Finding books by metadata

[NiLuJe]

I'm not *quite* sure why you'd sacrifice the Series field for that when tags are just fine, but, sure, it should work ;p.

[mst]

I'm not *quite* sure why you'd sacrifice the Series field for that when tags are just fine.

Because some of my bookshelves have longer comma separated names, which won't work as tags. And because tags are subsets of my bookshelves (now series).

[JSWolf]

Originally Posted by Frenzie

From the perspective of many of us directories are better. But indeed an interested party could certainly provide both.

I'll disagree with that. Kobo can organize eBooks a number of different ways to find what you want. Directories are not nearly as capable. You can put the same one copy of an eBook in different collections, you can sort by series, you can see books currently being read and more.

[JSWolf]

Originally Posted by NiLuJe

"Bookshelves" are a Nickel concept. There's no direct equivalent.

They are no longer called bookshelves. Kobo did away wit that in version 2 of the firmware and now that are called collections.

Originally Posted by mst

The +XYZ tags are a great idea, many thanks!

I made a custom category in Calibre that I called "bookshelves" and then forget that it was custom. I am converting this to series now and see how that works.

While you have a bookshelves column in Calibre, Kobo doesn't do bookshelves. What Kobo does is collections. Kobo changes the name. So you would be best using the correct terminology.

[Frenzie]

Originally Posted by JSWolf

I'll disagree with that. Kobo can organize eBooks a number of different ways to find what you want. Directories are not nearly as capable. You can put the same one copy of an eBook in different collections, you can sort by series, you can see books currently being read and more.

You can't disagree with my simple statement of fact. You might quibble that "many" KOReader users are still a very small niche and therefore not "many," I suppose.

Setting aside nitpicking of semantics, there's nothing to disagree with in what you wrote, only in what you left out. Assuming a perfect world, having to put it in a collection would be about as easy and portable as quickly putting it into a folder, but that's neither here nor there.

There are two big issues with metadata. I've only mentioned the technical problem so far, which luckily has been mostly resolved on modern Kobo software. Putting your latest Humble Bundle on your device is not nearly as bad as it once was.

The second and much more important aspect is that I don't want to painstakingly edit metadata to be able to find a document. I just want to quickly put it in a folder and go. I described what my world looks like here: https://github.com/koreader/koreader...ment-284213219 It's not like you have an on/off switch, between no metadata and perfect metadata. That just might work. No, in reality the majority of metadata is mostly useless. So I could waste my time in specialized programs like calibre, or I could just quickly place a file in an appropriate folder with an appropriate name.

Metadata offers many theoretical advantages, such as having the same file show up in multiple locations as you mentioned (e.g., in scifi, in Arthur C. Clarke, and in '80s fiction). It also easily enables doing rather silly things like finding the third book of every series. Your imagination's more likely to be a limiting factor there than with simple folders. But those advantages are meaningless to me when all I want to do is to quickly read something on the go or on something other than a monitor. Most of the time I don't want to play a librarian creating a catalog card; I just want to read something.

Incidentally, the ebooks I buy often have worse metadata than those I get for free from sources like Gallica, DBNL or Project Gutenberg. Archive.org varies.

If you always have perfect metadata, that's cool. If you like editing your metadata, more power to you. But without KOReader's basic file manager, I'd have returned the H2O I bought back in 2014 on the basis of its screen, naïvely assuming it was otherwise more or less the same as a BeBook 2 with better hardware. That's the ereader I want, and the closest you'll get is a Kobo Forma or Libra + KOReader. Metadata is a great addition, but it's like the current cart requirement in grocery stores that haven't installed a counter at the entrance and exit. A cart is occasionally useful, but most of the time it's a bother at best.

To be clear, I wrote a quick proof of concept metadata DB search GUI a couple of days ago (or rather, I transplanted a DB search onto the existing file search). Providing a basis for people to build upon tends to result in improvements more often than if they have to build it from scratch. So unless I won't have time, coming sometime soon.

[Ken Maltby]

I'm one of the "many" and find that a directory structure and file naming protocol, make it easy to find the books I want in my close to 8000 e-book library.

Luck;
Ken

[JSWolf]

The problem is with not checking the metadata of the eBook before you send it to your Reader. If KOReader is going to read the metadata from the OPF, then you need to make sure the metadata is correct in the OPF. For example, one of the eBooks in the Honor Harrington series has these tags in the metadata...

Action & Adventure, Fiction, General, Military, science fiction, space opera

That (IMHO) need to be edited.