Comment by majkinetor | Hacker News Reader

Comment by majkinetor | Hacker News Reader

Don't look any further. Calibre FTW.

Some features (there are thousands of them):

- You can make plugin for any magazine (python) including self-published

- You can scan books for ISBNs and get certain 1 on 1 match online without any effort or search foo. Without ISBN it will search by given criteria

- Batch operations for everything

- It first uses cover of the book (first page in PDF, epub etc), downloads covers if there aren't any, or generates them if nothing can be found

- You can add arbitrary metadata and give them types (bool, text, urls...)

- You can share via web server and access lib from browser and even add books from browser or simply copy them to special folder

- Its x-platform

- It supports zillion formats, and basically you can add zip or anything. Audiobooks could be handled better but ok.

- Its updated weekly for years

- You can have different libraries, groups, virtual libs etc.

- Awesome converter from-to number of formats

- You can both have online tags applied and your own tags. I have personal tags such as `reading`, `must read` etc. along with my own classification on books that isn't merged with online stuff and is kept separatelly.

- Awesome ebook reader

- Many more options

- There are plugins to put kindle encrypted books offline !

Simply, there isn't anything better out there. Anybody not using it simply doesn't know better.

The only thing it doesn't have is in text search, but hey... I am sure its comming one day or you can implement samo mumbo-jumbo on your own.

NOTE: Calibre is IMO not that good for research papers although YMMV. Zotero might be better choice for this.

How well does it work with physical books and magazines, rather than digital goods? Note the OP's wording:

> I would like to take photos of the covers of these, keep track of what box/shelf they're stored in

That is simply metadata.

Very well. "Add empty book", enter the ISBN (or scan the book's barcode) and it will pull all data, including cover image and even ratings from Amazon automatically.

https://dearauthor.com/ebooks/dear-jane-ebooks/dear-jane-can...

Calibre is in python2 and the author has no plans to move to Python3. Even after Python2 is completely deprecated. Not sure I would recommend Calibre right now as a long term solution.

Now explain how that affects users ?

If the program continued to only support Python 2, the plugins you mentioned in your post would need to be written in a language which most operating systems soon won't have an interpreter for, or will require jumping through hoops to get. Plus, it looks like Calibre doesn't come bundled with a python2 interpreter and relies on the system python2 install, so again, when python2 isn't supported on OS-X and Windows and <insert Linux distribution here>, and your favorite package manager doesn't have a copy any more, that will be pretty inconvenient.

With that said, it looks like there are efforts underway to port Calibre to python3: https://github.com/kovidgoyal/calibre/blob/master/README.pyt...

It can always bundle it so, no, nothing would have to be maintained, it could literally work forever (I count on this, Calibre library is among first things on my apocalypse readiness plan :). It does so on Windows ATM.

Plugins are not really a problem in Calibre, I could disable all of them and would still use it as it has enormous amount of functionality OTB. It might be even better to not rely on plugins like on any other system.

So not a problem really, even if it were true.

Python2 is going to hunt us at least several more years anyway.

I don't think this is correct. In the Calibre repository on Github, there is a page on the progress being made porting to Python 3[0], and a pull request that tracked the progress and has been merged into master[1].

[0] https://github.com/kovidgoyal/calibre/blob/master/README.pyt...

[1] https://github.com/kovidgoyal/calibre/pull/870

I also highly recommend Calibre, you may find there are plugins (installable within the application) that assist with physical book cataloging.

Calibre is good, but last time I checked, it didn't let you organize books in a tree-like directory structure, does it still just dump all your books in one "folder"?

https://manual.calibre-ebook.com/faq.html#why-doesn-t-calibr...

"If you are still not convinced, then I’m afraid calibre is not for you. Look elsewhere for your book cataloguing needs. Just so we’re clear, this is not going to change. Kindly do not contact us in an attempt to get us to change this."

In my experience, if a management tool provides good tagging, search and sorting, it doesn't matter how the items are organized inside. Dump an item into the collection, assign tags, it joins the rest in the search and sorting. Over time, you grow and nurture your set of tags.

(The exception is an outliner structure, however books wouldn't benefit from it. And directories aren't an outliner, they simply don't match the speed and convenience.)

Notably, tags themselves can use a hierarchy.

Its better then having folders. Folders are meaningless in organization of multimedia. Better to use tags as they are more flexible.

Then you create virtual folders via tag search. You can show them as tabs or have them in the menu to select. Single book can be in more then one virtual folder that way - for example in Biology and Programming.

But no trees of any kind.

Folders aren't "meaningless", they provide relational information (in form of a hierarchy), something that tags are really bad at.

It's not folders that are important, it's subfolders. Many fields have an implicit hierarchical organization, and reflecting that in some way is extremely helpful. As are tags. Ideally, you want both.

Folders are special case of tags.

Tags can be organized better - its easy to create tag exclusion groups, something that for example Gitlab recently added in the form of tag1::tag2 so when you put one of those any existing one gets removed

You can get folders from tags within VFS, for examle TMSU does this ( a x-platform file tagging system).

As a user? I don't care a lick.

Of course we can map any kind of relational system into a series of strings. What matters is support to do it in an intuitive way. Nested folders are intuitive, in terms of organizing, browsing, and highlighting the existing structure when adding new items.

Tags fail at least at the last two without UI support.

Slightly OT: I want to use tags for my files (general, not books) - is there a good manager out there for that (Linux, but preferably cross-platform)?

There isn't. I search for one for years.

Here are some FYI:

https://tmsu.org/

https://www.tagspaces.org/

http://www.tagflow.ch/en/

Calibre does have full support for multiple libraries, but they're not tree-like since there's only one level.

Tried it a few times. Almost unusable on macOS.

I use Calibre daily on macOS to manage epub/mobi eBooks. Curious what use case you found difficult?

Frustrating, rather than difficult.

1) Crashed on some of my files (don't know why, other readers open them correctly) during the import.

2) Can't really open pdf. In takes seconds to open epub, but a medium sized pdf (say 450 pages) takes up to two minutes easily.

3) After doesn't keep formatting in pdf files (tables\code blocks\etc). Making in useless.

4) By default Calibre will try to open pdf in Preview which, but this way you have to manually keep the count (how many pages you have read).

5) The reader itself is ugly. You can castomize it I assume, but I don't really want to bother.

6) Lot's of unneeded things (just an opinion) like news downloader.

7) Some smaller but still annoying probles I can't really remember.

All in all in makes it even harder to keep track of your books than a good folder structre, it's quite slow and buggy, it does LOTS of thing, but none of them is done great.

I agree with all points: I am lucky that my Calibre workflow avoids PDF files almost completely. Its support there is really lacking. The metadata tracking & organization is good, but anything beyond import + sync to devices can get dicey. With the devs refusal to port to Python 3, I wouldn't be surprised if a challenger begins to siphon away Calibre's users in the near future.

Reply on news.ycombinator.com