Unable to convert html to azw or epub

[Lucas Malor]

I discussed this with you many times... Even if Goyal intended to do this (and this is a big concession), this behavior is not conformant with HTML and EPUB 2 specs. If something must work in A way and on the contrary it acts in B way, it's a bug. An enhancement is adding some other functionality. There's nothing more to say.

I can sound repetitive, but as I said to Goyal and to Doctor Ohh, if someone can prove me this is not a bug telling me about HTML or EPUB documentations, ok, I can stop the "whining".

[PeterT]

Ignore HTML specs; they are NOT at all relevant to ePubs.

The ONLY relevant ones are ePub 2 as far as calibre is concerned.

Please post the relevant portions of the ePub specs where this is stated as REQUIRED functionality as opposed to OPTIONAL.

[Lucas Malor]

Ooooh, this is a good point, finally

http://www.idpf.org/epub/20/spec/OPS...m#Section1.3.5

1.3.5: Relationship to CSS

This specification defines a style language based on CSS 2. [] A conforming Reading System must render all OPS CSS 2.0 required subset properties.

http://www.idpf.org/epub/20/spec/OPS...htm#Section3.0

3.0: OPS Style Sheets
[]
3.3: Properties

Default values for all supported CSS properties are as listed in CSS2.

The following table lists all CSS properties and values supported by this specification. Where not all values given in the CSS2 specification are listed for a given property, those values not listed are not supported by this specification. The column “Alternate display” indicates acceptable fallback display for CSS values that a Reading System cannot display as intended.

and "position" is not listed, so it's not required by OPS. *But*, since it's not required, you *may* support it, and as far as I can see, the viewer of Calibre acts differently if the position is static or relative. So the viewer supports "position" property and, if so, you have to look at CSS2 specs, and I do not think that the viewer's behaviour is conformant to it.

So the viewer have to be conformant to CSS2 for the "position" property, or it should not support it at all, considering "position:relative" as "position:static". This could be a quick and easy solution, but IMHO it's not a good solution.

[DoctorOhh]

Originally Posted by Lucas Malor

So you're saying the flow mode is the correct way to see a document like my testcase?

Personally I did not checked your document for myself. You told me that your document worked in calibe's viewer. Did you somehow misrepresent that fact?

Originally Posted by Lucas Malor

Interesting... so can you explain me why Kobo and Kindle and KOReader and CoolReader can show the document as expected?

Why would I care? What those readers do has nothing to do with calibre. But I guess now you have the 4 ways (above) to view your document and calibre's viewer. You might want to check out the EPUBReader plugin for Firefox and you can have a 6th method available.

Originally Posted by Lucas Malor

Since, as far as I know, Kobo and Kindle viewers works in paged mode only, and I used page mode for KOReader and CoolReader.

No they do not use "Paged mode" they have one mode (whatever they call it) and they work in that one mode as designed. For many years calibre had one mode (flow mode) and worked in that mode only as designed, the fact that you couldn't single click between pages was not a bug. In response to users requests Kovid created "Paged mode" for the calibre viewer. There is no reason for me to compare this mode to any device or reader. This mode was created for calibre's viewer as an option for the users of calibre.

Originally Posted by Lucas Malor

As I said to Goyal, demonstrate me this is NOT a bug, using HTML, CSS, epub specs, in one word facts. Everything else is chatters and provocations.

The fact is that you told me that calibre supports your document. Everything beyond this simple fact is simply NOT relevant. Push the button and change modes.

[Lucas Malor]

Originally Posted by DoctorOhh

Personally I did not checked your document for myself.

I would have bet on it.

Originally Posted by DoctorOhh

No they do not use "Paged mode" they have one mode (whatever they call it) and they work in that one mode as designed.

What's in a name? That which we call a rose by any other name would smell as sweet.

Calibre has two modes, not one. And one of them does not work with the testcase. Yes, page mode does not work with the testcase, you can't say it's designed to work this way, it does not work. And it's not me that says it does not work, CSS soecs says it does not work. Read my previous post, since I could bet you've not read it.

Originally Posted by DoctorOhh

For many years calibre had one mode (flow mode) and worked in that mode only as designed, the fact that you couldn't single click between pages was not a bug.

I perfectly agree. In that case if someone asked to Goyal to add the page mode, that request was a enhancement request, since I do not think epub specs require page mode for a Reading System. But, as soon as you implement page mode, anything that is not compliant with the specs is a bug. You can't say it works with flow mode, so it's not a bug, since this is called a "workaround". It's very simple. I repeated this self-evident statement in 100 posts, and the fact you're systematically ignoring it demonstrates that you can't reply to it.


I want to add that I was not trying to impose some additional work to Goyal. He's free to decide to fix the bug or not, thanks to Gods. I was simply trying to understand if the problem was a bug or not, and I wanted to report it. If I was sarcastic it was a response to sarcasm. But please do not try to offend my intelligence. I'm too much sensitive for this and my doctor told me to avoid tirades on ebook readers.

[DoctorOhh]

Originally Posted by Lucas Malor

You can't say it works with flow mode, so it's not a bug, since this is called a "workaround".

I have never stated it is "not a bug".

What I have attempted to convey is that you have "reported the bug." The act of reporting "the bug" concluded on page 1 of this thread. In post 8 you acknowledged a perfectly fine solution to your problem by using flow mode. In post 13 you layed out clearly the circumstances to recreate "the bug".

Since post 13 you have engaged in a campaign of whining about "the bug." Nobody cares what you call it. You have a solution and the developer is aware of how to replicate "the bug." Your job as bug-reporter-in-chief is over, move on.

[Lucas Malor]

This is an interesting point of view, let me explain mine.

My point of view is that I discussed about my problem, Calibre bug or not, until post #19. After that post you dragged me in an off-topic discussion, trying to irritate me saying I'm whining. Do you know how can you translate this in jargon?

Why I don't simply use the flow mode? Well, it's axiomatic: because I opened this thread to discuss about this problem, which affects page mode. Everything I say about it is in-topic and it complies with the thread, the subforum and the forum. So, if you're not interested about the topic, as you demonstrated when you said you not even opened the testcase, you are free to write something else in some other thread. It's a complete waste of time to discuss about something you're not interested at all, no?

On the contrary, if you will want to write something pertinent, you'll be kindly welcome.