Book Comments On ``Interaction Design''

2009-10-31.

This book walks through the history of design activities in the computing history, with regard to both hardware design and software design. It presents a serial of great designs, starting from the invention of mouse, desktop window system, mobile devices, games and internet services. You will find some grand names in the IT industry, XEROX, APPLE, PALM, GOOGLE. It also talks about the design methodology. It emphasizes the importance of usability study, quick prototyping iterations and techniques. This book covers an extensive range of electronics design, most of which you should already be familiar with. However, this book is too early published to include the iPhone, which I think is the finest design, state-of-the-art.

No doubt anybody interested in design history must read this colorful book. Here I am going to share some thoughts during my reading.

Good design shares the same trail of simplicity, and yet simplicity is hard to achieve. A typical technology goes through three phases: geek, professional, and consumer. Geeks care only about the ability to do unusual things, professional users also care about stability and reliability, and what consumers care more is the simplicity to use and the style.

Simplicity example. Bill Atkinson developed the menu concept in Mac. Menu offers a unique way to provide functionalities, easy for developers. But rely on Menu is dangerous. I always find that it is not pleasant to go through the menus to find where the settings are. You know, some applications put it under ``File'', some put it ``Tools'' (visual studio), some put in under ``Edit''. Office team in Microsoft did a great job to eliminate Menu from its UI, that is really wonderful, and it is really a big step to say goodbye to Menu. If you ever use Office 2010 preview, you will find that it takes a step farther to remove the print dialog. It embeds the print settings and print preview in one place. I must say the design takes the advantage of current popular widescreen display, in the same time makes the software very handful and usable. Dialogs are very annoying and intrusive. Even in Mac, there are still modal dialogs, the ones that would never let you go unless you click on OK. What the hell I have to do that? For example, when I remove a USB stick from the computer without eject the device first, the same dialog always jump out and say that I accidentally remove the storage device and ask me to click on ok to dismiss it, and it doesn't offer a chance to say that I am not going to see this kind of stupid thing again. Is this a punishment or something? I would rather like two alternative ways: one is let me select an option to say no more, the other is a fading out notice. A company like Apple which lives on its design style still makes this kind of stupid things. I think the future is really going to remove all dialog based interaction.

Simplicity brings performance. Google's homepage is very uniquely simple in those days when it is born. Most of search web sites have tons of Ads and flashing banners. But google is different. The fast page load is very attractive. You don't have to be raped by the gif animations and flash. It is always responsive even you are in china. Being responsive is very important. What lacks in Microsoft's live services is the quick responsiveness. I used to install a mail client to read mail. Most interestingly, although running in client side, mail clients like thunderbird or outlook express lack the speed in gmail, especially when you are searching for mail. Google changes our impression of web services which used to be slow, instead it sets a very good example that everything could be made simple and fast. Methodology in design. I think good design comes when doing it rather than talking in a conference room, so called brain storming. Brain storming usually ends in endless arguments. There is no conclusion in those sessions, usually you will find them aimless and merely a waste of time.

Design is a very personal process. You do it, and show it to others; you get feedbacks and made improvments, and finally you really have a good prototype to show to managements or investors. People usually show more interest when they can actually see it and play with it. Good ideas joins into your efforts when you are showing it or doing it, not at you talking at a virtual idea. Never talk your design before you make a prototype of it.

There must be more than one solutions. Usually in big company, there are many people working on one product, and only on one solution. This is bad in my opinion. I believe for any product, there must be more than one solutions, and we can't make a good design if we didn't explore possible ways. But management would say, we can't allocate enough resource to do so, and we have a very tight timeline. This is not a good excuse to ship a product that is not good enough. If a product is not good but shipped, the cost to maintain and update, the cost to push it out of production line, is even worse comparing to the investigation effort at the design phase. But business people just want to ship, they don't want to ship good product, they just want to make good money especially when the company is big they have great pressure from investors. that probably is that why small company can have more innovative designs than the big ones. And usually when a big company buys a innovative new company, the small company just die silently. If you really love your product, you have to stick to yourself rather than sold it and then get stuck.

This book, ``Design and Interaction'', also emphasizes the importance to combine early usability testing and iteration of prototyping. It also introduces some interesting prototyping techniques using cards, papers, sketches. It also points out sometimes designers rely on their intuition to do design subconciously. On these points I totally agree they are valid in most of the cases. Anyway, I think, there is no final principle in guiding design, which is such a really artistic actitivity. If a company or a designer have to strictly follow some guidelines or rules, they really couldn't make first rate product. I heard that statistic plays a great part in google web page design, especially in choosing colors. That is , come out serveral design solutions and let thousands of users select and pick the final decision. I think it is possibly good for some products, but you can't always rely on usability test. The designers have to be the user of their product, otherwise it could not succeed.In apple, probably Steve Jobs is the one to represent thousands of users' desire.

Users sometimes have no idea what they really want. Take an example, WYSIWYG is very popular and is believed to be the right way for editing. Users are very easily attracted to this approach at first, but is it really the best way to do work? Formatting is the least thing to consider when you are writing, but now it just takes too much time. Even formatting has been easy in WYSIWYG editors, it is rare that people can produce documents that are truely beautiful in this way. I am not suggesting that everybody should write TeX like documents, yet I believe there must be a better solution to help user express their thoughts in a graceful way, and unfortunately you can not learn it from user survey, but in your talents.

Return to the book itself, now that you read to here, you see this book can really help people think. Everybody has his own experience in designing software or hardware products, and can have his own thoughts inspired by this book.