calibre 0.6 betas now available

[pilotbob]

I just wanted to say congrats to Kovid and other contributors. This is awesome. I am away for two weeks though so can't help test right now. I would be happy to help wtih documentation and still plan to do some screen casts on using this. I guess I should just stop being lazy and order a damn microphone.

BOb

[kovidgoyal]

Originally Posted by pars_andy

Sorry....forgot to attach

Looks like your reader is using a different firmware. I'll add support for it in the next beta

[kovidgoyal]

@akwant: Those messages can be ignored

@pepak: The available conversion options change depending on input and output formats. For example LRF supportf font embedding while MOBI does not. In order to generate the options list, ebook-convert needs to know the desired input and output formats.

You don't need to specify the output file just the file extension, in which case the output is derived from the iinput file name. For example:

ebook-convert myfile.mobi .epub

will result in myfile.epub

[pars_andy]

Thanks! I'll just go back to using the latest version of calibre 5 until you get a chance to sort it.

[pepak]

Originally Posted by kovidgoyal

@pepak: The available conversion options change depending on input and output formats.

Oh no! I was hoping that the unified code would mean unified arguments as well! This seems even worse than the multitude of converters in the 0.4 and 0.5 versions - not only will the arguments differ, but I won't even learn of a full list unless I do a help for every format :-\

Wouldn't it be better to simply show all possible options with --help, possibly marking those specific to a certain format with something like [epub,lrf]? That way the filename would not need to be the first anymore and --help would work as expected.

[pepak]

Another possible solution: Add a new option, e.g. --relaxed, which would not generate a fault on unknown option but simply skipped that option. Then input/output format wouldn't be needed for --help either, AND I could have an unified command-line that would work for every format.

[kovidgoyal]

The format specific options are grouped under input and output options. the vast majority of options are common to all formats. Except that in some cases some options are disabled for some formats.

[kovidgoyal]

Originally Posted by pepak

Another possible solution: Add a new option, e.g. --relaxed, which would not generate a fault on unknown option but simply skipped that option. Then input/output format wouldn't be needed for --help either, AND I could have an unified command-line that would work for every format.

You can easily have a unified command line for all formats. Just avoid format specific options.

[pepak]

Originally Posted by kovidgoyal

The format specific options are grouped under input and output options. the vast majority of options are common to all formats. Except that in some cases some options are disabled for some formats.

In that case it makes sense to allow ebook-convert --help which would display only those universal options.

You can easily have a unified command line for all formats. Just avoid format specific options.

I hope that will work, but I am afraid that I will have to use format-specific options to get a nice output.

[kovidgoyal]

Originally Posted by pepak

In that case it makes sense to allow ebook-convert --help which would display only those universal options.

No because the defaults for the various options change depending on the input and output formats. For example some input format might have a very well defined structure which would allow calibre to correctly detect chapters automatically, instead of using the default chapter detection expression.

[kovidgoyal]

Originally Posted by pepak

I hope that will work, but I am afraid that I will have to use format-specific options to get a nice output.

If you're going to try to tweak things a lot, it will be difficult to acheive a universal command line. Still it will be easier now than it was, since a large number of options that had completely different semantics are now common.

[pepak]

Guess I will wait for 0.6.0 final and then see how it works. Right now, the fact about filenames being the first arguments even for --help really fills me with gloom :-(

[kovidgoyal]

Originally Posted by pilotbob

I just wanted to say congrats to Kovid and other contributors. This is awesome. I am away for two weeks though so can't help test right now. I would be happy to help wtih documentation and still plan to do some screen casts on using this. I guess I should just stop being lazy and order a damn microphone.

BOb

Contributions of documentation and/or screencasts are most welcome. th eless time I have to spend doing that, the more time I have to spend on new features and bug fixing.

[=X=]

Major issue: viewer is not working I've tested LRF, ePUB, LIT, and PDF the the viewer always breaks (except LRF)

MINOR: On File import the "pdb" extension is not included so the file is not seen.

=X=

[pars_andy]

Originally Posted by pars_andy

I also get this error on some occasions if I move from epub output to lrf output in the conversion window.

Traceback (most recent call last):
File "calibre\gui2\convert\single.pyo", line 152, in setup_pipeline
File "calibre\gui2\convert\single.pyo", line 134, in widget_factory
File "calibre\gui2\convert\lrf_output.pyo", line 39, in __init__
File "calibre\gui2\convert\__init__.pyo", line 52, in initialize_options
File "calibre\gui2\convert\__init__.pyo", line 73, in apply_recommendations
File "calibre\gui2\convert\__init__.pyo", line 123, in set_value
Exception: Can't set value u'' in opt_serif_family

Just wanted to confirm that it is definitely beta 2 I'm running. Version number 0.6.0b2.