On Thursday 29 May 2003 16:37, Herouth Maoz wrote: > Quoting Dan Armak <[EMAIL PROTECTED]>: > > But you only memorize the milking process once, right? > > Yes, but the point is, you perform the task itself over and over again. > It's not that people don't abstract, they just don't abstract in the way > you want them to.
Well, yes, in the same way that they would mark heading lines as style=heading1 over and over again - every time they added a new heading. What's wrong with that? > > > Well, by your logic, people could never generalize enough that the > > process of milking a cow would deal with the abstract notion of _a_ cow > > and not with specific live cows. > > It's not logic, it's observation, and the observation is that the average > person does not abstract in the way a computer science graduate does. You > can't find fault in my logic, because I was not trying to prove anything, I > was making a statement about how the minds of people work, based on an > unscientific observation. Actually, I'm not a CS graduate, not even close to it yet. Neither is any of my friends and yet I believe they can all generalize in the basic way I've described. I only know what _my_ observation is telling me, and that is, that the average person _is_ capable of learning to select 'heading1' from a drop-down box instead of selecting 17 for font size from another box. I haven't performed this specific 'experiment', but I've shown people how to do other comparable things and my personal belief is that many people _can_ learn to do things like this. Now of course it might be that the people I know are a totally unrepresentative section of the population. I can't really judge that, I can't say I know many very diverse people. But even if you're right and people don't think in this way, surely they can be taught to? Thirty years ago, if you offered the average person to have and use a computer, he'd probably say, "but computers are things mathematicians and programmers use. It would take me a long time to learn to use it and I probably don't have the skills for it anyway." I don't know first-hand because I'm too young, but I imagine most people must have thought this way. Well, today a very large portion of the population in Israel uses a computer, or at least has access to one. But people still don't use them efficiently, and they still need to learn and gain experience to use them even in that inefficient way. So we need to make the computers, ie programs, better and easier to use. And a very good step in that direction, is making people stop using (or rather, stop forcing people to use) badly designed, non-intuitive programs like msword. To an existing msword user, the lyx idiom is unintuitive because it's different from what he's used to, and you're right in that he doesn't want to spend time retraining and learning, he wants to get things done _somehow_. But to someone who's going to use _some_ word processor for the first time, its better to start with lyx. I don't believe the learning curve of lyx (with a good set of templates and document classes) in that situation is much worse than msword's. Now there aren't many people who're going to use a wordprocessor for the very first time, except for children. So in many cases we _will_ have to retrain people, because we can't afford to wait for the whole old generation to die :-) However, we should at least try to move in that direction on a case-by-case basis. In some cases lyx is not yet preferrable to msword, but in some it is. > > > So why can't it be "mark as style = heading3" every time? They don't > > really understand what they do or how it works in any case. It's just a > > matter of getting used to a different interface - and then there's a big > > saving of effort since everyone's output looks the same via a common > > document class, and noone spends time manually formatting fonts to reach > > the same result, or even thinking about fonts at all. > > You mean, it's just a matter of changing the way people percieve the world, > process the data they percieve, and make decisions according to that? I don't see why you say that. A new wordprocessor with a new interface isn't going to change the way someone sees the world. The way one sees word processing, at most. > > Easy, very easy. > > Basically computers are more easily programmable than people. Therefore, > change the computer program, not the person's brains - especially given > that your real expertise is computer programming, not psychology. True. And 99.9% of our effort and time as programmers does go into programming computers. But at some points we need to introduce truly new things, and to deprecate some truly bad old things, and then we need to educate the users about these changes. It's like that when moving someone from windows to linux (even if he'll never once see a cli), and it's like that when moving from msword-like word processors to lyx. If you can see a way to let people use good new interfaces and idioms without retraining them at all, please tell me, as I can't see it. Not something that works in 100% of the cases. > > > Tell the user: when you want a big heading, select 'heading1' from this > > list. When you want a small heading, select 'heading2'. When you want > > normal text, select 'standard'. Then when you print the file, it will > > magically all be correct. Just forget about the fonts. Don't worry about > > them. > > "Tell the user"? Do you mean that I have to clone myself for every program > I write (Assuming I'm the author of LaTeX or whatever), and kibitz over the > head of every user in the world, convincing him that he can think the way I > think? Believe me, it didn't even work with my mother and my sister, let > alone the less intelligent population of the earth which, despite being > less intelligent, still needs to use a word processor. Someone needs to tell Joe User how to use the basics of _any_ new program he uses. It can be a manual (printed or online), it can be a guided tutorial, or it can be a real person. And you don't need to tell him much beyond that paragraph (the Tell the user:... one). (To make him use styles, that is.) > > Using a word processor is more complicated than driving a scooter. You can > easily train a person to do things by rota - when you see that, you do > that. It's much much harder to train people to change they way they think, > rather than the way they behave. Well I _am_ talking about rote training, am I not? When you want a heading, click this item on this list. When you want a bigger heading, click this other item. If you rename the items to 'heading' and 'bigger heading' or whatever, I think it'll be easy enough. > > And asking the average joe to RTFM is not fair either - next time you see > an average joe read the manual of his car, I may be convinced that you can > convince the user to RTFM. People basically learn practical things - from > house cleaning to knitting to milking a cow - not from books but from > example and direct training, or by trial and error. Books are for > academics, not for everyday people. I've never said he should RTFM. The things neecessary to use lyx (well, without the more complex parts like the equation editor, perhaps) can be fit into two or three pages. That's reasonable to ask a user to read when using a completely new program. What info does he need, anyway? Especially after knowing to use msword on a basic level, so he's comfortable with the bare idea of a word processor. Just give him first of all half a page of explanation about lyx taking care of _all_ formatting, and that he doesn't need to worry about the size of the heading's font, etc. ever again. As long as he uses styles properly, of course. So you can say, once again, "When you want a heading, put a cursor in the heading line and select 'heading' from this drop-down box. You can see the font on screen becomes big, because the program now knows it's a heading. And when you print it it'll be the right size for a heading. You don't need to think about what exact size it will be. Just trust lyx, it will be the right size." Endorsement from one's employer, or success stories from other people would help greatly in gaining this trust. > > We are not talking about stupid people here. We are talking about people > behaving like the slightly evolved apes that we all are. Some of us are > more comfortable with symbols, logic, abstractions, etc. Most of the > people, however, work differently. Complaining about it won't help - you > have to work with the humans as Evolution made them. Since Evolution is not > God, people are not perfect, and certainly are not adapted to anything > *you* want them to adapt to. No, but they _can_ adapt. They adapted to usnig a computer at all, in the first place. Evolution is also about survival of the adaptable. > > > So you don't think we can or should go for linux on the desktop for > > non-programmers? > > I do. I just think we should start realizing that we are dealing with > non-geek humans, and maybe even *W00t!* consult with people who specialize > in human perception, behaviour and learning. Yes, I know, social sciences, > brrrrrr. > > Then we should start writing applications that fit humans, not try to fit > humans to existing applications! Well, don't you think msword is _not_ something that fits humans - the kind of humans that aren't used to its quirks yet? And if you also think lyx doesn't fit humans, what (existing or not) program do you think is best? -- Dan Armak Matan, Israel Public GPG key: http://cvs.gentoo.org/~danarmak/danarmak-gpg-public.key
pgp00000.pgp
Description: signature
