CC and the cloud: discussion

[eschwartz]

Originally Posted by Purple Lady

Google drive is incompatible with Calibre - https://www.mobileread.com/forums/sho...d.php?t=205581

It is not incompatible with Save-to-disk, though.

[Purple Lady]

Originally Posted by eschwartz

It is not incompatible with Save-to-disk, though.

If you had the Calibre ID in your template like it is used in Calibre folder names it would be.

[eschwartz]

Originally Posted by Purple Lady

If you had the Calibre ID in your template like it is used in Calibre folder names it would be.

No, it still wouldn't -- because calibre needs those ids, and regular old folders don't. They will still contain "one book per directory, assumes every ebook file is the same book in a different format".

It was still braindead of Google to do it, though.

[chaley]

Google drive is incompatible with any app that requires access to metadata.db, such as calibre cloud, my options 1 and 3, and any app we might make to accompany option 2. The base option 2 would work with save-to-disk folders, so it would work with drive.

[Purple Lady]

Originally Posted by eschwartz

No, it still wouldn't -- because calibre needs those ids, and regular old folders don't. They will still contain "one book per directory, assumes every ebook file is the same book in a different format".

It was still braindead of Google to do it, though.

I said it would be incompatible because Google would change the folder names on you if you have numbers in the folder name. Anything that changes your folder or file names on you is incompatible with everything as far as I'm concerned.

[cybmole]

my vote - are we voting yet? - is for support drop box - forget the others. - should we have a poll?

PS I failed with an attempt contact the c cloud developer ( re another issue) - there's no contact us on his website & it is unclear if he is still maintaining /developing the app

- so the sooner you add drop box support ( making that other app redundant) the better - please!

[kaufman]

Originally Posted by chaley

2- a function to CC that can import a book from a calibre-style folder on the device, which is one that contains formats, metadata.opf, and cover.jpg. The disadvantages are that the folder must somehow get to the device, and that the metadata.opf file in the folder must be up-to-date.

I would vote for this one if it had the ability to download directly from Dropbox, rather than from a local directory. If someone has to go from Dropbox to the device, I would rather that CC did it. Also, this fixed the problem of "somehow the folder has to get to the device"

Also, if you do this, please also include a download all missing books, and perhaps a delete all books that are no longer in dropbox.

[eschwartz]

Originally Posted by kaufman

I would vote for this one if it had the ability to download directly from Dropbox, rather than from a local directory. If someone has to go from Dropbox to the device, I would rather that CC did it. Also, this fixed the problem of "somehow the folder has to get to the device"

I first this idea. (Because I can't second it, can I?)

Also, if you do this, please also include a download all missing books, and perhaps a delete all books that are no longer in dropbox.

That would require reading the database and/or doing complicated things. People would even think CC is supposed to sync with the Dropbox library!

[kaufman]

Originally Posted by eschwartz

That would require reading the database and/or doing complicated things.

Why? CC knows the author, title and ID of each book in it's database and it can figure it out for the books in dropbox from the folder structure.

[chaley]

Originally Posted by mrbrimi

Getting back to the original subject; my preference would be option #2 but then I have lots of memory available. But agree that could just as easily import my library into CC right now.

Why does #2 require lots of memory? You would (somehow) download the folders containing the books+metadata.opfs you want and add them to CC.

Guess my question is which "Cloud" services you'd choose? You're not going to keep everyone happy.

Good question. See below.

Originally Posted by gsparks2

Option 2 is attractive to me.

Originally Posted by Terisa de morgan

I would prefer #2, really, because if would help me to update the CC db if I have changed my library but not updated it.

We seem to be developing a consensus.

Originally Posted by cybmole

my vote - are we voting yet? - is for support drop box - forget the others. - should we have a poll?

The big ones seem to be Dropbox, Box, and Copy. I am ignoring Drive which also is discussed, but I won't support something that can eat calibre libraries.

Originally Posted by kaufman

I would vote for this one if it had the ability to download directly from Dropbox, rather than from a local directory. If someone has to go from Dropbox to the device, I would rather that CC did it. Also, this fixed the problem of "somehow the folder has to get to the device"

The base CC will import from a local directory. At some point we will provide a pay-for app that can fetch these directories from [?], ask CC to do the import, then delete the local copy.

Also, if you do this, please also include a download all missing books, and perhaps a delete all books that are no longer in dropbox.

This isn't going to happen. It requires CC itself to be integrated with the cloud, which is precisely what option 2 is intended to avoid.

Note that the combination of wireless connections and the smart eject plugin gives you the "all books not in CC" function. WD connections and the in-library mark on the device view give you an easy way to remove all books in CC that are no longer in calibre's library.

[Terisa de morgan]

Originally Posted by kaufman

I would vote for this one if it had the ability to download directly from Dropbox, rather than from a local directory. If someone has to go from Dropbox to the device, I would rather that CC did it. Also, this fixed the problem of "somehow the folder has to get to the device"

Also, if you do this, please also include a download all missing books, and perhaps a delete all books that are no longer in dropbox.

I agree with all of this.

[kaufman]

Based on information in this thread, I checked out calibre cloud today, and while it has nothing on calibre companion, it is an implementation of most of what I would like for option 2.

Calibre cloud connects to a Calibre library in dropbox (and other cloud solutions) and has an option to send a book to another app.

If calibre companion published the ability to receive epub, you would only have to read the metadata in the received epub and add it to the library. On the next connect, you could grab the missing metadata (computed columns, etc) I can't speak for anyone else, but this would give me almost anything I wanted. Bonus points if you check if the book was already in the CC database and ask if the person wanted to add a duplicate.

The nice thing about this is that it also works for people who get epub in emails or from the Internet.

Is this something that is doable, or am I missing something.

[chaley]

Originally Posted by kaufman

Based on information in this thread, I checked out calibre cloud today, and while it has nothing on calibre companion, it is an implementation of most of what I would like for option 2.

Calibre cloud connects to a Calibre library in dropbox (and other cloud solutions) and has an option to send a book to another app.

If calibre companion published the ability to receive epub, you would only have to read the metadata in the received epub and add it to the library. On the next connect, you could grab the missing metadata (computed columns, etc) I can't speak for anyone else, but this would give me almost anything I wanted. Bonus points if you check if the book was already in the CC database and ask if the person wanted to add a duplicate.

The nice thing about this is that it also works for people who get epub in emails or from the Internet.

Is this something that is doable, or am I missing something.

I really don't want to add a cloud connect to CC. It adds a lot of development and maintenance for almost no payback. It isn't clear if you are asking for that or not. We will consider a separate pay-for app to reach into the cloud and get books from a calibre library.

The other problem is that many people use formats other than epub, and these formats do not contain metadata. There is also no reason to think that epubs contain correct metadata unless they have been saved-to-disk or sent-to-device. The metadata in unprocessed epubs often has the wrong title and author info, meaning that it won't match the book in calibre. We see this all the time with scan books in library. That is why I want access to metadata.opf, which probably does contain good metadata.

[kaufman]

Originally Posted by chaley

I really don't want to add a cloud connect to CC. It adds a lot of development and maintenance for almost no payback. It isn't clear if you are asking for that or not.

Sorry, I wrote that while standing in line at the airport so I probably wasn't clear.

No, I wasn't asking you to do a cloud connect.

Earlier in this thread, someone mentioned a app called calibre connect. If will read a calibre database on Dropbox (and other services) and allow you to send the book files to another app. Right now on my device, the two apps it shows when I select send to app are moon+Reader and ezpdf. What I was suggesting for you is to make the following changes to calibre companion

1) register cc as a handler for epubs
2) add code to accept the epub and read in its metadata and add it to the CC database, and store the epub in the CC folder
3) on the next connect, attempt to grab the rest of the metadata the way you would after finding a file with scan on connect.

I like this approach because it gets cc a good percentage of the way there with relatively little work on your part. For some people, this may be all they need. It does have some minuses

1) it only really works for epub, because only epub have most of the metadata embedded

2) it relies on the epub to have good information

3) it relies on a third party app

As far as 1 goes,thats true, but it's the most common format, so at least you arn't starting with something obscure. If you later wanted to add other formats, perhaps you could have a minimal popup that let people type in information.

As far as 2 goes, since you are only pulling in books that are part of a calibre library, there is some reason to believe that the metadata will have been cleaned up already.

As far as 3 goes, that's true, but it's an app you don't want to write, and you don't need to ask the developer to do anything special to support you.

[chaley]

Originally Posted by kaufman

Earlier in this thread, someone mentioned a app called calibre connect. If will read a calibre database on Dropbox (and other services) and allow you to send the book files to another app. Right now on my device, the two apps it shows when I select send to app are moon+Reader and ezpdf.

I think you mean Calibre Cloud.

What I was suggesting for you is to make the following changes to calibre companion

1) register cc as a handler for epubs
2) add code to accept the epub and read in its metadata and add it to the CC database, and store the epub in the CC folder
3) on the next connect, attempt to grab the rest of the metadata the way you would after finding a file with scan on connect.

I like this approach because it gets cc a good percentage of the way there with relatively little work on your part. For some people, this may be all they need. It does have some minuses

1) it only really works for epub, because only epub have most of the metadata embedded

2) it relies on the epub to have good information

3) it relies on a third party app

As far as 1 goes,thats true, but it's the most common format, so at least you arn't starting with something obscure. If you later wanted to add other formats, perhaps you could have a minimal popup that let people type in information.

The problem here is that by encouraging someone to use calibre cloud we are also encouraging them not to use CC. It is true that the searching capabilities of calibre cloud are minimal, but those along with the library capabilities of the reader could easily be enough for a lot of people.

There is a second issue: CC would show up along side all of the reader apps such as Moon+ and the like, making it look like a reader. We already get a fair amount of static because CC doesn't contain a reader app, and this would make it worse.

As far as 2 goes, since you are only pulling in books that are part of a calibre library, there is some reason to believe that the metadata will have been cleaned up already.

This isn't true. A calibre user must take special steps to update the metadata in the book stored in the library (polish etc) and many people do not. For example, I don't. The calibre forum is full of threads asking why the metadata in the book is not right. Because of this, the book in the library often the original metadata and seldom has the latest metadata in the calibre library. The calibre metadata is in metadata.opf.

As far as 3 goes, that's true, but it's an app you don't want to write, and you don't need to ask the developer to do anything special to support you.

We don't mind writing the app if we get something for it. That is why the separate pay-for app is as attractive as it is.