The Old Joel on Software Forum: Part 3 (of 5)

About Face 2

Went down to the bookstore recently and was looking through the advanced computing topics and found, next to Joel's fine tome, About Face 2 by Cooper. Hadn't seen it before so flipped through and every page I looked at had insights that I agreed with. Like don't make the user do unnecessary stuff that has nothing to do with their task. He calls unnecessary stuff 'excise'. Andyway, bought, read it, liked it. Then went to check Joel's list just to confirm it wasn't there and actually the first edition was! Well this is a 10 years later revision.

The really funnest part was, after finishing, reading the reviews on Amazon for both editions. The ratings are evenly divided between 5 stars 'best book I ever read' and 1 star 'don't bother the author is a pompous blowhard and the nonadvise he gives is utterly useless.' One things for sure - every one has a passionate opinion about this book!

Dennis Atkins
Friday, January 23, 2004

Go back and get his other book - 'The inmates are running the asylum' as well.

'About Face' is all theory, which is what causes the 'pompous blowhard' quotes.

'Inmates' is theory put into practice. Read the section on the in-flight movie system, how they turned junk into a winner.

The '1 star' crowd can jeer all they like, you can't deny the existence of a real-life working system.

(btw, 'Inmates' has an anti-programmer slant - upsets many people)

AJS
Friday, January 23, 2004

Imo, you should never go to just one online place and read the reviews of books you are interested in purchasing. I always do a Google search and try to find reviews that aren't associated with a book store.

One Programmer's Opinion
Friday, January 23, 2004

Huh?
About face (1.0) is a very practical guide to designing WIMP (Windows Menus Pointers) interfaces. It's not theoretical at all: it's very hands on and specific. To the point that it talks about whether menu items should be capitalized and stuff like that.

The Inmates are Running the Asylum, on the other hand, is a "usability for manager types who don't get it yet" book. It's a great read, but it's neither theory nor practice: it's motivation.

j b
Friday, January 23, 2004

I agree with JB. About Face (first ed. anyway) is very practical and VERY VERY good. It's on my short list:

About Face
Code Complete
Don't Make Me think (more for web useability, but very very good and to the point)

Cooper invented (Ruby ?) which became the Visual Basic form designer.

The real Entrepreneur
Friday, January 23, 2004

I think that Cooper's tone, both in AB2 and even moreso in Inmates, continually hammers home the point that developers shouldn't be designing interfaces sets a lot of people on edge.

Funny thing though, letting a message like that set you on edge, kind of reinforces his points in a way.

Obviously there are programmers who are, or would make, excellent UI designers, but the separating design (even beyond just interface design) and development into distinct disciplines provides a lot of value.

herb
Friday, January 23, 2004

'About Face' says here's what you should do. Menus go like like this, buttons like that, cross your t's & dot your i's. That's theory. No-one says it will work.

'Inmates' (in part) says here's a real live working we-got-paid-for-it-and-people-use-it system, and this is how we did it. That's practical. See! It does work! Nyaa!

Plenty of theory out there in any field you like. Writing a book about it is easily done. 'Look - here's a car/plane/spaceship/browser/etc I've designed!'. Fine, I'll be impressed when you build me one.

Ideas (theory) are cheap. So is talk. Posting to forums even cheaper.

Cooper built systems based on his ideas. For that matter, so did Joel.

AJS
Friday, January 23, 2004

Hmmm...

So if I was a doctor, and I wrote a book of specific instructions of how to perform specific procedures, that would be "theory"?

And if I wrote an inspiring narrative about how many lives I saved doing surgery that dedicated about 4 pages to the actual techniques used, with the rest of the book dedicated to talking about why surgery is important and why only people like me should do it (leaning heavily on why psychiatrists SHOULDN'T do it), that would be "putting theory into practice"?

Cooper is awesome. But in inmates he never gives enough detail about about the process of persona creation (the core technique Cooper uses in his work) to enable you to do it yourself. He acknowledges this readily in his speaking engagements, and has been promising a detailed follow-up book to Inmates for the last 5 years. In the meantime, people trying to use these techniques in the field have to pretty much make it up as they go along.

j b
Friday, January 23, 2004

Anyone read both editions? (About Face vs. About Face 2?) Are there substantive differences? Does he address web design issues at all in 2?

j b
Friday, January 23, 2004

Granted, I read AB about 5-6 years ago and AB2 last summer so take this with a grain of salt, but...

depends on what you mean by 'substantive' ;)

Yes, it's a different book, not just an attempt to make AB more up to date. The main thrust of AB was that the UI should reflect the user's mental model not the software's implementation model. The main thrust of AB2 is if you are going to design the UI to reflect the user's mental model, you've got to understand what you mean by 'user' and talks a great deal about personas, types of personas, the importance of digging through their different motivations, etc. There are definitely common threads through both books; mental/implementation model, the different software postures (sovereign, transient, etc.) and how they affect UI design etc.

As for Web Design, he touches on it but certainly not in any kind of thorough way. IIRC, there is a small section devoted to Web UI Design, but other than that it's mainly some "of course, when designing for the web, things are different" qualifiers.

I enjoyed all three books and do think Cooper has some very useful things to say about UI design. The only real problem I have is, in all his discussion about mental models vs. implementation models, he seems to neglect that there's a model that supersedes both (and that the design process needs to consider as primary), a "canonical" model I guess I'd call it; a representation of the real world things/interactions that make up the problem space.

herb
Friday, January 23, 2004

I didn't read version 1 but I did read all the comments on amazon about version 1's missing bits. Version 2 seems to specifically address the complaints on amazon. Chapter 37 is "Designing for the Web". Chapter 38 is "Designing for Embedded Systems". An afterword ties it all into the asylum book, and there is a full index and bibliography.

Dennis Atkins
Friday, January 23, 2004

"About Face" is a really good, practical book. I highly recommend it.

"Inmates" is a cheap rehash of the same ideas, aimed at non-technical people. Don't waste your time or money with "Inmates".

runtime
Friday, January 23, 2004

If Dennis Atkins thought the Amazon reviews were funny, I guess he'd find this thread hilarious.

'Inmates' isn't a programming book, nor is it aimed at programmers. That's the point.

Comparing doctors to developers doesn't work. Most developers couldn't even find the bathroom, let alone wash their hands properly.

I remember VBPJ had an article on basic algorithms once. A follow-up letter from a developer thanked them profusely for it. I think your average doctor is a bit more clued up than this. Lawsuits help.

The worst design book I encountered compared programmers with artists. He theory was that programmers should use similar methods, ie draw an outline (framework), fill it in. Fair enough, until he said forget about the details. Painters do buttons as a blob of paint (it saves time), why not code like that? Who need bounds checking, error handling, asserts, etc.

Will programming become paint-by-numbers? Will developers become licenced? Will it make a difference?

AJS
Saturday, January 24, 2004

The amazon reviews are two opposing views stated separately and not a back and forth so its different. I was interested that the book was controversial enough to arouse such polarized feelings.

Some complained because he uses big words like Kafkaesque. Well, if you aren't intimately familiar with that word you probably have limited programmin experience. :)

Others complained that he makes up his own terms. But when he does so, he explains 'I am making up a new term. I am not using the other term because it doesn't do what I want. This new term means specifically ABC." So I don't have a problem with it. A glossary would be nice but the index suffices. Of course you can sort of pick up the meanings through context too if you missed the initial descriptions.

Others complain that he is down on programmers. Not true. he is down on programmers who make bad interfaces that frustrate their users and waste their time. The fact that making such an interface is incredibly hard is why he suggests that only specialists do so. Of course there are the rare developers like Joel who can do both. But he's got a point that its so hard to do it suggests the use of a specialist. Me, I do both and I still agree with him.

One thing I like about the book is he goes over UI elements and explains what's good and bad about them very specifically. Example: Balloon Help was good because of ABC but flawed because of DEF. ToolTips are similar but better because of GHI. That sort of thing. Maybe I don't agree with all the details or thing there are other things that are a factor. But still, his reasoning is well thought out and worth thinking about. I guess I don't see him as telling you what to do with your life but making sure you are aware of all the different issues. I've always felt this way about people teaching. Maybe they are not 100% correct about everything. Maybe the world from their mouth are not the words of a supernatural, infallible, always right superbeing. Maybe instead they are letting you know about the ideas and thought they've had so you can start from a base a little bit higher, having had some fat to chew and things to think about.

One example of his controversial issues is he points out that file systems are bad from a UI standpoint. Hard to accept since file system is an implementation detail that is so in our face, but he is right. Users don't understand the file system. Trees? Recursive deeply nested directories? No way Jose. Documents are either in "My Documents" or they are on the Desktop or they don't exist. "The computer ate my file." is what happens if a file goes outside "My Documents". Extremely popular software admits to this. In iTunes, you don't ever need to know where your music files are stored. In Outlook, you don't ever need to know where your mail files are stored. In iPhoto, you don't ever need to know where your digital camera pictures are stored. It all gets put in a database front end. Software that forces you to deal with the file system is substantially less popular and useful.

Dennis Atkins
Saturday, January 24, 2004

That is certainly true, oddball pieces of software that force you to use the filesystem, say MS Word, are clearly less popular than iTunes.

Saturday, January 24, 2004

In what way does Word force you to deal with the file system?

Almost every 'regular user' of Word I know of keeps all their created word documents in My Documents.

Dennis Atkins
Saturday, January 24, 2004

And every professional user of word I've ever met doesn't.

You point is?

Or are "symbolic link" directories now not part of the filesystem?

Saturday, January 24, 2004

I think you're making a great point that you are not the guy to put in charge of developing an interface. Safe to say you didn't care for About Face 2.

Dennis Atkins
Saturday, January 24, 2004

?

MS Word, and the office apps consider My Documents to be just the same as any other directory in the filesystem, as far as I'm aware.

Most people would consider MS Word (et al) to be fairly successful.

You might even assume MS had done some useability testing.

But what would they know, right?

Sunday, January 25, 2004

The point of all this is that most users don't understand the file system. So, when designing a program to be used by most users, the program will be more comprehensible to the largest number of users if they do not have to have any awareness of the file system.

Such as the idea that your files are in a folder somewhere.

Programs like Outlook do not expose the file system to the user. Sure, the 'advanced' users like ourselves know where everything is kept, but the common user does not know and should not need to know anything about the implementation detail of files being stored in folders. It's not just that they don't know abut file allocation tables and sectors and blocks, it's that they don't get the whole infinitely nestable tree idea.

Likewise, iTunes and iPhoto and even my IDE reduce the need to deal with knowledge of files and their location within the system. The file system is an implementation detail that most users don't deal with well and even advanced users have trouble with. Where is my file from last year? Where are the photos I took in 1997? Dang if I know - probably backed up to a CD somewhere. Or maybe they are on one of these hard drives if only I could remember the name of them.

My Documents is one way of having a safe default that matches the ordinary user. Longhorn supposedly will move beyond this, replacing the entire file system metaphor with a database metaphor instead. There is a reason why things are moving in this direction.

Dennis Atkins
Sunday, January 25, 2004

The Palm Pilot probably owed its success to lack of a file system. Gets along quite well without 'save \ save as' dialogs too. (On hardware that is. The PC software is horrid)

There's no need to educate people regarding the difference between a file in ram and one on disk, and where on the disk it might be.

Those who insist you do should not be designing systems. That's you, -blank-.

Eliminating the file system is not a new idea. Raskin's 'The Humane Interface' discusses the Canon Cat word processor that lacked a file system. When users went to other machines, they wondered 'what's this file/directory crap?'.

AJS
Sunday, January 25, 2004

So going back to the topic... would you guys suggest reading both About Face and About Face 2.0? (this qiestion is only for those who read both)

Gamut
Sunday, January 25, 2004

I agree that the idea of file systems and directories is unintuitive. I am not sure it is possible to substitute it. I tend to feel that file systems are the only piece of theory that should be taught.

Stephen Jones
Monday, January 26, 2004

Can't comment on About Face v2.0, but the original is superb. Get 'Inmates' as well.

Recommended are Donald Normans books (Psychology of Everyday Things), 'Humane Interface' by Raskin (Mac designer), and a couple of books on graphics design can't hurt, eg 'Non-designer Web Book' by Williams & Tollet.

'Humane Interface' covers Mac vs Windows, points out little flaws like you should be able to activate the Start menu by shoving the mouse to the bottom left corner and clicking. Instead, you need to come back a little way, meaning a little bit of thought is required, slowing things down. Google for 'mile-high menus', same thing.

I'm not a Mac user, so these bits are interesting. Windows 'almost' gets there, in the case of the Start button, Raskin comments 'missed by a few pixels'.

AJS
Monday, January 26, 2004