ARCast.net - Why Software Sucks with David Platt

Announcer: It's Thursday, June 21, 2007, and you're listening to ARCast.

Ron Jacobs: Hey, welcome back to ARCast, friends. This is your host Ron Jacobs, coming to you today from Tech-Ed 2007 in Orlando, where I recorded a session live with David Platt at the Virtual Tech-Ed stage.

So there were people kind of wandering by. There was even somebody sitting right in front of the stage talking on their cell phone the whole time! And you know, just kind of ignoring everything that's happening on the stage, but hey, this is what happens when you record in that kind of environment.

But we were talking about why software sucks. And whenever we talk about this people start to notice, you know, people start to notice, they tune in and listen. Of course it helps that David Platt is probably one of the funniest guys on the planet, at least for a software guy, you know, and so we had a great time talking about it. We've talked about this before on ARCast TV, but if you haven't heard it it's a great book and David's book, "Why Software Sucks." You ought to get it. Let's welcome David Platt!

[applause, whooping]

David Platt: How you doing', man?

Ron: I'm doing' well.

David: You having' a good Tech-Ed?

Ron: I'm having' a great Tech-Ed, it's really fun to be here with so many people, like what is there, 13, 14,000, something, I heard?

David: Something like that. You can't fit them into just any convention centre, you know?

Ron: Yeah, that's right and that's why we keep coming back here to Orlando because this place is gigantic.

David: Yes.

Ron: And it's wonderful, and nice to be here, although I did not go to Disney World, did you?

David: Not yet.

Ron: No. Are you going to go?

David: I don't know, I was thinking staying Friday through the Shuttle launch.

Ron: Ah. Oh yeah, that's right, I saw that, it's going to be Friday evening, yeah.

David: Yeah, but I got to ship out on Sunday for Chicago and I hate to miss my daughters on Saturday, you know it's not...

Ron: Yeah, yeah, I was thinking' about that too. OK. Anyway, so that's enough small talk.

David: We're here to talk about software, aren't we?

Ron: That's right. so we're going to talk about today... you wrote this interesting book - in fact we had you on ARCast TV right after it was released, called "Why Software Sucks."

David: That's right, that's right. That was last October in Half-Moon Bay, wasn't it?

Ron: That's right, it was a very nice place; in fact if anybody wants to see the video of us standing with the ocean in the background on a beautiful sunny day it's right there.

David: Wasn't it a beautiful one? I love that.

Ron: [laughs] So how has the reaction to this book been?

David: Well, it's funny. It splits about 75-25. Three-quarters of the people say "wow, that's absolutely fabulous, Platt - you give a voice to our fears, our pain, you finally validated our feelings, you... hale-bloody-lujah, it's finally time somebody respectable" - I'm not respectable, but "somebody knowledgeable came out and said this stuff and affirmed that it's not just us being stupid." Right? The other 25% are from geeks who are absolutely furious that I've ripped the lid off their sausage factory.

Ron: [laughs]

David: And they say, you know, "Platt, you ought to be, need to be, taken out and shot, slowly and painfully." Of course, if it was hard to write it should be hard to use.

Ron: Yeah.

David: So that only those who have proven themselves worthy through intellectual toil will be able to benefit from the fruits of our labors. And 20 years ago that was not an unreasonable attitude when all users were programmers themselves. But when you see how the user population is changing just the last five years, even, you know it's no longer acceptable.

Ron: You know what's funny, is that that attitude... actually runs quite deep in the geek community.

David: It does.

Ron: I can recall once, when I was proposing a feature that would have been in a Microsoft product,

David: Yeah?

Ron: And this feature was actually intended for other geeks to use, but - it had potential for being misused, and so they...

David: Any powerful feature is like that...

Ron: And so they said, well - but it would have made life easier for a lot of people.

David: Hmm.

Ron: Right? And so they said, well, look: we only want people who really know what they're doing to do this, so we're not going to put that easy feature in there because... people who don't really know what they're doing might hurt themselves with it, so they left it out.

[laughs]

David: That was probably not an unreasonable thing in a consumer product. I mean...

Ron: Yeah, but it wasn't a consumer product, though it was a technology intended for developers.

David: Ah, well that's a little different, I mean developers are a specialized market and when I say, well, no, programs need to have... fewer obscure features that are of use only to a tiny fraction of a minority and it's much more important to make the main case run smoothly, developers get extremely angry. They're afraid that I'm going to take away the customization from THEIR programs; I don't want to, I'm not talking about them. I'm talking about the... they don't realize that they are NOW a tiny minority. Twenty years ago we were not, but now... the developers, us, you, me, all these guys here in Tech-Ed are a tiny minority of the people that actually use computers.

Ron: Yeah.

David: There were, back in 1994, there were about two billion users of the World Wide Web - ah, sorry, two million users of the World Wide Web; it was this thing for geeky academics. And that was the watershed year in the web, and then there got to be over 10,000 servers in 1994.

Ron: Wow.

And if you look now there are over a billion users of the World Wide Web... if the trend goes on at current rates there are going to be over a billion MySpace accounts by the end of the year.

Ron: [laughs]

David: Even accounting for duplicates, so there are over a billion users of the World Wide Web today. And China and India haven't really come online yet, and when they come online with a billion people apiece there are going to be a whole lot more. So there's this huge, huge increase in the number of users, and in the users that don't have a technological background.

For example my wife has a business selling knitted crafts, scarves, hats, things like that.

Ron: Yeah.

David: And in Massachusetts, to pay her sales tax she has to do it online.

Ron: Yeah.

David: She can't mail a check, they won't take a check, you know, and it's you have to do it online because it costs the state that much less to do it and so she is forced to use the web. And so we have ordinary non-technological people - you cannot get any less technological than my wife, you know -- doing and using the web because that's kind of the way the world works today.

And so we've got this huge new user community that -- I don't think that the developers understand that, they don't even acknowledge that it exists let alone figure out how to take care of them yet.

Ron: So, give me a couple of examples: you cite in your book some examples of... just really bad ideas, so let's just talk about a couple of them.

David: All right. Well, since this is Microsoft let's talk about Office.

Ron: OK.

David: In Office, if you reached up to click the File menu...

Ron: Yeah?

David: And if you overshoot by a couple of pixels then... usually in the morning when I've had too much coffee... and you wind up dragging the menu bar around on the screen.

Ron: Right.

And I've never met anyone, not even one person, not even my friends on the Office team that has done that intentionally, because they wanted to move the menu bar. There is, you know, some sexually frustrated geek with nothing better to do on a lonely Saturday night, who's geeky to get a date. So he went to the lab, and he wrote some code, and difficult code that the other geeks would respect as "Wow, man that's cool! You wrote some nice code, good job."

Ron: Yeah, in fact they call that the cool bar!

David: The cool bar, do they? I don't call it the cool bar. All right, maybe the geeks who wrote it call it the cool bar. Yeah, OK, I guess that wouldn't surprise me if they did, and again it just goes to show the difference in mindset.

But it's actually very bad from the user's standpoint because it forces a user to be more precise: it forces the user to be more like a computer. You know, we're not precise animals, that's why we build computers in the first place.

Ron: Yeah.

David: We're good at things, we do things, people can do things. Like a newborn infant can recognize faces, which computers can't do even today, no matter with gazillions of dollars poured into that. And a newborn infant can do it in two or three weeks.

Computers are good at some things and humans are good at other things. That particular feature forces the user to become more like a computer. It says, "If you do not become more precise, we will punish you by doing things that you do not want done." At some point, I do not know what version that came out in, but there was a version without it. Then there was an upgrade that had that feature. Then all of a sudden, the users had to get more precise. And that's exactly how we should not be driving the industry.

On the other hand, let's take a good feature from Word, and that is their automatic correction. I think a whole faster that I type. It's not that I think all that fast. It's just that I type really slow.

So, I got these ideas in my head and I am trying to get them down and typing them as fast as I can. Trying to get them down before the words flee, and I will type -- the one hand is faster than the other -- I will type "hte" instead of "the". I will type h-t-e space, and Word looks at it and says, "OK. Buzzy, that's cool. We know what you mean, no problem. And automatically just switches the "hte" into "the." That's cool.

Again, human beings are not precise animals, and I'm not being precise here, but the computers are being precise on my behalf. So, the computer is being precise, it's being diligent and it's being thorough, so that I don't have to. The computer is doing what it is good at, so the human being can do what the human being is good at. So, the computer is being a computer and allowing the human being to be even more human.

So, that first feature - does not recognize, or understand humanity, and tries to force the user to become something other than human. Whereas, the second feature - understands, and respects, and enhances the user's humanity. And that is a brilliant idea. If you don't think it is not a brilliant idea, try working in Word for a while, and then, switch back to FrontPage, or something, that does not have that feature.

It drives you absolutely bonkers. I get to the end of a page and I have got 25 typos. I have to stop and go back, fix them. It breaks my flow completely. I have no idea how I dependent I was on this thing, until and until, I got taken away. There is an example, a good one, recognizing, understanding, enhancing the humanity of the user.

Ron: And what is interesting about that is that early on, we had spell checkers. And the spell checkers would just tell you, "You have spelled that wrong, " and maybe even highlight it in red. You think, here is a great option because it could interpret the meaning and get it to you in the right place instead of just nagging you about - you're a lousy speller.

David: It is much better that way. It is not just the implementation that take this word and compare to a list of words that I know about, and then, underline it with red if it is wrong. That is OK. But what does the guy really want? And let's see how close we can get to that. It is thinking not just about how the program is implemented internally, but who the user is and what the user wants, and why the user is using this program, why the user paid his money to you for that program. We need to see a whole lot more of that. And less of the first feature.

Ron: OK. So there's Office. But then, I'd like to think about Web applications. There are so many people here writing Web applications and they are doing a lot of things that just people find frustrating.

David: I know exactly what you mean. A classic example, I'm sorry, but they're not live on video, but I'll be showing this when I talk on Thursday morning - tomorrow morning. I was in Sweden and I went to Google.com. The Google server sniffed the IP address, figured out that I was coming from Sweden, and bang, up came Google in Swedish. They knew I was coming from Sweden, and they know that all human beings like to be spoken to in their mother tongue. Every human being does.

So they said, "OK. This guy is coming from Sweden. It is more likely than anything else that his mother tongue is Swedish, and that he would like Swedish. So we are just going to give him Swedish." And they did.

Suppose they were wrong and in case they made an error or in case the person from Sweden, was a Swedish student wanted to practice his English. Or a guy on the road like me who does not speak Swedish. So they had a link at the bottom that says, click here for Google in English. So, they made the best guess for most of the people, most of the time -- hey, how are you doing? -- and they had a link.

So it just takes one click to fix it permanently. They set a persistent cookie permanently, if it is ever wrong. And that is very good. Again, thinking about who the user is, and making the best guess as to what the user probably wants and then, making it very easy to fix, just in case they got that wrong.

There is one for a guy from Canada. The button says, "I'm feeling luck, aye."

Ron: [laughs]

David: But if you go to the ups.com website, the first thing that you have to do is you have to select your country. They will not talk to you. They will not do anything at all until you select your country. You can't track a package. You can't schedule a pickup. You can't find your "by locations". You can't read their annual report. You can't do one darned thing until you drop down the box and select your country. This takes a lot of clicks. If you're in Sweden, it takes 30 clicks to do that.

And then you know what else? After you have made that selection, if you want them to remember it - you say, I am in Sweden. They say, well, OK. Come on, I want you to remember it. Remember I am in Sweden, darn it! You have to check, specifically check a box that says, "Yes, remember this please." So, what they are doing, is they are forcing the user to understand how they have written the computer program. And forcing the user to do things in a way that makes sense for their computer program.

Instead of Google putting themselves in the user shoes, and doing their best to make things as easy as they possibly can for the user, ups.com is saying, "Hey! You exist for our benefit. You shall do what we tell you to do in the order that we tell you to do it, and these are orders. Thou shalt do the things that we say." Would you accept that kind of guff from a human user? No. You will punch the guys lights out in the first five seconds.

Ron: [laughs]

David: Why do we meekly accept these from computers? We don't have to and we shouldn't.

Ron: Well, you know what occurred to me is that we're entering an era where, more and more, we are doing business with companies solely through the web.

David: Very true. You're telling about a bank the other day.

Ron: Yeah, I have this bank. I have never met them. I have never been into one of their buildings. Never talked to a human being from them.

David: Yeah, Fidelity. I haven't talked to a human being in years. They are just doing their online stuff and it's actually not bad.

Ron: So, I have that one bank that I really love their web site. But then, my other bank, which I use a lot more, and they have a little building right here near my house; this is probably why I end up with them. Their web site stinks. It's absolutely hell. One of the simplest things that they do, drives me batty is that you'll be going into something, kind of drilling into your account or a transaction or something. If you hit the back button on your browser, it says, "Web page expired." And then, if you say refresh, IE shows you... But to us, we understand, but to a normal person, it looks very scary; like, "Don't press this retry button if you're buying something, or you might buy it twice, or whatever."

David: Yes, yes. I know what you are talking about.

Ron: That's lame.

David: I quit my bank, because they tried to force me into that nonsense. Again, they were trying to force the user to do what they wanted the user to do. Human beings are pattern oriented animals. We learn one pattern that works, and we follow the same pattern over and over again. I used Fleet Bank. They had the same thing. You can't use the back button of the browser to go back a page because that would break their security. If you use the back button, they will lock you out for 10 minutes or so.

"Sorry, it is your job to be more precise to use our software." "No, it is not my job, and I will take my money to somebody that wants it. No, I do not exist for your purposes. Get that through your head. Me customer. You vendor. Me have money, you want. You be nice to me, or else."

Ron: That's reminds me of...

David: Stop me if I'm too technical here.

[laughter]

Ron: That's reminds me of -- I had an incident where I called the customer support for this company on the web -- where I had to pay for a year's subscription to do something.

David: Yeah, all right.

Ron: It was for Microsoft and I was trying to pay with a purchase order, which completely threw them off because they were only set up to deal with credit cards.

David: For what it's worth, I'm a small business owner and I get stuck that way myself. It's just trouble to take PO's.

Ron: Yeah. So, I got in touch with a live human being. They said, "Yes, we can do it." But it kind of threw them off so bad that it took about four or five weeks, from the time when I signed up to the time when my account actually worked.

David: That's crazy.

Ron: And so, I called up their customer service and I said, "You know, since I wasn't able to use my account for the first four weeks, can we just make it start today because that's when it's finally working?"

They said, "Well. Sir, no we can't."

I said, "Why not?"

They said, "Because the system will not allow me to change the start date."

I said, "No. You see, we're the humans, they're the computers. We're in charge, not them." [laughs]

David: Great, great.

Ron: But they never did fix it.

David: I tell my daughters, when the cats are chasing my daughters around the house. "Hey, you got to know. Humans are in charge. The cats are not in charge."

Ron: Yeah.

David: When my kids are, when they are out and it bugs me, "Wait a minute. You stop. We are the grownups. The grownups are in charge, the children are not in charge." The cats don't pay any attention, the kids don't pay any attention and the ops aren't paying any attention either. But you are right.

We created them. You know what this software business is? This is human kind creating it in its own image.

Ron: Yeah, yeah.

David: I wish our image was a little bit nicer.

Ron: OK. So we could moan about this all day, and it's really fun to do but...

David: But there's so many bad examples out there, it's not even sporting. It's like dynamiting fish.

[laughter]

Ron: OK. So what's your advice to developers, architects -- what are you telling them to do?

David: The number one thing I tell them to do is -- the number one problem they have is they're out of touch with their user base.

Ron: Yeah.

David: What you need on a design team is you need a virgin on the design team, by which I mean somebody -- not the traditional meaning of the term -- but somebody who does not know how your application is constructed internally.

A classic example is just opening Notepad, typing in some characters, you go to close it and Notepad says, "Oops. Do you want to save the changes?" Or, opening a document that already exists, making a couple of changes, you click close, and it says, "Do you want to save changes?"

What's happening there is, obviously, you read the program in, write the changes in buffer, and before the buffer gets written back to disk, the program asks, do you want to do this or not. What that is doing, is that is forcing the user to learn and understand how the program has been written. Nobody thought about whether the user is going to like this or not. They just took the internal implementation model and exposed it directly to the user.

Now, the reason that they did that is because that's how the program is written internally. If you had a person on the design team, that had some authority, so people would have to listen to him, he would say, "Why are you doing this? This is not what the user is thinking. The user is not thinking whether they want to save this or not. The user did something. The user is expecting it to be there. If they don't want it to be there, they'll throw it away."

A really good example of this is Microsoft Windows. People start using it and they say, "How do you save?" And the answer is you don't have to save. It's just done. If you want to throw something away, you have to do that explicitly. And, you can do that, you can undo, undo, undo. That works fine. They are not just constantly asking you if you want to save, save, save.

So, what you need is, you need somebody who does not know how the program is written internally. So, who cannot possibly be easily influenced that this can be easy to program and that is hard to program. And, can represent only the user to say, "Gosh, the user finds this easy, the user finds that hard." Therefore, do the things that the user would like.

A classic example is Carbonite, the backup service that I use, the Internet back up service. When you sign up for a Carbonite account, they ask you what files do you want to back up on your PC. On the PC, you can choose the contents of the My Documents folder, you can choose all of My Data, where ever it is on the disk. Or, the third choice is, you can say pick and choose. Choose to back up this file and not that one, and so on.

The President of Carbonite told me no one has ever chosen that third one, ever in the history of the world.

Ron: [laughter]

David: But the geeks insisted on putting that one in, they wouldn't work on the project until they were allowed to put that in.

Ron: Wow.

David: So, that's what geeks do, they have control. And so, what we need is somebody to say, "Hey, look man. Show me a customer, show me somebody who's going to pay me 50 bucks a year who's going to use that. I don't care what you think, you're not paying me!"

Ron: Yeah!

David: This is what our users want: somebody to represent the user, somebody who does not know how the program is built internally, who can represent only the user experience and nothing else. That's the number one thing that you need.

Ron: OK. I think that is really critical, that when we focus on a user experience -- I've talked a lot about this on ARCast in various episodes -- because it occurred to me, the way you feel annoyed when you go to a business and they're ignoring you...

David: Yes, yes.

Ron: Or, you're in a restaurant, you can't get the waiters attention, things like that. It's a negative user experience. We have those all the time with the computer, but somehow we've come to just say, "Well, that's the best it's going to get."

David: Absolutely. And that's what I try to say in my book here, "Why Software Sucks." It does not have to be that way. It's that way because of design decisions made by the guys who built it. Maybe they were programmers, maybe they were project managers with deliberate design decisions, and had they made different design decisions they could have given us a better user experience.

It's no harder, to write good software, than it is, to write bad software. Do you do any cooking in your house?

Ron: Oh, yeah.

David: OK. Then, agree or disagree, if you would, with the following statement: it is no harder to make good food than it is to make bad food.

Ron: I would agree, yeah.

David: It's the same thing here. If the programmer geeks say, "Gosh, we can't make this easy to use." You know what I come back to them and say, "You're telling me you're not smart enough to do that?"

Ron: [laughter]

David: And it gets them, bang, right between wind and water.

Ron: Ah!

David: You give them a challenge like that and they can't turn it down.

Ron: See, there you go. I think a part of it is persuading the management. They go, "Well, my boss will never pay for us to think about that, or, having a designer come in, or an interaction designer help us." How do you advise them?

David: Well, that's a good one. I do see that a lot. They say, come on, you're the software geek, you use computers all day everyday, you ought to know what a user wants. That's like saying, you get a root canal, so that makes you qualified to give a root canal.

Ron: [laughter]

David: That analogy often gets to them. The other thing I say is, the programmer is a programmer because they are good at having a conversation with a silicon chip, mediated by a language compiler. They use their interface design, as having a conversation with a human being, mediated by a graphics package. There are no two more things on God's good earth that are more different than a silicon chip and a human being. In order to be good at communicating with one is always mutually exclusive that you are not good at communicating with the other.

The three things I say for convincing the boss, is - number one, you show them constant friction. An extra click here and a click there and a click there and a click here. Just one extra user interface click times a thousand PCs once an hour times a thousand PCs at $50 per hour employee time is $125,000 per year.

Ron: Wow.

David: It is a hell of a lot of money. I did that calculation for a hospital in Boston. It adds up a whole lot faster than you think, number one.

Number two is the problem of catastrophic error if the user interface is bad. You go to cut and paste something. You thought you cut but you didn't. You thought you pasted but you didn't. And so you end up amputating the wrong guy's leg. Catastrophic error. This is a problem. And so, a better user interface can have fewer of these catastrophic errors.

And the third thing you can offer your boss is public humiliation by me. You show them Vincent Flanders' web page, "Web Pages That Suck." Or my blog, Suckbusters.com, and you say, "You want to be there by name? Do you want your picture saying, 'Hey, this guy is stupid!'" That often gets their attention.

The other thing that a boss sometimes says when you try to make an application simpler to use, is they say, "Well, gosh. You're going to have fewer features here. How the hell am I going to sell something with fewer features. Nobody's going to buy something with fewer features."

Well, you know, that's like buying your clothing by the pound. At one time it might have made a certain amount of sense. I show them the picture of the Carbonite Web Backup Service. They were... You know what their feature list is? They show a guy lying in a hammock drinking a beer, and say, "Hey, pay us $50, and you can scratch this off the list of things to do for the rest of your life."

That is, to my mind, the ultimate feature. Something that means that you don't have to touch it again. That is... That, to my mind, convinces the boss, convinces the...

Ron: Well, you know, you have a point there about the features. I remember the Office guys were saying about Office 2007 that when they went out and were doing their early design work, they asked users, "What features would you like to have in Word or Excel?"

They found that half the things that people asked for were already in the product. They just didn't know it.

David: They couldn't find it. Yeah, I know what you mean. Even there. On one hand, they have made things more visible with the ribbon. That represents a lot of thoughts. They said that something is down on their menu item. You don't really know what it is. You don't know how it is categorized. You won't see it until you pop up the dialog box, and nobody ever does. So, they're going to make them more visible on the ribbon.

On the one hand, that's good. On the other hand, I use the print button a couple of times an hour, and I don't know where the hell it got to. Just click the toolbar print button and print one copy on the default printer. Either they took it away, which they might have, or they hid it so deeply that I can't find it, which to me the two situations are indistinguishable.

Right now I have to click on the little disk icon, go to the print menu, and pick print, just quick print on there. It is four or five clicks when it used to be one. This is something I do frequently, so there's a piece there with which I have a serious disagreement.

Ron: Isn't that ironic that they call it quick print when it takes four clicks to get to it?

David: It used to be a lot quicker, and if you think I'm going to call it advanced, you're wrong. Now, if they have data that says, "No, Platt, see you're unusual, because most people don't even print once a day. Here's the customization feature. You put that in if you really give a damn, but nobody else wants it. Here's the data to prove it." Well, yeah. OK. I'd have less of a problem with that.

You know what I would really like to see? You know the Microsoft they call the Customer Experience Enhancement Program, or whatever it is? It sends messages back over the web saying, "The user's doing this. The user's doing that." I really love that thought, because now you will have hard data that says, "Over time..."

If you do a focus group, or a run through, you will find... Get a decent handle of a user's first experiences. And a lot of times those are important, particularly on websites. The user's first experience says whether there is even going to be a second experience. If you get that wrong, it doesn't matter what the second experience would have been.

When you have an application like Word that you're going to use every day for a long time, their experience over time becomes more important. Now they will be able to measure that. So, OK, the user... For example, the insert overtype. You hit the insert key, and you go into overtype mode. "Oh, damn. I didn't mean that." And you Alt-Backspace to undo it, and hit the insert key to go back into insert mode again.

I have never... You guys in the audience, has anybody ever done that on purpose? Going into overtype mode because they wanted to overtype data in one of your programs? One guy did. Maybe one guy. OK. And this is tech, it is full of geeks, and even with a bunch of geeks only one guy, and he's even saying only half the time.

But, if I had this service, I would be able to see definitively that yes, people do use this feature, or no, they don't. Or the insert is always followed by an undo and then go back into over... Going into overtype mode is always followed by an undo and then going back into insert mode. I would be able to definitively prove that yes, this is an error. No, it's not an error. Then I would have the hard data that I need to be able to say that "Yes, they like this, " or, "No, they don't."

It is very hard to get that hard data out of a user. If you ask them, they might not remember. They might try and be nice and tell you what they think you want to hear. Maybe they don't want to admit to having trouble, because they don't want to appear stupid. It's a...

[pause]

A fair number of people are uncomfortable with Microsoft seeing what they had. Now, of course there's a privacy statement that says, "Yes, we promise to only use this for good and not for evil." I wonder how many subpoenas that is likely to stand.

At the same time, yes, that is an attitude that a significant number of people are going to have. But, if you get even five percent of users turning it on, often because they don't even recognize that they should turn it off, because they never read privacy statements. Nobody ever does.

If you only get five percent of users, you will get a whole lot more information than you had before. What I would really like to see Microsoft do is to offer that as a commercial service that third party vendors could buy. So instead of them having to develop their own customer, user interface tracking kinds of things, they could just buy this service from Microsoft. It would be a whole lot easier than rolling it themselves, and possibly getting it wrong. That would be a fabulous innovation.

Ron: That's a great idea.

David: I've been trying to find the right guy. Tell everybody I know at Microsoft about this, and I haven't gotten it through anybody's head yet.

Ron: Well, OK. We've got to wrap. But let me just say, if you have not read "Why Software Sucks..." I know people think, "Oh, that's a book for end users," or whatever, but I think developers should read it, and think about it. It's a hilarious book. You'll really enjoy it. It's an easy read. I highly recommend it.

Thank's so much, David, for being on ARCast.

David: Thanks for having me here, Ron.

[cheers]

Ron: David Platt, ladies and gentlemen, with "Why Software Sucks." It's a great book. You really ought to go out and get it and read it. Like I said, it's really fun, too. I read the whole thing on the flight back from Half Moon Bay last October, and just enjoyed it tremendously.

He's got a lot of good things to say. Because we're geeks we've got to remember we're computer geeks. We know how all these things work. We know how they think. We know how programs operate. And that's why sometimes we build things that seem natural to us. But when the world uses them, it just sucks.

[laughs]

And, you know, I think most people are pretty frustrated with their computer, so it's part of our problem to deal with. We've got to do something. We've got to make it better. You can do it. I know you can. Hey, we'll see you next time on ARCast.

Announcer: ARCast TV is a production of the Microsoft Architecture Strategy Team, www.arcast.tv.