Help with Library Sync

[JSWolf]

Originally Posted by Puni

I haven't tried calibre in Linux, perhaps that might be a different experience when adding, going to be a weekend project.

Same code on Linux as on Windows. Your filenames will still be renamed.

[theducks]

Originally Posted by Puni

That sounds pretty good, I'll try moving my calibre management server to a linux docker container, share out the storage path over SMB, and have my PCs autoimport from the share. Thanks

I've never heard of someone sharing an Auto Import folder.

Multiple users can use the Content Server. If the logged in user has the proper Content Server rights, they can add that way and make changes.

It is a Web Interface, not a folder share.
AFAIK The user does net even need an account on the host because of this.

[DNSB]

Given that calibre supports multiple OS, you will find the path truncation on any supported OS. That way if I move my library to a different OS, it will not crash and burn. If calibre and whatever you are reading on supports metadata, the actual file name does not matter.

On Windows, if I enable the LongPathsEnabled registry setting (Windows 10, 1607+), it will make no difference to how calibre stores filenames though a 32,767 character path would be allowed.

[BetterRed]

Originally Posted by DNSB

. . .

On Windows, if I enable the LongPathsEnabled registry setting (Windows 10, 1607+), it will make no difference to how calibre stores filenames though a 32,767 character path would be allowed.

I've considered enabling it, it can be done in the Group Policy editor, but I don't have a compelling reason to do so. Nor do I trust all of my installed applications to cope, some started life in the mid 1990s on OS/2 and were last built more than a decade ago.

Q. if it's enabled, do you know if calibre's Save to disk will allow a long path name?

BR

[DNSB]

Originally Posted by BetterRed

Q. if it's enabled, do you know if calibre's Save to disk will allow a long path name?

I've had LongPathsEnabled enabled since it first became available. Save to disk follows what I mentioned in my previous message:

it will make no difference to how calibre stores filenames though a 32,767 character path would be allowed.

Which makes sense since if I was using a Windows 7 system with 3.48, using >260 character paths would make the libraries incompatible.

[Puni]

Originally Posted by theducks

I've never heard of someone sharing an Auto Import folder.

Multiple users can use the Content Server. If the logged in user has the proper Content Server rights, they can add that way and make changes.

It is a Web Interface, not a folder share.
AFAIK The user does net even need an account on the host because of this.

It's a different use case, single user many devices at different locations. I keep lots of docs in calibre and I don't want to manually pull every file since I last used a device. I keep a small local server at each location that replicates all my critical binaries. I was using sync on my laptops to keep my calibre folder replicated but if I simply have the server replicate and present it as a share, I can have the laptops auto-import from the share instead. No race condition, and fast. If I had to hit a remote calibre server via VPN over WAN, it would take a long time for some of the larger files with lots of technical drawings. That's why sync is important, keeping the documents and tools you need on hand rather than on-demand. Even though each location has pretty good bandwidth, the latency means it's very hard to take advantage of it.

[Puni]

Originally Posted by JSWolf

Same code on Linux as on Windows. Your filenames will still be renamed.

Yep looks like that's the case. Even on very short authors/filenames, it always romanizes DBCS names and has some weird truncation for the longer ones. Japanese kana gets romanized using hepburn-shiki but Japanese kanji is romanized using Chinese Pinyin so it's mostly incomprehensible. It looks like the independent client importing from a read-only share is working well though so I can at least see the titles from within the calibre application on multiple devices.

[JSWolf]

Originally Posted by Puni

Yep looks like that's the case. Even on very short authors/filenames, it always romanizes DBCS names and has some weird truncation for the longer ones. Japanese kana gets romanized using hepburn-shiki but Japanese kanji is romanized using Chinese Pinyin so it's mostly incomprehensible. It looks like the independent client importing from a read-only share is working well though so I can at least see the titles from within the calibre application on multiple devices.

The renaming and shortening does not matter at all. It's the template you create for when you save to disk that matters. That's how you get the filenames to be as you want when the eBooks are not part of the library.

[slm]

Originally Posted by JSWolf

The renaming and shortening does not matter at all. It's the template you create for when you save to disk that matters. That's how you get the filenames to be as you want when the eBooks are not part of the library.

You are being blind to the ways people other than you may use Calibre.

For example, I sometimes want to access a book when I am not at home and don't have access to my home network.

I can get it from Dropbox, but because of the naming issue, I need to use an intermediate step:
To access the books from my server, therefore, I have Calibre create a catalog periodically, with both the full names and the paths and upload the catalogue to my server. Since the library itself is in a Dropbox folder, I can access a book file by finding the book in the Catalogue and then using the indicated path to get the file from Dropbox.

All of this works fine for my purposes. Save to Disk would be worthless for this particular problem.

I assume others have other reasons for using Calibre differently that you do.

[DNSB]

Originally Posted by slm

All of this works fine for my purposes. Save to Disk would be worthless for this particular problem.

For my uses, I do a Save to Disk to a directory that I sync to DropBox. This gives me directories based on the first author in the author_sort list with the books saved in those directory as [series series_index - ]title - first author in author list.

i.e.

Andrews, Ilona/Rusalochka - Ilona Andrews.epub

Robb, J. D_/[In Death 58] - Payback in Death - J. D. Robb.epub

Robb, J. D_/[In Death SS 03] - In Death Christmas Collection, The - J. D. Robb.epub

Tillman, Charles/[Big Easy Bounty Hunter 05] - More Monsters, More Mayhem - Charles Tillman.epub

Template used is (I am considering changing the series index to 3 digits since one series I'm reading has reached #86):

{author_sort:sublist(0,1,&)}/{series:|[|} {series_index:0>2s||] - }{title} - {authors:sublist(0,1,&)}

Originally Posted by slm

I assume others have other reasons for using Calibre differently that you do.

Of course, calibre is a massive Swiss army knife and most of us never do more than scratch the surface of it's capabilities. To quote one comment about Microsoft Word, 90% of the users never use more than 25% of it's capabilities, the problem is that no two users agree on which capabilities to use.

[Wiggo]

Originally Posted by slm

You are being blind to the ways people other than you may use Calibre.

For example, I sometimes want to access a book when I am not at home and don't have access to my home network.

I can get it from Dropbox, but because of the naming issue, I need to use an intermediate step:
To access the books from my server, therefore, I have Calibre create a catalog periodically, with both the full names and the paths and upload the catalogue to my server. Since the library itself is in a Dropbox folder, I can access a book file by finding the book in the Catalogue and then using the indicated path to get the file from Dropbox.

All of this works fine for my purposes. Save to Disk would be worthless for this particular problem.

I assume others have other reasons for using Calibre differently that you do.

Extremely cumbersome - take an app like Calibre Sync, connect to Dropbox and all your problems are gone.

[DNSB]

Originally Posted by Wiggo

Extremely cumbersome - take an app like Calibre Sync, connect to Dropbox and all your problems are gone.

Oddly, my Kobo ereaders have issues with running Calibre Sync, not to mention that the instructions seem to require that I move my calibre libraries to a cloud service which is NOT A GOOD IDEA!

To move your Calibre library to cloud services:

1) Open Calibre on your Laptop/PC/Desktop
2) Select "Switch/create library..." from the Calibre menu
3) Select an empty folder on your Cloud drive in "New Location" drop down
4) Select "Move current library to new location", then press OK
5) Wait for cloud sync to complete.
6) Then you are ready to use Calibre Sync.

Whereas the Save to Disk and sync changes works rather well and without placing my libraries in jeopardy.

See I am getting errors with my calibre library on a networked drive/NAS? or you can check SQLite Over a Network, Caveats and Considerations for more detail.

[Wiggo]

As @slm already has his library on Dropbox the cloud issue doesn‘t really matter.
Additionally maybe it’s only a copy, then even less.

And regarding Kobo I would connect it with your mobile, as Calibre Sync is available for iOS and Android - it‘s really easy.

[DNSB]

Originally Posted by Wiggo

As @slm already has his library on Dropbox the cloud issue doesn‘t really matter.
Additionally maybe it’s only a copy, then even less.

And regarding Kobo I would connect it with your mobile, as Calibre Sync is available for iOS and Android - it‘s really easy.

Sadly, I rarely read on my mobile and trying to copy books from a mobile device to a Kobo ereader is not something I want to try again. I've played with Calibre Sync on iOS (iPhone and iPad Pro) and on Android (Google Nexus 7 (2013) and Lenovo Tab M8 HD) and I was not that impressed. If I put my entire library on DropBox, it was rather painful finding an individual book in the ~14,000 available from a mobile device.

For me, Save to Disk and copy to DropBox works with minimum effort on my part and is just as easy without the risks associated with having my library in the cloud.

[slm]

Save to Disk and Calibre Sync undoubtedly work fine in the contexts they are being used.

My approach (which is only partly described above) suits me, in part because I can access my books from PocketBook or a computer even when I'm not connected to my home network.
I have, in effect set up a web page that shows the Catalog I've created, navigates to individual books with cover and description and lets me download or email any of ebook files. This works for me.

It may well have been overkill to write the php and batch scripts but they are done now.