Translating Sigil

[user_none]

Now that metadata can be translated in the GUI please be aware that the Dublin Core Metadata that EPUB uses already has some translations at http://dublincore.org/resources/translations/#dces1-1 . This should make translating the metadata attributes and descriptions easier.

[Jellby]

Damn! I completed the translation yesterday, back to work now!

[Toxaris]

Oh well, I was almost afraid to get bored...

[Jellby]

What is the "SP" language in the language strings? Is it supposed to be the programming language?

[meme]

Can you give an example/section of text/code?

[Jellby]

No, because I only have the translation file from transifex, but it says:

<location filename="../../Misc/Language.cpp" line="278"/>


By the way, could all the language names be imported from Calibre instead? I know the format is different, but the language list is much more extensive there, and copying them one by one is a pain.

[meme]

SP was a subtag for Serbian, though its not clear what it was for. I'll remove it as Serbian is already there. Of course, anything missing can be added later if someone identifies it.

The list was what was available already in Sigil, so I assume it was a list of standard names valid in EPUB books.

As for importing from calibre - what do you want imported where? If I import the language names/codes from calibre into Sigil's source code then you still have to update the Transifex translations for each one. Or do you mean copy translated names from calibre's files.

Maybe user_none might have a suggestion here as he's more familiar with calibre and the translation setup.

[Jellby]

I mean the language codes, names and translations, which in calibre (launchad) are in a separate .po file, although calibre seems to follow ISO 639-3, while the ePUB spec refers to ISO 639-2...

[user_none]

Originally Posted by Jellby

By the way, could all the language names be imported from Calibre instead? I know the format is different, but the language list is much more extensive there, and copying them one by one is a pain.

Calibre and Sigil handle language codes and translations in very different ways. It wouldn't be possible to make them match.

[Jellby]

OK then, I thought it would be easier.

Now what's killing me is all this Dublin Core stuff, too many terms I don't know the exact translation, or overlapping translations.

[user_none]

Originally Posted by Jellby

OK then, I thought it would be easier.

Now what's killing me is all this Dublin Core stuff, too many terms I don't know the exact translation, or overlapping translations.

It's very strange that Spanish isn't one of the reference translation languages...

[meme]

See user_none's post above with pre-translated Dublin Core text. Edit: which as he just posted, doesn't help at all

[Griffel]

Dear all,

For some while I follow this forum in some areas by reading.
Today I decided to participate by becoming a member.

I'm working on the German Sigil translation.

And I'm wondering about the amount of metadata entries: 454, this means 227 pairs of metadata fields.

Originally Posted by user_none

Now that metadata can be translated in the GUI please be aware that the Dublin Core Metadata that EPUB uses already has some translations at http://dublincore.org/resources/translations/#dces1-1 . This should make translating the metadata attributes and descriptions easier.

As far as I understood, the Dublin Core Metadata set consists of 15 fields, according to several sources.

The links to translations I saw handle only 15 fields.

Are the other fields really part of the standard?
Or are they non-standard and optional?
I'm speaking of the amount of the fields, not of the entries.
If the fields are non-standard and if the field names are free, wouldn't it be a better idea to let the users decide to add field names by themselves?

A lot of the fields seem to be related to movie business, like "dancer", "woodcutter", "wood-engraver" and so on. They don't make sense in an ebook.

Are all these fields really needed? They will cause a lot of work (and may be frustration) for all translaters.

I couldn't find any list of all this fields of "Dublin Core Metadata" in a German translation. And I assume, other translators to other languages will have a similar problem.

For all who are interested in the German translations, here are some hints:
SELFHTML
Wikipedia
http://www.community-of-knowledge.de...ta-initiative/

Cheers,

Griffel

[mmat1]

Originally Posted by Griffel

If the fields are non-standard and if the field names are free, wouldn't it be a better idea to let the users decide to add field names by themselves?

Hello,
I would guess, they are standard. Of course, usage is optional.

A lot of the fields seem to be related to movie business, like "dancer", "woodcutter", "wood-engraver" and so on. They don't make sense in an ebook.

They make sense in some very rare situations. I.E. Woodcutter is defined as a person who cuts an image into wood (see definition in Sigil). This reminds me on the the way, how pictures are produce until ~1850/60. I.E. the early works of W. Busch are painted on a shelf, engraved by a woodcutter and then printed. Well, who remembers the name of the woodcutter today ?

Are all these fields really needed? They will cause a lot of work (and may be frustration) for all translaters.

I couldn't find any list of all this fields of "Dublin Core Metadata" in a German translation. And I assume, other translators to other languages will have a similar problem.

As user_none says, there are some translations (i guess, just the 15 basic attributes).

[Griffel]

@mmat1

To make my theme more clear, please have a look to a screenshot I took from the translation program "Qt Linguist": Screenshot of "Qt Linguist"

It's only a small extract of the 454 entries in the section "Metadata".

Are you aware of all these 227 categories?

From my point of view none of the entries of the screenshot is related to books or ebooks. These entries are typical for movie business.

Only a subset of few categories in the Metadata section makes sense in an ebook.

We all know the multiple and annoying kinds of categories from the end credits of movies.
But, in book business, these categories are not in use.

For example "woodcutter". No one in book business uses this category. Typical for book business is the category "illustrator". No one cares about the technique an illustrator uses. But the category "illustrator" is not part of the core. Therefore it makes sense to provide "illustrator" as additional field, but not "woodcutter".

I showed the screenshot to my girl friend. She is involved in book business for a long time. Her answer was: "It looks like an alien speech."

Only few categories of metadata are in use in book business. Most of them are described in the core. But some other additional categories make sense. The challenge is, to identify these categories and to shorten the long list to an appropriate subset.

I highly recommend to do so. Even if it's not easy. I would appreciate to support this action.