Rethinking the user interface

[kiwidude]

@Kovid - I understand your concerns about extending the startup wizard but clearly there is no solution to keep everyone happy - either questions that one set of people click past, or no questions that a different set of people later spend hours scratching their heads over

The reason I suggested putting on that wizard (other than it already exists) was because if the user does not decide on author sort up front before they add their first book, their library gets in a crappy mess that is just awful to fix. Manage Author Sorts, recalculates, tweaks - ugh. I wouldn't have thought two questions asking how you want your titles and authors displayed in a program that is about managing books was a big issue but maybe my grandma has a high pain threshold

As for plugboards, as another option I am throwing out there - is there any harm in perhaps creating a default plugboard once the user selects a device in that startup wizard? Just a basic "series - title" one? So maybe the user doesn't get prompted at all? I don't know enough about plugboards and other devices to know what harm it would cause? I only know about Kindles and for them it is a must have.

[kovidgoyal]

The question about how to sort authors and how to sort titles would be in addition to the questions about: 1) interface language 2) library location 3) device choice 4) email delivery (for a kindle). It's the cumulative effect I worry about.

Having series prepended to titles by default for kindles may be worthwhile. Offhand I cannot think of any problems it should cause. The downside is that it would be surprising behavior, that will be hard to turn off. In my experience, you shouldn't do surprising things without confirmation from the user, it tends to generate a lot of noise from surprised users

[theducks]

Kovid
How about a second level on the start up wizard?

The first level is what we have now, except there are now 2 buttons:
"I'm ready to go with the most common settings as default"
"I am fussy, please let me choose the defaults"

If you pick button 1, the settings are what the majority of users (I will get to this ) use for the chosen device.

Button number 2, just makes the choices that button 1 sets visible.
If the form has not been previously completed for the previously selected device, the form is preloaded with the settings button 1 would use. Prior usage would show the settings AND a Mark Suggested button (reset).

OK Now I am back to the Majority of Users thing
Add a "help us help others" button, that the user can choose to click, that would send additional (to the statistic collection) non-user-identifiable information, on their wizard chosen device and preferences to be used to create a DB of commonly used preference settings that could be used to create a settings pattern.
This might answer the age old question: "What settings do you use with a model K?"

[kovidgoyal]

Having a single button for additional config options in the wizard is a compromise I can live with. Of course, the question remains, who is going to code this I'm the default guy, but my calibre queue is full to the bursting, with things I consider to be a lot more important than this, so if you leave it to me, it will be a while.

[nickredding]

Originally Posted by kovidgoyal

The issue I see here is, there is some demand for a version of calibre that does nothing else but "put book on reader". This version is presumably supposed to be targeted to those people that have no needs beyond "put book on reader". The problems I have with that are:

1) Why use calibre at all in that case? What does calibre buy you over just using the file system?

You are obviously not familiar with the 80% of which I speak. These are people that expect a window to pop up asking them what they want to do when they plug a device in to a USB port. They would typically not know how to get to the file system on the device, and if they could, they wouldn't know what folder to put the books in, and if they did they wouldn't realize the ebook format might not be compatible and needs to be converted. Furthermore, if you explained all this they would likely not retain it if it's something they only do once a month.

2) Where are these books going to come from? If they are DRMed amazon books, then just having amazon deliver them to your device is a much better solution that copying them to your computer and then copying them from your computer to your device, via calibre or otherwise.

True, but there is a large universe of free books out there that can be accessed with a few clicks of the mouse via a web browser (one of the few technical skills the 80% have consistently mastered). But what then?

4) Again, what is so complex about calibre? There are extra buttons on the toolbar, whose function you may not know. As I outlined before the actual basic use case with calibre is extremely simple. If you are the kind of person that gets fazed by that, you're never going to be able to use calibre to the extent that you get any benefit over just using windows explorer, without someone else (like the family tech guy's) help.

The problem with the calibre UI is it looks complicated and unfamiliar. Unlike 95% of Windows/Mac apps it's not icon-based in terms of object management (it's list-based). In short, it is intimidating and fairly shouts at the user "this is for serious users only!!" Well, at least you'd agree with that!

This is a serious question. calibre today has over 6 million installs, the vast majority of which are by people that are in no way good with tech.

You have no evidence of this. 6 million is a drop in the total population bucket of computer users. And how many of those 6 million installs are actually being used?

I remain not convinced that calibre is actually hard to use. Undoubtedly, there are people who find it hard to use, my point is that, would they actually find anything that exposed more functionality than the add books/send to device button easy to use?

My point is that calibre looks unfamiliar and complicated.

My vision for calibre is that it is a tool for people that are serious about maintaining a collection of personal document/books/etc. I am not trying to compete with tools designed purely for throwaway content consumption. This probably means that calibre is never going to be universally used, and for people like us, that know what calibre is capable of, that is a shame, but I believe that there are more than enough people in the world for whom calibre will remain an excellent value proposition.

We agree on this, calibre is for the 20%, not the 80%.

@nickredding: If you are willing to maintain an alternate UI that only exposes a couple of functions, I am open to discussing it. But I remain skeptical of the need for it. Perhaps, if you flesh it out some more. Exactly what would it have and what would it leave out? How would it simplify the functions it has, compared to the current UI? My minimum requirement for such a UI be that it provide a prominent (and permanent) switch to advanced UI button. And that I will not have to maintain it

I have no reason to swim against the stream with calibre, and your comment about the button for a "permanent" switch makes your attitude to this crystal clear.

EDIT: In summary, I am unable to envision a subset of calibre that would actually be useful to a lot of people in the just put book on reader category

I think I have adequately explained why you are completely wrong on this. End of my participation in this discussion.

[kovidgoyal]

Unlike 95% of Windows/Mac apps it's not icon-based in terms of object management (it's list-based). In short, it is intimidating and fairly shouts at the user "this is for serious users only!!" Well, at least you'd agree with that!

Hmm the 80% you know are obviously very different from the 80% I know. You are seriously coming in here to claim that 80% of humans dont know what to do with a list?!

You have no evidence of this. 6 million is a drop in the total population bucket of computer users. And how many of those 6 million installs are actually being used?

Wait, you feel free to make sweeping claims about 80% of the people in the world, and then have the temerity to ask me to provide evidence. Unfortunately, unlike you, I actually can provide that evidence: http://status.calibre-ebook.com/ those represent calibre installs that have been started atleast once in the last three months. Given that the total size of the ebook market is, as best as I can tell, between 10 and 20 million people, I'll leave you to do the rather elementary math.

My point is that calibre looks unfamiliar and complicated.We agree on this, calibre is for the 20%, not the 80%.

No, we dont. I say that calibre is for people that are serious about maintaining a large personal collection of ebooks. That may be 20% of all the people in the world, it may be 2%, it may be 0.2%, I have no idea. I claim that calibre is the tool of choice for such people, irrespective of their tech level. I provide evidence to support my claims, all I see from you is hand waving.

I think I have adequately explained why you are completely wrong on this. End of my participation in this discussion.

You haven't even understood what I was saying, let alone "adequately explained" anything.

[saxondawg]

I'm bewildered by the claim that Calibre has a difficult, user-unfriendly interface. I'm a fairly simple guy and it made perfect sense to me from the beginning. The thing about Calibre is that it goes a long way in being all things to all people. For the power user who is willing to dive into plugboards, etc. (not me), there are options to take it as deep as he wishes. For those who want a nice cataloging of a book collection, with little customization, that can quickly be done as well. It seems like there is a certain amount of unreasonable having-it-both-ways in some of these requests: They want it all simpler, but they want it to cater to specific author sorting (which adds a level of complication). For someone else, it won't be author sorting but some other pet concern about his or her collection.

While I took to Calibre from the beginning and use it every day, my brother was actually confused by it and couldn't convert books. I asked a couple of questions and discovered he was trying to convert .pdfs to kindle e-books. The problem was not with Calibre, but his lack of understanding--a lack many casual users will share--that .pdfs are a special case. Maybe those are the 80% to whom everyone refers. The myth is that almost any software will be simple enough for the quick-and-dirty crowd to use. My brother probably downloaded Calibre, in haste, just because he found out his Kindle wasn't taking the .pdfs he wanted on it. He had that single purpose and wasn't about to stop and learn the program, watch the videos, etc. There is a learning curve for almost any program on a computer. Read the amazon.com Kindle customer discussions and you'll see very quickly just how ignorant the very casual user is. And most of them will never take the time to master software or hardware.

I use Microsoft Word for a living, as an author, and find that nearly everyone I know uses it without knowing 90% of the features right under their noses--even though they've been Word users for years. They can hit the button for italics, maybe two or three other functions. Otherwise it's a typewriter. They can't use styles, review, or other sets of features. And their eyes widen in panic if you offer to show them. But they'll tell you Word is "too complicated."

I agree with Kovid's sentiments in this thread. He has provided a program that is as simple as possible for the very casual user, with minor tradeoffs here and there, while offering powerful features for those who want to go farther. As his installations climb above 6 million, this will bring more and more people who think the program should be this or that to their liking--but the number in itself says he must be doing something right.

[nickredding]

Originally Posted by kovidgoyal

I say that calibre is for people that are serious about maintaining a large personal collection of ebooks.

Exactly, and I am saying it could be useful to a much larger audience ("the 80%") who don't fit your description of the target audience.

You haven't even understood what I was saying, let alone "adequately explained" anything.

No, I understand exactly what you are saying and so does everyone else reading this thread.

EDIT: I can tell from your tone that you are offended by my commments and that was not my intent. I've spent hundreds of hours over the last year contributing to calibre and I consider it to be a very valuable tool which I will continue to contribute to in the future. However, you don't need to be so defensive--just say (as you have here) that this is my intended audience for calibre, full stop.

[Synamon]

Originally Posted by nickredding

Exactly, and I am saying it could be useful to a much larger audience ("the 80%") who don't fit your description of the target audience.

I'm not sure that's true. Probably half of all users will never even learn how to download books and transfer them by cable to their reader. Those people are Amazon 1-clickers (or B&N or Google or whatever) and will never use calibre, so trying to simplify the interface to appeal to them is a waste of time and effort. There is an interface for them already, at Amazon, B&N, etc. The casual reader who reads less than 10 books a year doesn't need to use calibre to organize their books either. calibre really is for the 20%.

[chaley]

I would like to pose a question. I am not trying to be snarky, but it will probably come across that way.

I have put in more than 15 full-time person-months adding to calibre many of its complicated features. A partial list includes plugboards, the template language, most of the custom columns code, a lot of column coloring code, many searching enhancements, restrictions, and a lot of the sony device capabilities. I did these things because they were useful to me, or in some cases because the problems were interesting (custom template functions fall into this category, as do many template functions) or because the requester was a pleasant person.

The question: Why should I care about whether or not these features are usable by the "80%"? I think I should care about them working correctly, where in the end I define "correct", but after that what is in it for me?

Of course, Kovid could refuse to accept my changes, but why should he do that? The functionality is useful to some number of people. Refusing to integrate it would tell me to go away, and tell the "some number of people" that their needs don't matter. These are the people who engage with me, who work with me, and who in the end matter to me. I suspect that we could replace "me" with "us", but I don't want to put words in anyone's mouth.

Disclaimer: Kovid and I have argued several times about function discoverability and approachability. He is almost always more on the side of the 80% than I am.

[fesja]

Originally Posted by chaley

Of course, Kovid could refuse to accept my changes, but why should he do that? The functionality is useful to some number of people. Refusing to integrate it would tell me to go away, and tell the "some number of people" that their needs don't matter. These are the people who engage with me, who work with me, and who in the end matter to me.

That's the perfect recipe to build an unmaintainable software, saying yes to every feature because someone has asked for it. I don't know which features did you implement, so this is just a general thought about usability and how to grow a software program. You shouldn't add a feature until many many people has asked for it. If Calibre has 6M users, you shouldn't add a feature unless is going to be useful to at least 2M.

That's what theory says to avoid having thousands of features that get mixed, that produce bugs, that have to be maintain or that complicate the interface.

I highly recommend you all should read the book 'Getting Real', specially this chapter http://gettingreal.37signals.com/ch05_Start_With_No.php. It applies very well in this case

[kovidgoyal]

Originally Posted by nickredding

Exactly, and I am saying it could be useful to a much larger audience ("the 80%") who don't fit your description of the target audience.No, I understand exactly what you are saying and so does everyone else reading this thread.

And I am saying that it will never be of use to people that do not need to maintain a large collection of books. If you read a handful of books a year, which you dont intend to re-read anyway, why bother with maintaining a book collection? You need to step back and think about calibre's value proposition. Why should someone want to use calibre? What do they get out of it? Is making calibre simpler going to harm the interests of the people that actually have a solid reason to use calibre? IMO trying to simplify the "send book to device" use case further will actually be counter productive to calibre's adoption, by the people that actually have good reason to use it.

I am firmly of the opinion that human beings are intelligent. As long as a complex system is presented clearly, with sensible defaults, in a consistent manner, they can learn how to use it. You may well believe otherwise, but I have to say that my experience with calibre has re-inforced my view. I realize this flies in the face of blogger "wisdom" about people, but I choose to believe the evidence of my eyes and my mind, rather than currently held "opinion".

@fesja: All of chaley's additions have complicated only one part of calibre, the preferences, a part that, as per your philosophy, should not even exist. So people like you are free to ignore its existence. Simply pretend that calibre has no preferences button, rather like an app designed by someone at 37signals And if you, and the author of the book you quote, really believe that software must never have any advanced functions, because they scare off people, I am afraid we are never going to see eye-to-eye.

@saxondawg: Thanks, it is good to hear from someone that uses only the basic use case of calibre. As I have been saying all along, calibre is actually very simple to use, for that basic use case. It gets complicated when you ask it to do complicated things. Dont try and you will never see any of that complexity.

EDIT: I can tell from your tone that you are offended by my commments and that was not my intent.

I was offended, but that is in the past

[BeccaPrice]

I'm a reasonably typical non-technical user - much of this discussion has gone over my head. But I'm having fun learning about some of the more (moderately) advanced features of the program. I'll never use it to it's fullest; I have no need for recipies and regex (whatever that is) and things like that...

But I found Calibre to be pretty intuitive to use for basic use - add a book, modify metadata, convert... yeah, setting up some of the preferences in advance (custom columns ftw - now if I can only remember which of the books in my collection I've read already, so i can set that flag appropriately) would have been nice, except I didn't know what I needed until I'd used the program a fair bit. Being asked to set up preferences that I didn't understand in the setup wizard would have only confused me and turned me off the program.

When I feel more daring, or have a need for some of the more advanced uses, I can learn about them. Even if I don't need them *right now*, I'm glad those more advanced uses are there.

So I guess what I'm saying, Kovid, is that I like the UI just as it is.

[theducks]

Originally Posted by BeccaPrice

I'll never use it to it's fullest; I have no need for recipies and regex (whatever that is) and things like that...

Bet you find a many uses for REGEX

It will be later, but it is oh so handy for renaming, cleaning metadata (and other things)
( Be sure to find the tutorial by Manichean in these forum when you are ready)

BTW REGEX = Regular Expressions
which are search and replace patterns that do way more than the simple
Find: foo Replace: Bar

[etopian]

I personally have been working on a simpler interface to compliment Calibre; I don't want to rewrite Calibre, but only those parts that I don't like; I would still like to use Calibre to do tasks like book conversion... I think Kovind has done great work, and I think he has a very specific design philosophy that I think many people don't agree with... So a few months ago I set out to code up something that was more to my liking... Things like not wanting to put everything under a single program-managed library, and a simpler user interface. If you are interested in helping me test it out, let me know... Screenshot is: http://etopian.net/wordpress/wp-cont.../01/gnosis.png and http://etopian.net/wordpress/wp-cont...s/gnosis-2.jpg It's written in C++ Qt, which I found to be more to my liking than Python; though development has been slower; it's very quickly getting to be a usable product... though nothing as sophisticated as Calibre.