Kobo Patcher

[GeoffR]

Originally Posted by dmapr

If anyone is using the line spacing patch, does it work for you on the 2.10.0/3.0.0 firmware? For me as soon as I touch the slider it reverts back to Kobo default values. I ended up going back to 2.8.1 because of that.

I'm using it with 3.0.0 on my Glo and it works very well, no problems at all that I can see. I only use built-in fonts though, I havnt tested with sideloaded fonts. And the only other patches I have applied are the custom font sizes and custom margins, so I haven't tested if there is any conflict with other patches.

Unlike the 2.8.1 patch, the 3.0.0 patch seems to work with kepubs just as well as with epubs. With 2.8.1 I found the patch messed up the screen layout in kepubs if the spacing was too small.

[ShellShock]

Originally Posted by DNSB

I just checked with an Aura HD with FW 3.0.0 and the line height went from 0.9 to 1.35 when I moved the slider (checking the value stored in KoboeReader.conf after making changes). I didn't see any changes in the ebook since I set all the line heights from the stylesheet.

Regards,
David

Confirmed on my Aura HD.

[DNSB]

Originally Posted by dmapr

Thanks, does that change correspond to a single notch on the slider or did you move it all the way?

The change from 0.9 to 1.35 is the full right to full left movement of the slider. I stopped at 3 spots in between and the spacing shown in the KoboeReader.conf matched what I expected. I changed the default line spacings in the kpg.conf since I find all too often that dropping the line spacing below 0.9 allows ascenders and descenders to overlap which I find very annoying.

After I wrote the previous message, I checked with a .kepub.epub that doesn't have fixed line spacing. The line spacing moved wider/narrower with the slider and I did not see any unusual jumps. I didn't notice any difference in behaviour between using Charis SIL, Deja Vu Serif or Caecilia fonts.

patch_name = `My line spacing values` 
patch_enable = `yes`
# Bypass code-injected values - the wonderful know-how from Kobo programmers
replace_bytes = 817CEE, 57 F5 16 EE, 00 46 00 46
replace_bytes = 817D3C, 57 F5 EE ED, 00 46 00 46
replace_bytes = 817D62, 57 F5 DC ED, 00 46 00 46
replace_bytes = 817D98, 57 F5 C0 ED, 00 46 00 46
replace_bytes = 817DF4, 57 F5 92 ED, 00 46 00 46
# Values for replace
replace_float = 817E38, 1.05, 0.9
replace_float = 817E40, 1.07, 0.95
replace_float = 817E48, 1.1,  1.0
replace_float = 817E50, 1.35, 1.05
replace_float = 817E58, 1.7,  1.1
replace_float = 817E60, 1.8,  1.15
replace_float = 817E68, 2.2,  1.2
replace_float = 817E70, 2.4,  1.25
replace_float = 817E78, 2.6,  1.3
replace_float = 817E80, 2.8,  1.35
</Patch>

Regards,
David

[dmapr]

Thanks David. I wonder what the heck is wrong on my end.

[forster01]

delete

[dmapr]

Originally Posted by dmapr

Thanks David. I wonder what the heck is wrong on my end.

It was a combination of two factors: something has changed in the way nickel deals with fonts and the version of Charis SIL I was using. What I thought was Kobo default settings was actually nickel's interpretation of what the line height should be for that font. Swapping out for the Charis SIL Compact has fixed the problem.

[Quexos]

Noooo... it was only my imagination when I called Kobo evil ...

Also a long time ago I said somewhere in some post that it was strange that despite time passing, why Kobo wasn't offering a full screen option and some Kobo defenders replied that it took time and was something complicated to implement ...
Well it turned out that one simple guy (Metazoa) did it on his own, fast and well done.

Originally Posted by DNSB

Up to a short while back, I considered most of the comments about Kobo trying to block hackers from modifying their code as paranoia and/or conspiracy theory.

After all, zlib does offer a saving in space so that might be an excuse for the earlier changes. Now that Kobo has gone to using XOR to obfuscate the code, some of those conspiracy theories are beginning to sound rather likely. The only use for XOR is to hide information though not very well.

Regards,
David

[DNSB]

Originally Posted by Quexos

Noooo... it was only my imagination when I called Kobo evil ...

Originally Posted by Quexos

Well it turned out that one simple guy (Metazoa) did it on his own, fast and well done.

Try selecting text on the last 3 or so lines of the screen when using the reduced footer. A total pain in the patoot.

Regards,
David

[spruitje]

Additional patch to increase the dictionary font size in 3.0.0, using information from the Modifications for bad eyes (for 2.8.1) by tshering:

<Patch>
patch_name = `Dictionary font size`
patch_enable = `yes`
# Aura HD
replace_string = 115FB16, `%1px`, `39px`
# Glo, Aura
#replace_string = 115FB16, `%1px`, `31px`
# Touch
#replace_string = 115FB16, `%1px`, `24px`
# Mini
#replace_string = 115FB16, `%1px`, `28px`
</Patch>

[byblo]

@dmapr: any advance on the locked fonts and locked space line ?

[ShellShock]

Originally Posted by DNSB

The change from 0.9 to 1.35 is the full right to full left movement of the slider. I stopped at 3 spots in between and the spacing shown in the KoboeReader.conf matched what I expected. I changed the default line spacings in the kpg.conf since I find all too often that dropping the line spacing below 0.9 allows ascenders and descenders to overlap which I find very annoying.

There is something weird going on here though. This is what I am using

<Patch>
patch_name = `My line spacing values` 
patch_enable = `yes`
# Bypass code-injected values - the wonderful know-how from Kobo programmers
replace_bytes = 817CEE, 57 F5 16 EE, 00 46 00 46
replace_bytes = 817D3C, 57 F5 EE ED, 00 46 00 46
replace_bytes = 817D62, 57 F5 DC ED, 00 46 00 46
replace_bytes = 817D98, 57 F5 C0 ED, 00 46 00 46
replace_bytes = 817DF4, 57 F5 92 ED, 00 46 00 46
# Values for replace
replace_float = 817E38, 1.05, 0.8
replace_float = 817E40, 1.07, 0.9
replace_float = 817E48, 1.1,  1.0
replace_float = 817E50, 1.35, 1.1
replace_float = 817E58, 1.7,  1.2
replace_float = 817E60, 1.8,  1.3
replace_float = 817E68, 2.2,  1.4
replace_float = 817E70, 2.4,  1.5
replace_float = 817E78, 2.6,  1.75
replace_float = 817E80, 2.8,  2.0
</Patch>

When I set the slider to the fifth position, the Kobo eReader.conf setting is

readingLineHeight=1.2

which is what I would expect, but when actually reading a book, the line-spacing looks a lot more then 1.2em to me. This is the CSS in the book:

.para
{
    font-size: medium;
    font-weight: normal;
    /* line-height: 1.2em; */
    margin-top: 0;
    margin-bottom: 0;
    margin-left: 0;
    margin-right: 0;
    text-align: left;
    text-indent: 1.5em;
}

i.e., I am using the Kobo eReader.conf setting. This is with the built-in Georgia font, with some Advanced tweaks to make it darker.

If I change the style to

.para
{
    font-size: medium;
    font-weight: normal;
    line-height: 1.2em;
    margin-top: 0;
    margin-bottom: 0;
    margin-left: 0;
    margin-right: 0;
    text-align: left;
    text-indent: 1.5em;
}

(spacing set from the book) I get much closer line-spacing. So the CSS seems to give me the correct spacing, but if I rely just on readingLineHeight, the line-spacing is out of whack.

A thought: I had always assumed that the readingLineHeight in Kobo eReader.conf was in units of ems? Is this not the case? (But ems is the only unit that makes sense).

[GeoffR]

Originally Posted by ShellShock

A thought: I had always assumed that the readingLineHeight in Kobo eReader.conf was in units of ems? Is this not the case? (But ems is the only unit that makes sense).

Yes the readingLineHeight units changed in recent firmware. I think readingLineHeight=1 in the new firmware corresponds roughly with readingLineHeight=1.3 in the old firmware, so the units are not in em any more (if they ever were). To get a css line-height of 1.2 you would need readingLineHeight of less than 1, perhaps about 0.92.

[ShellShock]

Thanks Geoff, it sure is screwy.

[dmapr]

Originally Posted by byblo

@dmapr: any advance on the locked fonts and locked space line ?

Not that I know of. Metazoa just mentioned that he hasn't done it yet, but didn't say if he had a chance to even start looking into it.

[JonnyXDA]

FYI Metazoa has updated the patch for 3.0.0 firmware support.

http://rghost.net/50190463