Author: Roy Eassa
Date: 08:02:48 06/11/02
Go up one level in this thread
An interesting reply to the Editorial, at that same site: I study computer science at a university where Microsoft recruits heavily. I know many of the people they have hired from my university in the last couple years. I myself have been interviewed by Microsoft (and have a standing invitation for an on-site interview). In the course of their recruiting process, I have spoken with several Microsoft employees, particularly in the area of mobile computing. All of these people are very intelligent, bright, clever, hardworking individuals. I bear no ill will to them, nor do I think they are unqualified to be working on such advanced systems. I will also qualify my following statements with the caveat that I have spoken with developers and engineers only - I have not spoken with any usability experts or designers at Microsoft. So my comments may be wrong. What I find is that their development process takes usability as an afterthought -- something to work into a subsequent revision, after receiving customer feedback. Products are not tightly designed around customer needs. This happens because their development process is decentralized and executed by developers/engineers. Each project (PPC, for example), has a manager, who is usually a business person. He or she is in charge of a group that is largely composed of engineers who manage a particular component of the larger project (which can be as granular the data entry screen in a to do application, for example, or as large as a calculator program). The project manager gives a list of features in checklist form to the engineers managing each component. This is a problem because it results in little integration among components, and because it empowers individual engineers with the ability to design everything, including the user interface. Again, I believe that MS employees are very intelligent and capable people. However, they are not the right type of people to be designing user interfaces for their products because these people are expert computer users, who have developed a tolerance for user interfaces that waste time and are inconsistent. They do not see the obvious usability flaws in the interfaces they design; instead, they just accept things they way they are and think that such problems are normal. Compare the number of taps required to do something in on PPC vs. number of taps required on Palm. For example, compare number of taps to enter a new to do item, or to schedule something. (note: they are improving -- the difference has been closing in newer versions of PPC). To be fair, the PPC allows "richer" items -- on PPC, I can store the location of a scheduled appointment in a specially designated blank for the location, whereas on Palm, I need to include it in the generic text field for location. Some people may think that this means the Palm is limited. But I disagree -- examine a day planner, and you will not find a place to put location - just a blank line, waiting for you to fill it up. The additional value of having a special blank for a location is negligible when it comes at the cost of several additional taps required to schedule something. The to do application is another excellent example. On Palm, I need only push the to do button, and start writing in the graffiti area to add a new task. On PPC, I need to click the new button, and input the task into a number of fields. While these additional taps do not seem to be a problem, if you consider that most people hate managing their schedule, and could just use paper if they wanted to, each second shaved off the amount of time to do something is highly significant. (I'm so lazy that I know an additional tap to do something may be the difference between managing my time/tasks on Palm and doing it on paper or not doing it at all. I doubt I am unique.) Add in the time to retrieve the device from pocket/bag/etc, and turn it on and wait for the screen to be drawn. If the amount of time is too high, the device no longer saves you time, it wastes it, and there's really no reason to use it. Yet the people at Microsoft seem not to understand this; again, I believe that it is because the people driving these project are engineers and developers who are used to dealink with slow, unintuitive interfaces, and love computers so much that it makes up for any time they waste on them. To make things worse, they add usability experts in after their development cycle. Anyone with experience in this area knows that usability must be involved in the whole process -- its impact afterwards is minimal. Furthermore, they drive their revisions off of "feature checklists" given to them by customers who purchase their products. In many cases, these will be middle managers, who do not actually use the devices themselves, but instead have come up with a list of things required for the devices to be on the network. People rarely add "easy-to-use" to the list of features they need -- instead they add things like 'manage a computer from my handheld,' which results in ridiculous things like the Remote Desktop client, which requires users to use their 320x320 PPC screens to view their much larger desktop screens. I have heard that when MS-Word was still competing against WordPerfect and AmiPro, the Word group kept two charts on their wall: sales of WordPerfect vs Word, and a feature checklist of Word vs. WordPerfect. They kept working until they achieved feature parity with WordPerfect. While they may have come up with one or two novel features, the central driving force of their development process was their competition's feature checklist. Handheld computers must be trivial to use; the "killer feature" is the ability to help you quickly do something else in the real-world -- not be a "cool toy" which is useful to themselves (there are exceptions to this of course but this seems to be true most of the time). Using a feature checklist, having people who cannot differentiate between an excellent interface and a decent one, and considering usability as an afterthought do not result in the creation of this "killer feature." Palm understood this early on; the evolutionary progress of PalmOS (which of course would probably be better if it was faster) is a testament to their grasp of this. As long as Palm continues to add features without making their OS distracting and difficult, they will own the market.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.