0.8.36 deleting news

[luka6000]

Hello developers,

I'd like to share with you my thoughts about this new feature of last 0.8.36 calibre version:
"When automatically deleting news older than x days, from the calibre library, only delete the book if it both has the tag News and the author calibre. This prevents accidental deletion of books tagged with News by the user."

With this version, calibre is not showing news author correctly as set by recipe - it just uses 'calibre' within library although epub files have right metadata.

I'd like to convince you, that this new feature is just wrong. First of all, calibre is not and can not be named as author of content it creates using recipes. I understand that if recipe author doesn't bother to set content author correctly, calibre must set something but why not use 'unknown' instead? Please remember, that this content is probably copyrighted and I'm not talking about money involved but just plain simple GNU GPL rule of keeping the original author always in place. You want your rights to be respected, please respect rights of other authors.

So, what to do to make is better and right? Just use Producer (in epub) or Contributor (in opf) fields instead Author. Put something like 'calibre news downloader' there and it will be just so true. It will also make it differentiate from standard calibre producer field so it would be very easy to recognize old news.

Please, give this feature a second thought. Thanks.

[kovidgoyal]

1) calibre has always set the author of downloaded news to calibre. There was an unintended loophole earlier that recipe authors could use to get around this, that has now been closed.

2) Many ebook formats do not support metadata beyond title and author, and the code is designed to work with them as well. That is not going to change.

3) Setting the author to calibre is not a means to claim authorship of the content. To imply that it is, is silly. I doubt anyone suffers under the illusion that calibre claims to produce the content downloaded by the fetch news system. Authorship information as well as copyright notices at the bottom of articles are preserved by the download process. Setting the metadata author field is simply a way of signalling that the ebook file has been created by calibre. A way that actually works across multiple ebook formats.

[kovidgoyal]

Oh and you can still set the publisher field to whatever you like in your recipe. Since most news sources typically have multiple authors and a single publisher, that is arguably more appropriate to set anyway.

[luka6000]

Originally Posted by kovidgoyal

1) calibre has always set the author of downloaded news to calibre. There was an unintended loophole earlier that recipe authors could use to get around this, that has now been closed.

ok, so why not rethink this idea? It's just the information within library metadata, so why not make it true instead of putting false author?

2) Many ebook formats do not support metadata beyond title and author, and the code is designed to work with them as well. That is not going to change.

Library metadata does support lots of fields - why use 'author' for tracking recipe download? There are better metadata for that.

3) Setting the author to calibre is not a means to claim authorship of the content. To imply that it is, is silly. I doubt anyone suffers under the illusion that calibre claims to produce the content downloaded by the fetch news system. Authorship information as well as copyright notices at the bottom of articles are preserved by the download process. Setting the metadata author field is simply a way of signalling that the ebook file has been created by calibre. A way that actually works across multiple ebook formats.

I understand 'why'. I'm just saying that it's not the only way and maybe you could give it a second thought?

As for now, calibre is actually replacing original author with 'calibre' and it does that only with library metadata. If you insist that there is a need to mark a file created by calibre, maybe make it like 'NY Times by calibre' if there is 'NY Times' author already set. Maybe let it be configurable?

All I'm saying is that technology is not an issue here. Calibre should use different field for tracking news download if you decided that it should track this kind of content. And calibre library is already capable of doing this better. The issue with file formats supporting only title and author is a different problem. Are there any plans on how long should new calibre features all work on every file format? Is it good for development?

[kovidgoyal]

I have told you why, you are free to disagree with my reasons, it is not going to change.

[DoctorOhh]

Moderator Notice
Read the "PLEASE READ BEFORE POSTING - The purpose of this subforum" sticky before posting in the development forum. Thread moved.

Originally Posted by luka6000

As for now, calibre is actually replacing original author with 'calibre' and it does that only with library metadata. If you insist that there is a need to mark a file created by calibre, maybe make it like 'NY Times by calibre' if there is 'NY Times' author already set. Maybe let it be configurable?

You do understand that NY Times is the Publisher not the Author, don't you? You might consider modifying your recipes to set the appropriate publisher.

[luka6000]

I have told you why, you are free to disagree with my reasons, it is not going to change.

There are many calibre configurable options that are far less important than setting the author/publisher data wright and jet this is set by calibre without question. Will news tracking feature work if I decide to manually change the author from calibre to anything else? Someone might do that after seeing setting it wrong. Anyway, thanks for your replies.

Originally Posted by dwanthny

You do understand that NY Times is the Publisher not the Author, don't you? You might consider modifying your recipes to set the appropriate publisher.

Remember that there are file formats that do not support Publisher data

Cheers!