Scrambling copyright ebooks to help troubleshoot problems ??? - Page 7

jackie_w · 10-18-2015, 03:07 PM

Quote:

Originally Posted by RbnJrg

Watch the ebook I attach. That book was scrambled with your original utility; I don't know why, the .xhtml index resulted with the numbers unchanged

The original book is in public domain (is "The mistery of the yellow room") so if you need it to find the cause of the unscrambled numbers in TOC, I can upload here.

You mean on the 'Exblgknua' page? You're absolutely right. I'll investigate.

ETA: The reason is because that page doesn't actually contain any digit text.

The page structure (simplified) is:

Code:

<body>
<h3>...</h3>
<ol>
     <li>first item ...</li>
     <li>second item ...</li>
      ... ... etc
</ol>
</body>

The numbers are auto-generated by the ordered list <ol>/<li> HTML. I'm not sure what I can do about that ... nor how much it will upset HarryT.

I've added it to a "things to think about list" (no pun intended)

eschwartz · 10-18-2015, 03:45 PM

That's structural, and anyway I think HarryT's issue was that some numbers can identify the book (ISBN).
Not sure why knowing the book title matters, but...
that isn't the case here.

HarryT · 10-18-2015, 04:55 PM

Quote:

Originally Posted by eschwartz

That's structural, and anyway I think HarryT's issue was that some numbers can identify the book (ISBN).
Not sure why knowing the book title matters, but...
that isn't the case here.

I was under the impression that the goal was to produce a book that's structurally similar to the original, but from which the original couldn't be identified.

Of course, if that's not the desire, then there's no issue.

DaleDe · 10-18-2015, 08:24 PM

Quote:

Originally Posted by PeterT

I'm of the opinion that anything within a <.....> should be left unchanged.

Certainly, if the goal of this whole exercise is to produce a product that can be used to find formatting bugs or other page problems then tags are essential to troubleshooting potential problems, even (perhaps especially) if they are bad tags.

Dale

eschwartz · 10-18-2015, 11:11 PM

Quote:

Originally Posted by HarryT

I was under the impression that the goal was to produce a book that's structurally similar to the original, but from which the original couldn't be identified.

Of course, if that's not the desire, then there's no issue.

Okay then.

I thought the goal was to produce a book that's structurally similar to the original, but doesn't run afoul of copyright law when uploaded to MobileRead for debugging purposes.

I can imagine some people might also want to prevent the book being identified at all, but I'd venture to say most people are satisfied even if the bare minimum to stave off copyright concerns is done.

jackie_w · 10-19-2015, 01:47 PM

Quote:

Originally Posted by HarryT

I was under the impression that the goal was to produce a book that's structurally similar to the original, but from which the original couldn't be identified.

Quote:

Originally Posted by eschwartz

I thought the goal was to produce a book that's structurally similar to the original, but doesn't run afoul of copyright law when uploaded to MobileRead for debugging purposes.

There's a fairly straightforward decision to be made, I think. Troubleshooting an ebook problem will almost certainly require the book to be exploded into its constituent parts for further investigation.

Which of the following are true from MR's POV?

There must be no visible sign of Author, Title, ISBN (and maybe other metadata yet to be identified):

when reading the book on a reading device or with a more generic book viewer, e.g. calibre Viewer.
when the troubleshooter views the exploded ebook contents, using their troubleshooting tool of choice.
Is Publisher's name also a 'must-not-see' item?

Quote:

The moderating team has discussed the question you raised, and we're happy to give you the go-ahead for this suggested work, provided that the following conditions are met:

- All text, table of contents, headings, textual metadata, etc, must be replaced with scrambled text.
- All images must be replaced with scrambled images or removed.

Provided that your suggested plug-in produces output which satisfies these conditions, there would be no problem in uploading such output to MobileRead's forums.

Based on HarryT's post #63 (part re-quoted above) I'm pretty sure item 1 is true.

What about item 2? This is a more challenging project (not quite ready to give up yet, though

)

I've seen many books where the ISBN is hard-coded into the filename of nearly every constituent file in the book (HTML, CSS, NCX, images), including the OPF file itself.

I think the calibre utilities should be able to handle the issues detailed below (need to make sure, though).

If the HTML filenames contain the ISBN then any internal link anchor reference will also contain the ISBN e.g.
Code:
```
<a href="name_with_ISBN.html" id="id1">
```
If the image filenames contain the ISBN then the HTML image alt attributes may also contain the ISBN, because some books derive the alt from the filename (who knows why?) e.g.
Code:
```
<img alt="name_with_ISBN" src="name_with_ISBN.jpg">
```
There's also the HTML title attribute to consider for possible revealing info, e.g.
Code:
```
<h2 title="BookTitle AuthorName ISBN, Chap 1">Chapter 1</h2>
```
Comments in the CSS and/or HTML files may contain revealing info

Does there come a point where the troubleshooter doesn't have enough of the original to work with to solve the problem? Hard to say until you try it.

From my own experience, many of the 'this book doesn't work right on my Kobo/Sony' problems I've been involved with have come down to idiotic HTML/CSS decisions by the ebook creator. Once the book's in front of me the cause of the problem becomes clear quite quickly. Text content is usually irrelevant to the process. Others may have different experiences.

RbnJrg · 10-19-2015, 04:43 PM

Jackie, I think there is no problem is we know what book is, if the content (text, images) are scrambled (of course, the best is not to be able to indentify the book). Why do I say that? Because in the rules of MR is present the possiblity of upload a short sample of a copyright ebook. In post #46 of this thread, HarryT says:

Quote:

Uploading a short sample of material is of course, "fair use", and is permissible.

So, if it is ok to upload a short sample where everybody in MR will know what book is, then it doesn't matter if we know what book is if we don't know what is the text of it (IMHO, I can be wrong

)

jackie_w · 10-19-2015, 06:03 PM

Quote:

Originally Posted by RbnJrg

So, if it is ok to upload a short sample where everybody in MR will know what book is, then it doesn't matter if we know what book is if we don't know what is the text of it (IMHO, I can be wrong

)

If HarryT (as spokesman for MR) thinks it's important to obscure the ISBN on the copyright page, I have to assume it might also be problematic to MR if it's visible elsewhere. I'd rather not guess the answer.

Legal debate and philosophical discussion are not really my forte

RbnJrg · 10-19-2015, 06:35 PM

Quote:

Originally Posted by jackie_w

If HarryT (as spokesman for MR) thinks it's important to obscure the ISBN on the copyright page, I have to assume it might also be problematic to MR if it's visible elsewhere. I'd rather not guess the answer.

Legal debate and philosophical discussion are not really my forte

Then rules must be revised

If it's allowable to upload a short sample of a book under copyright where all of us will know what book is, then I don't see why it can't be uploaded a book with the ISBN visible but all other text and images encripted. Of course, it's preferable to have the ISBN obscured, but in cases when that is not possible, I don't see a major problem for that. Again, I may be wrong

Let's wait an official word about this topic.

JSWolf · 10-19-2015, 06:45 PM

Quote:

Originally Posted by jackie_w

I wasn't aware of this Sigil plugin, however it's a long time since I used it. Has it been extended to open AZW3 and KEPUB books these days?

IIRC Sigil used to make some fairly significant structural auto-changes during opening. Is this still the case? IMO this is not necessarily helpful when the aim is troubleshooting. Does anyone else have an opinion?

The problem with the Sigil plugin is that you need Sigil to use it and in using Sigil, the eBook's structure can be changed. Also, it doesn't work with kepub and KF8. So in this case, I think yours is the better one to use.

JSWolf · 10-19-2015, 06:51 PM

Quote:

Originally Posted by Cinisajoy

My question is would this random generator just pick a random letter for each individual letter or would it substitute one letter for a particular letter? Let me show you what I mean.

The Cat in the Hat.
Would it look like this
Atr bho pw cfe qxt
Or
Atr swa yu atr twa?

The first one would be virtually untranslatable, the second one is a cryptoquiz.

Each letter would be replaced with a random letter.

JSWolf · 10-19-2015, 06:54 PM

Some people have been replacing the text in eBooks to post here and no moderator has batted an eye. So why should this be different?

JSWolf · 10-19-2015, 07:06 PM

Quote:

Originally Posted by jackie_w

If HarryT (as spokesman for MR) thinks it's important to obscure the ISBN on the copyright page, I have to assume it might also be problematic to MR if it's visible elsewhere. I'd rather not guess the answer.

Legal debate and philosophical discussion are not really my forte

You might be able to figure out what book it is from the ISBN.

jackie_w · 10-19-2015, 07:38 PM

Quote:

Originally Posted by JSWolf

You might be able to figure out what book it is from the ISBN.

I think the point RbnJbg was originally making was that the short extracts people have been uploading with no objections might be totally identifiable (cover, metadata, text, ISBN) because there are no restrictions other than the amount of text.

It comes down to whether a full-length undecipherable text needs to have much stricter rules than a short fully visible extract to stay legally risk-free. Even though there's less readable text in the full-length version.

JSWolf · 10-19-2015, 07:54 PM

Quote:

Originally Posted by jackie_w

I think the point RbnJbg was originally making was that the short extracts people have been uploading with no objections might be totally identifiable (cover, metadata, text, ISBN) because there are no restrictions other than the amount of text.

It comes down to whether a full-length undecipherable text needs to have much stricter rules than a short fully visible extract to stay legally risk-free. Even though there's less readable text in the full-length version.

If you are scrambling the ISBN number, also scramble the ID as the ISBN is sometimes used there. But you will have to use the same scrambled ID in the OPF in the NCX if the ID in the NCX matches that in the OPF. If they do not match than scramble the ID in the NCX.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Sony ereader troubleshoot	chrisms	Sony Reader	3	10-02-2013 02:08 PM
Out of copyright ebooks/writers you'd most like to see digitized?	pstjmack	Reading Recommendations	18	09-14-2012 08:46 PM
whispersync not working: can anyone help me troubleshoot?	rheostaticsfan	Amazon Kindle	4	10-27-2011 08:09 AM
troubleshoot battery life	kkinser	Amazon Kindle	2	04-18-2011 09:05 PM
The copyright issues of fan fiction eBooks	Kirok	Lounge	33	12-08-2008 06:54 PM

10-18-2015, 03:45 PM	#92
eschwartz Ex-Helpdesk Junkie Posts: 19,421 Karma: 85400180 Join Date: Nov 2012 Location: The Beaten Path, USA, Roundworld, This Side of Infinity Device: Kindle Touch fw5.3.7 (Wifi only)	That's structural, and anyway I think HarryT's issue was that some numbers can identify the book (ISBN). Not sure why knowing the book title matters, but... that isn't the case here.

10-19-2015, 06:54 PM	#102
JSWolf Resident Curmudgeon Posts: 79,756 Karma: 145864619 Join Date: Nov 2006 Location: Roslindale, Massachusetts Device: Kobo Libra 2, Kobo Aura H2O, PRS-650, PRS-T1, nook STR, PW3	Some people have been replacing the text in eBooks to post here and no moderator has batted an eye. So why should this be different?