Disappointed in organization options

[rchiav]

It's not so much that I have to use folders. I'm looking for an easier way to manage workflow. I understand that tags work, but the interface is not really optimized for managing workflow. I can copy and paste the files much quicker than I can search for a tag or the lack of a tag and add and remove tags.

I just read in another thread that someone is implementing the ability to save searches, which would also be helpful. I have some suggestions about that that I'll post in that thread though.

[Starson17]

Originally Posted by rchiav

It's not so much that I have to use folders. I'm looking for an easier way to manage workflow. I understand that tags work, but the interface is not really optimized for managing workflow. I can copy and paste the files much quicker than I can search for a tag or the lack of a tag and add and remove tags.

You probably can copy and paste, but that's pretty much all or nothing. It's in a directory or not. I find that the workflow for managing ebooks is far more incremental. I add covers when Google images is open. I may work on a particular author, or a particular time period when I posted books. I may try to solve a particular type of error in the database. Each of those is handled with a different search to locate records that may need work. The idea that I'll enter the ebook and completely add all the correct metadata doesn't fit for the way I work on the database.

If it works for you, that's great.

I just read in another thread that someone is implementing the ability to save searches, which would also be helpful. I have some suggestions about that that I'll post in that thread though.

Calibre currently remembers searches, and the save search feature will be a great improvement. As a matter of interest, that work's being done by Chaley. IIRC, he keeps his author in reverse name order and he wrote a tweak to allow others to do that. I don't do it that way, most don't, but if it works for him, that's great. Same for you - do what works, but keep an open mind.

[chaley]

Originally Posted by Starson17

Calibre currently remembers searches, and the save search feature will be a great improvement. As a matter of interest, that work's being done by Chaley. IIRC, he keeps his author in reverse name order and he wrote a tweak to allow others to do that.

Yes, I am, and yes, I do.

Of course, you are right to point out that calibre is not designed to manage automatic workflow chaining along a multi-step process. I also think that such functionality will not arrive soon, given the amount of other stuff to do that seems to be higher on the priority list.

What works for me for workflow is to use a set of tags with a common prefix to organize what I am doing. For example, I use the tags T:inprogress, T:done, T:waiting, and T:a few others. The advantage for me is that I can query about the state of books by asking for 'tag:T:', query about specific states by asking for the exact tag such as 'tag:=T:done', or find books that have been forgotten by asking for 'not tag:T:'.

I agree with starson17: tags help organize much more than is (usually) apparent when starting using calibre. That being said, as time goes on, I am finding that I also want tags with testable values. For example, I would like to have READ:xx/yy/zz, and then be able to ask for books READ before/after some date. I am considering playing with adding columns to the GUI that show the value of some tag, a poor-man's custom column implementation that would facilitate sorting and selecting on tag values. I can see reasons not to do this, so I am still thinking about it. In addition, Kovid might have alternatives in mind that would obviate the benefits. We will see...

[Starson17]

Originally Posted by chaley

We will see...

Lots of fun stuff still to do

[rchiav]

Really, I have an open mind. I've just been involved in software development for quite some time and I think it's better for software to allow users more flexibility in how they chose to work. Having written software since I was 8 years old (and I'm 38 now), I have a pretty good understanding of what's possible and that you want to enable people to do what it is that they want to do without having to do very much to get there. If a fair amount of people want to keep track of where they are in a workflow process, they should be able to do that with 1 or two clicks, or a drag and drop. If tags are to be leveraged to accomplish this, the UI should support adding, removing and finding these tags in a faster and more well defined way than it does right now.

There's a lot of really good stuff in Calibre, so I don't think it's a question of whether or not the developer or developers have the ability to create such an interface. Obviously they do. It seems to be more of a design decision to funnel people to work one particular way, or an indifference to that type of functionality.

It wouldn't be very difficult to make workflow management much easier and quicker and still use tags if so desired. Having read through a bunch of posts in this sub forum, the reoccurring theme seems to be that user after user gets told to change the way they think and work to adapt to how the application is written instead of others having an open mind to the possibility that maybe the application could be adapted to better fit people.

Again, it's not a matter of "if" something can be done. It's a matter of how user friendly and intuitive it is. Right-mousing to select "Edit meta information" -> "Edit meta information individually" and then having to manually delete or type (which could be typo'd) a tag, or clicking another button that brings up a less error prone way to edit tags but will still take 3 or 4 more clicks is not something that seems very efficient. Sure, it associates information with the book that you can find later, but if it's something you're going to do regularly, shouldn't that be something that only takes 1 or 2 clicks? Then you have to manually type in a query string to filter on that information, or scroll through the tag list? Again, it accomplishes the goal but it's not something I'd consider user friendly.

There was another thread where a woman asked about sharing the same books with her husband but setting it up so that he had his own library and she had her own library. Tags were again given as the solution. No doubt you could use tags for that, but this isn't intuitive to people and it doesn't work the way most people think. The software could still leverage tags to accomplish this, but if the interface clearly identified people and what books were in each persons "library", this person would have been happier than anything. This also wouldn't take very much work.

I probably shouldn't have even written this so this will be my last comments on the subject.

[Starson17]

Originally Posted by rchiav

Really, I have an open mind. I've just been involved in software development for quite some time and I think it's better for software to allow users more flexibility in how they chose to work. Having written software since I was 8 years old (and I'm 38 now), I have a pretty good understanding of what's possible and that you want to enable people to do what it is that they want to do without having to do very much to get there.

If you can articulate an improvement, the bug tracker is the best place to suggest it. Even better, if you can write the code, do that. Kovid went to a lot of trouble to write code that's easy to read and to provide instructions on how to set up a development environment on each of the 3 platforms he supports. There's no doubt that many improvements can be made to Calibre. I just finished one today that I have hopes will be integrated, and Chaley is in the middle of one.

There was another thread where a woman asked about sharing the same books with her husband but setting it up so that he had his own library and she had her own library. Tags were again given as the solution. No doubt you could use tags for that, but this isn't intuitive to people and it doesn't work the way most people think.

Tags are the way Calibre customizes the information it provides to the user about the ebooks it has stored. What she wanted was a way to mark books for herself, for her husband and for both, without having to set up multiple libraries or keep multiple copies of each book. Tags are perfect for storing that type of information about the books. Directories are not.

Any bit of information you want to connect to an ebook is flexibly stored via tags, whether it's workflow status, book content or book ownership. If your concern is just about how many clicks it takes to do a search, see Chaley's stored search thread. He's working to avoid repetition in typing searches. He also has some thoughts about additional fields. Calibre typically releases improvements about every 6 days (except when Kovid is goofing off)

However, I can't agree that just because many users are unfamiliar with the advantages of tags, Calibre should be redesigned to use another interface you may prefer (that so far you haven't articulated very clearly). Most users just need some help in understanding how well tags work to do what they want done.

The software could still leverage tags to accomplish this, but if the interface clearly identified people and what books were in each persons "library", this person would have been happier than anything. This also wouldn't take very much work.

If she follows my suggestions about tags, she'll have a nice window with 2 tag buttons marked "Mine" and "His." She can press either tag and see only the books she wants to see. I don't know what's hard about that. Later she may learn that by doing it this way (instead of separate libraries, which she clearly said she didn't want) she'll be able to set up the content server and both users can access all their books from anywhere in the world.

[animedude01]

Also, you have to remember that Kovid wrote Calibre to scratch an itch. He obviously doesn't have a complex workflow process to didn't see the need to add that in.

Also, the vast majority of users don't have a need for any complicated workflow processing. They process the books as they add them and really just use it to load their books onto their readers.

That's what I do. I normally only imports a few books at a time as I purchase or download them, fix the metadata and load them onto my reader. Books I've finished reading I tag as READ and that's about it. I don't need to do much conversions as my new PocketPro can read both mobi and epub and I generally only use my DR1000 for PDFs.

I think the majority of Calibre users do this too.

Yes, there are more sophisticated users and power users that want it to do a great deal more, but Kovid doesn't feel like it's necessary so doesn't add it in. Being open source, there are other users who have been adding various bits and pieces to do things they want.

If a sophisticated workflow process is important to you, and you've been programming for 30 years, the absolute best thing you can do is to start contributing to Calibre to make it even more powerful. I'm sure there are other users who would love that

Again, most users simply shouldn't be doing things that are overly complex as it just adds to their management burden. The simpler, the better!

[pilotbob]

Originally Posted by rchiav

Any new book that I brought in would immediately fall into the "in-progress" category because I haven't checked it. It you could auto apply that to any imported book, it would be very helpful.

That would probably be a 3-4 line plug-in. Did you create a ticket to request this. I bet Kovid or one of the other calibre devs could whip up a custom plug-in for you in a few minutes.

BOb

[kovidgoyal]

Originally Posted by pilotbob

That would probably be a 3-4 line plug-in. Did you create a ticket to request this. I bet Kovid or one of the other calibre devs could whip up a custom plug-in for you in a few minutes.

BOb

I dont see a use case for an automatically applied tag. If the idea is to mark books you have still to work on, why not mark books you have worked on istead.

With an automaically applied tag the workflow would be:

1) add book, book gets in-process tag.
2) work on book, remove tag name in-process

And you can find in process books with

tag:=in-process

Instead

1) add book. No automatic tags
2) work on book, once done and tag worked-on

and you can find in process books with

not tag:=worked-on

[rchiav]

Originally Posted by Starson17

If you can articulate an improvement, the bug tracker is the best place to suggest it. Even better, if you can write the code, do that.

What version of python is this developed with, or is any recent version good enough?

However, I can't agree that just because many users are unfamiliar with the advantages of tags, Calibre should be redesigned to use another interface you may prefer (that so far you haven't articulated very clearly). Most users just need some help in understanding how well tags work to do what they want done.

I'm not saying that tags shouldn't be used. I'm saying that for some things, there should be some abstraction between the user and manually editing and searching for tags. In the example of workflow, and this is off of the cuff.. you could have right mouse context menus that let you pick a status of a book, and the GUI simply added and removed tags. You could have those user defineable under preference or have a set list. A less specific implementation of something similar could just be a right mouse menu of "quick tags" that would allow you to add or remove a tag that you set up as a quick tag. This would require 2 clicks though.. one to remove a tag and one to add a tag when it changed status. If there was a list of tags just used for status, when a new one was selected, the function could scan and remove any of the other status tags which would only require 1 click.

For filtering on the status, you could have pre canned category of searches that will find the various statuses. I think you'd also want there to be precedence in searches so if you selected to view all ebooks with a specific status, you could leave that view applied and either select another filter/search and/or enter something in the search bar.

Right now, someone is going to have to read up on the syntax of the search bar or scroll through all the tags. And if they decide to use the search bar, what's the likelihood of the average person remembering the syntax of a compound search statement two weeks from now when they try to use it again? What's the likelihood of them being able to quickly get what they want with a couple of clicks on a pull down menu?

You could apply the same type of automation to the people example or other things like weather someone wants to track if they've read a book or not. You could also make these menus user enabled in preferences so you don't have to see them if you don't want to.

I think the open nature of tags is great, but sometimes software needs to be more focused for the specific task at hand as well.

[pilotbob]

Originally Posted by kovidgoyal

I dont see a use case for an automatically applied tag. If the idea is to mark books you have still to work on, why not mark books you have worked on istead.

With an automaically applied tag the workflow would be:

1) add book, book gets in-process tag.
2) work on book, remove tag name in-process

That's a good point. But, as you know, sometimes users want what they want. I thought this was part of the point of creating the plug-in system.

BOb

[kovidgoyal]

No, the purpose of the plugin system is to modularize calibre code to help promote robustness and ease of development/maintenance. In addition, a secondary goal is to allow people to customize aspects of calibre's behavior by writing their own plugins. The existence of a plugin system should not be construed to mean that I will happily implement every feature request that comes across my way. And I should note that the user interface of calibre is not pluginized. Plugins are typically for backend logic.

As far as user interfaces go, my philosophy is to provide as flexible an interface as possible, so users (with a little thought) can customize it to do what they need. At the same time I try to keep the interface as consistent and easy to use as possible, for the vast majority of people that want to use calibre do a few simple common things.

The key point is with a little thought. There are always going to be users who as you say, want what they want, and aren't willing to learn/experiment a little to see how they can use calibre's UI to accomplish what they want.

Typically these are users who are used to a certain paradigm and are unwilling to move off it. That's fine, it's their call, but I'm not going to compromise calibre's UI integrity to cater to them.

@rchiav: I'm not implying you are such a person. This post really has nothing to do with any individuals at all.

[pilotbob]

Originally Posted by kovidgoyal

In addition, a secondary goal is to allow people to customize aspects of calibre's behavior by writing their own plugins.

Well... that right there is exactly what I said isn't it?

Originally Posted by kovidgoyal

The existence of a plugin system should not be construed to mean that I will happily implement every feature request that comes across my way.

I don't think I ever said or expected any such thing. I just pointed out that you could probably easily help this person with a plug-in to customize their and theirs alone calibre installation.

Of course, I know your time is your own and if you can't or don't want to, I would not fault you. However, let's face it, you probably could have helped with a code change or plug-in to add a tag during import in less time it took you to write the post I am replying too. I know how good you are and how well you know your system.

BOb

[ficbot]

I was initially upset that Calibre was going to copy my books into its own directory no matter what, but now I see the advantages and am quite happy to have Calibre manage my files. They all go into one directory so it's easy to back everything up, which is the main concern for me given how much I have invested into my collection so far. But I found that everyone who spoke about tags was right, it really does make it easier. For example I have some mysteries which are novels and some which are short story collections, so I can tag them as 'Genre: Mystery' and at the same time as 'Type: Short Story' and then if I want to read a book which is a collection of short mysteries, it is easy to find them, just as it is easy to find mysteries which are NOT short stories---or short stories which are NOT mysteries.

I think the only thing I would add at this point is a separate database field for 'source' because right now I am using tags for that and I have about six devoted to it. A single text-entry field for it would allow me to streamline my tag list by getting rid of that whole section. As far as 'workflow' goes, I never add it to Calibre until I am ready to work on it (e.g. if I download a library book and don't feel in the mood to seek out the summary, cover art or whatever, I leave it on my desktop and don't bring it into Calibre until I am ready).

[kovidgoyal]

Originally Posted by pilotbob

However, let's face it, you probably could have helped with a code change or plug-in to add a tag during import in less time it took you to write the post I am replying too. I know how good you are and how well you know your system.

BOb

For this one person sure, but if I got into the business of writing code snippets on demand...

That would not end well