 5. Page forward
position: 10225
ctx start: 8254
ctx end: 17950
# 6. Page back
position: 8282
ctx start: 8254
ctx end: 17950

So the code of the patch works absolutely correctly, but something else obviously doesn't. Note how the current position and the context start are not the same in step 2. Also note that if you perform the abovementioned steps, step 3. actually puts the "Prelude" heading on the second line, while step 4. will put it back on the first line. Finally, paging back and forth will not always lead to the exact same position as where you started from (not even with all Amazon books: "Huckleberry Finns Abenteuer und Fahrten" has occasionally - but interestingly, not always - exposed the same behavior).

This is either a bug in the reader software, or a bug in the book's format, or a combination of both. In any case, it's not a bug in the progress bar patch, and I can't do anything about it. A position before the start of a chapter is a position in the previous chapter, period. In other words: we are going to have to live with it.

[eureka]

BTW, Time To Read feature advertised for Kindle Paperwhite is already implemented (or, at least, partially implemented) in current Kindle Touch firmware. Though it's called Reading Timer. It's implemented as reader plugin, located at /opt/amazon/ebook/lib/ReadingTimer.jar and it's entry is presented (but commented) in /opt/amazon/ebook/config/reader_plugins.xml.

Just thought it's worth mentioning here just as hidden Cover View mode.

(Speaking cynically, with all these implemented but disabled features Kindle Touch is looked like hardware platform for beta-testing of Kindle Paperwhite firmware).

[veezh]

Originally Posted by eureka

(Speaking cynically, with all these implemented but disabled features Kindle Touch is looked like hardware platform for beta-testing of Kindle Paperwhite firmware).

Well, let's hope so... Maybe that means it won't be that difficult to implement JB Patch for the PW.

I'm looking forward to the PW but *not* to going back to pre-JB Patch wonky 'justification' and lack of hyphenation (screenshot from PW commercial attached).

[dukesilver]

Originally Posted by eureka

BTW, Time To Read feature advertised for Kindle Paperwhite is already implemented (or, at least, partially implemented) in current Kindle Touch firmware. Though it's called Reading Timer. It's implemented as reader plugin, located at /opt/amazon/ebook/lib/ReadingTimer.jar and it's entry is presented (but commented) in /opt/amazon/ebook/config/reader_plugins.xml.

Just thought it's worth mentioning here just as hidden Cover View mode.

(Speaking cynically, with all these implemented but disabled features Kindle Touch is looked like hardware platform for beta-testing of Kindle Paperwhite firmware).

oh nice! so how tough might it be to enable that feature? I'm not a dev/hacker, so I have no clue.

[elbunuelo]

Hi Ixtab. I've been trying to install JBpatcher but I'm stuck at the installation screen showing "invalid init.xargs" and then error 006.

I've read that this is due to a tampered init.xargs file. I honestly don't remember editing that file in any way, but I may have done it.

Here's the content of my init.xargs file


-Dch.ethz.iks.concierge.basedir=file:/opt/amazon/ebook/lib
-Dch.ethz.iks.concierge.jars=file:/opt/amazon/ebook/lib
-Dch.ethz.iks.concierge.log.enabled=true
-Dch.ethz.iks.concierge.log.level=4
-Dch.ethz.iks.concierge.debug.classloading=true
-Dch.ethz.iks.concierge.storage=/var/local/java
-Dch.ethz.iks.concierge.strictStartup=true
-Dcom.amazon.kindle.util.log.logLevel=info

-istart portability-impl.jar
-istart kaf.jar
-istart KindletSystemPlugins-2.0.jar
-istart AudibleHandler.jar
-istart ACXToDoHandler.jar

Thanks for your help

[ixtab]

Originally Posted by elbunuelo

Hi Ixtab. I've been trying to install JBpatcher but I'm stuck at the installation screen showing "invalid init.xargs" and then error 006.

I've read that this is due to a tampered init.xargs file. I honestly don't remember editing that file in any way, but I may have done it.

Here's the content of my init.xargs file


-Dch.ethz.iks.concierge.basedir=file:/opt/amazon/ebook/lib
-Dch.ethz.iks.concierge.jars=file:/opt/amazon/ebook/lib
-Dch.ethz.iks.concierge.log.enabled=true
-Dch.ethz.iks.concierge.log.level=4
-Dch.ethz.iks.concierge.debug.classloading=true
-Dch.ethz.iks.concierge.storage=/var/local/java
-Dch.ethz.iks.concierge.strictStartup=true
-Dcom.amazon.kindle.util.log.logLevel=info

-istart portability-impl.jar
-istart kaf.jar
-istart KindletSystemPlugins-2.0.jar
-istart AudibleHandler.jar
-istart ACXToDoHandler.jar

Thanks for your help

Are you sure that you have Firmware 5.1.x installed? This looks like a 5.0.x one. If so, you have to upgrade your firmware first, because JBPatch doesn't do anything useful on FW 5.0.x anyway.

[elbunuelo]

That should have been the first thing to check out. I'm sorry, after the update it worked perfectly.

Thanks