Kobo Patcher

[PeterT]

So set all books to finished on initial transfer That way they will open at 1st page

[dmapr]

Originally Posted by davidfor

But, he also needs to consider the reading status of the book. This is stored in the database for the book.

- Unread - ReadStatus=0 - Open to first TOC entry
- Reading - ReadStatus=1 - Open using reading position stored in database
- Finished - ReadStatus=2 - Open to first page of the book.

That is why I suggested looking at the path for opening a finished book. As there is already an execution path that will do what is desired, finding a way to use that would solve the problem.

That is exactly what I passed on to Metazoa, that the Unread books should be treated as Finished, while behavior for Reading should remain the same.

[Nick_1964]

You can change the reading status with the kobo plugin for calibre, and you can repair the toc with Sigil automatically, so if the book doesn't work the way you want,you can adjust it, or is that not the way you want it ? (excuse me if this sounds crazy)

[byblo]

Originally Posted by Nick_1964

You can change the reading status with the kobo plugin for calibre, and you can repair the toc with Sigil automatically, so if the book doesn't work the way you want,you can adjust it, or is that not the way you want it ? (excuse me if this sounds crazy)

That is the 'very' expensive way to do it

Having to modify each epub...
Also, doing that will show all the epubs in the ebook marked as finished already, which is not desirable.

The trick with the patcher is to find a way to force the application to always start at page 1 (only when it is marked as 'not readed yet' or 'finished' in the DB)

[PeterT]

The big issue with ALL of these patches is that they are very VERY version dependent; as soon as a new release of F/W comes out, all the work that has been done is negated, and in fact as we've seen, some of the existing patches are not portable to newer versions of the F/W due to massive restructuring of the innards of the code.

[roger64]

I would also be interested to begin with the cover page. This is typically the kind of small things where the user could have been offered a choice and where this kind of patch can help.

Taking into account the latest release of Kobo Patcher 0.7, I consider that 2.8.1 is now for me a stable version. If Kobo really made a breakthrough, offering a new killer feature, of course I would reconsider this judgment, but chances that this kind of thing could happen seem dubious at best.

[ShellShock]

Originally Posted by PeterT

The big issue with ALL of these patches is that they are very VERY version dependent; as soon as a new release of F/W comes out, all the work that has been done is negated, and in fact as we've seen, some of the existing patches are not portable to newer versions of the F/W due to massive restructuring of the innards of the code.

Yes, but I am happy to accept never upgrading the firmware again. The patch has given me features much more valuable to me than any official firmware upgrade ever has.

[dmapr]

Originally Posted by roger64

I would also be interested to begin with the cover page. This is typically the knd of small things where the user could have been offered a choice and where this kind of a patch can help.

Taking into account the latest release of Kobo Patcher 0.7, I consider that 2.8.1 is now for me a stable version. If Kobo really made a breakthrough, offering a new killer feature, of course I would reconsider this judgment, but chances that this kind of thing could happen seem dubious at best.

Originally Posted by ShellShock

Yes, but I am happy to accept never upgrading the firmware again. The patch has given me features much more valuable to me than any official firmware upgrade ever has.

Same here.

[Nick_1964]

Originally Posted by byblo

That is the 'very' expensive way to do it

Having to modify each epub...
Also, doing that will show all the epubs in the ebook marked as finished already, which is not desirable.

The trick with the patcher is to find a way to force the application to always start at page 1 (only when it is marked as 'not readed yet' or 'finished' in the DB)

I just do it by a batch assignment, set all fonts to Amasis, with the Kobo Plugin you can set all books as not read (not finished) or read.
Put justify off, set top and left margin to 0,5 and put this to all the 2800 books in calibre et voila.
I use the Kobo Utilities Plugin for this.
Sure it would be nice if the book will do it by itself, but I also have a couple of Sony's so I have to do it in Calibre anyway.
The first time it takes a while for Calibre to finish the 2800 books, but now I just do it by every new book I add to Calibre, I have calibre running as service so I can remotely load the books that I want by wifi.
i am very happy with the patch the way it is now and won't leave the 2.80 firmware, but I think there will always be something that I want for the patcher, that can be easy done by calibre, because when Kobo adds something, a silent update,name it, and you can't fall back on the patch again,then at least my books are adapted the best way for the Kobo

[evanft]

Originally Posted by ShellShock

Yes, but I am happy to accept never upgrading the firmware again. The patch has given me features much more valuable to me than any official firmware upgrade ever has.

Agreed. I don't care about anything added with FW 2.10. I'll stick with what I have until (if?) the patch works with it.

[ShellShock]

Spoke too soon! In 2.10 you can apparently cross reference in dictionaries. Now that is one feature I would like to have.

[GeoffR]

I wrote a program so I could apply the patches in kpg.conf from Linux. It isn't finished but it works for the patches I use, so I probably won't get around to completing it. I've attached it in case anyone else has a use for it.

The main limitation is that the replace_utf8chars and replace_zlib methods are not implemented. This means it doesn't work with the 2.8.1 footer patch or the cyrillic keyboard patch, but it should work with the others including the 2.10.0 and later footer patches.

To use:

# Compile the program
gcc -Wall -o patch32lsb patch32lsb.c

# Create KoboRoot.tgz containing the patched libnickel
mkdir ./usr
mkdir ./usr/local
mkdir ./usr/local/Kobo
./patch32lsb -p kpg.conf -i libnickel.so.1.0.0 -o ./usr/local/Kobo/libnickel.so.1.0.0
tar -czf KoboRoot.tgz ./usr/local/Kobo/libnickel.so.1.0.0
rm -rf ./usr

# Transfer to Kobo device
cp KoboRoot.tgz /media/KOBOeReader/.kobo/

Edit: The program I originally posted (v0.1) would refuse to apply the `Brightness fine control` and `margins fine tuning` patches, complaining of "replace_int mismatched data". I've attached v0.2 which will apply those patches.

Edit: I've attached v0.4 of the program which implements the replace_string method as well as a new encode_xor8 setting. These can be used together to apply patches that use the replace_xor_XX methods, see comments in the source file.

Edit: Attached v0.5 which adds a -k switch for testing, and source code changes to allow it to be compiled for Windows with the mingw compiler. It should compile on the Mac or any other system with gcc and GNU libc, but only tested on Linux and Windows so far.

Edit: Attached v0.6 which understands the \x style escape sequences in strings. Adding \x00 to the end of the string as a terminator is not necessary with this program, but it won't do any harm if it is there.

[tshering]

Originally Posted by GeoffR

The main limitation is that the replace_string, replace_utf8chars, and replace_zlib methods are not implemented. This means it doesn't work with the footer or keyboard patches, but it should work with the others.

The original patcher is a python script with a windows wrapper around. So you can easily extract the script in order to run it on Linux. Of course the patcher is copyrighted, but for personal usage I think there is no problem.

[GeoffR]

Originally Posted by tshering

The original patcher is a python script with a windows wrapper around. So you can easily extract the script in order to run it on Linux. Of course the patcher is copyrighted, but for personal usage I think there is no problem.

Do you have a pointer to how to do this? I had a look in the kpg.exe executable but I coudn't see anything that resembled a script.

I have done a little C programming for Linux in the past, but I am pretty much clueless about Windows and Python.

[tshering]

GeoffR,

I sent you a PM.