Waving the red flag
Locked
 
|
Hello Dan and everybody,
Smalltalk is wonderful. Making it work in virtually every browser by coding the base system in Javascript is very clever and could be marvelously useful. Enriching the current code to make Lively Kernel suitable for awesome presentations well beyond what PowerPoint could dream of has much appeal for me. I love neat things that are useful and can be acquired inexpensively in money and time. When such things have unbounded utility, Pavlov sets in and I slather and drool. Keep it up, man. I don't mind that my shirt gets wet.
But wait. How long does it take for someone to be comfortable navigating around in a Lively Kernel world? I'm sure it's not one of those things that takes weeks to get into, but I worry that it might take an hour or two. What I want is a system that computer experts can become competent with in only a few minutes. It would be truly great if novices could get there even faster. But who knows how to build such a system?
Today, I think no one knows how to do that. However, the late Jef Raskin, father of the Macintosh and author of "The Humane Interface", did. Given a charter to assist in getting around in a patient's chart which was impossible to read when fully displayed and awkward to navigate when magnified to be readable, Jef used zooming to good effect. He wanted to call it a Flying User Interface, not only because he liked flying and it felt like that, but especially so he could call it (phonetically) a Phooey. He was like that.
Anyway, he discussed the system in his book, but he left out some details. When computer experts were trained to use the system. they became comfortable and competent in less than TWO minutes. But when utter novices, who maybe recognized the mouse as a thing to push around, not speak into as Scotty did, they became fully functional with the system in less than ONE minute.
I really like that. I want that. With such a system I could teach a three year old to use it, or a 93 year old, or even a college professor. I am NOT kidding, the first and second examples may have time to spare, but the prof does not.
Why does it work so well? My theory is that for tens of millions of years, our ancestors made it back to the nest, or we would not be here today. Thus the talent for geographic navigation is built into our DNA. We do not forget where the fridge is or where the couch is. Often we can get to such places in the dark. If our computer world is so arranged, people won't get lost so often. If we can follow links by rolling into a thumbnail and can return by recrossing that border it will seem natural to us.
What would it take to make it easy to construct and navigate such a zoom world in Lively Kernel?
Richard -- Richard Karpinski, Nitpicker extraordinaire 148 Sequoia Circle, Santa Rosa, CA 95401 Home: 707-546-6760 http://nitpicker.pbwiki.com/ |
|
|
History sometimes disrupts good stories, and this is one of those times. The first really good zooming interface was done many years before Raskin by Nicholas Negroponte's Arch-Mac group at MIT in the 70s. (It was called the Spatial Data Management System) Raskin was greatly taken to task by the community for somehow not mentioning this in his book. Another zooming interface after Negroponte and before Raskin was done by Vaughn Pratt at Stanford. And yet another by Randy Smith at PARC. And so it goes .... Best wishes, Alan From: Richard Karpinski <[hidden email]> To: Lively Mail List <[hidden email]> Sent: Sun, April 11, 2010 11:08:01 AM Subject: [lively-kernel] Waving the red flag Hello Dan and everybody, Smalltalk is wonderful. Making it work in virtually every browser by coding the base system in Javascript is very clever and could be marvelously useful. Enriching the current code to make Lively Kernel suitable for awesome presentations well beyond what PowerPoint could dream of has much appeal for me. I love neat things that are useful and can be acquired inexpensively in money and time. When such things have unbounded utility, Pavlov sets in and I slather and drool. Keep it up, man. I don't mind that my shirt gets wet.
But wait. How long does it take for someone to be comfortable navigating around in a Lively Kernel world? I'm sure it's not one of those things that takes weeks to get into, but I worry that it might take an hour or two. What I want is a system that computer experts can become competent with in only a few minutes. It would be truly great if novices could get there even faster. But who knows how to build such a system?
Today, I think no one knows how to do that. However, the late Jef Raskin, father of the Macintosh and author of "The Humane Interface", did. Given a charter to assist in getting around in a patient's chart which was impossible to read when fully displayed and awkward to navigate when magnified to be readable, Jef used zooming to good effect. He wanted to call it a Flying User Interface, not only because he liked flying and it felt like that, but especially so he could call it (phonetically) a Phooey. He was like that.
Anyway, he discussed the system in his book, but he left out some details. When computer experts were trained to use the system. they became comfortable and competent in less than TWO minutes. But when utter novices, who maybe recognized the mouse as a thing to push around, not speak into as Scotty did, they became fully functional with the system in less than ONE minute.
I really like that. I want that. With such a system I could teach a three year old to use it, or a 93 year old, or even a college professor. I am NOT kidding, the first and second examples may have time to spare, but the prof does not.
Why does it work so well? My theory is that for tens of millions of years, our ancestors made it back to the nest, or we would not be here today. Thus the talent for geographic navigation is built into our DNA. We do not forget where the fridge is or where the couch is. Often we can get to such places in the dark. If our computer world is so arranged, people won't get lost so often. If we can follow links by rolling into a thumbnail and can return by recrossing that border it will seem natural to us.
What would it take to make it easy to construct and navigate such a zoom world in Lively Kernel?
Richard -- Richard Karpinski, Nitpicker extraordinaire 148 Sequoia Circle, Santa Rosa, CA 95401 Home: 707-546-6760 http://nitpicker.pbwiki.com/ |
|
|
In reply to this post by Richard Karpinski
[re-send to the list as originally intended]
Hey, Dick -
You didn't answer my little questionnaire. Please
everybody, do this.
Regarding zooming UI's, in addition to Negroponte's early work
that Alan mentioned, you might be interested in this paper:
Generalized and stationary scrolling, ACM
SIGCHI 1999
Authors: Randall B. Smith, Antero
Taivalsaari
ABSTRACT:
We present a generalized definition of scrolling that unifies a wide range of existing interaction techniques, from conventional scrolling through pan and zoom systems and fish-eye views. Furthermore it suggests a useful class of new scrolling techniques in which objects do not move across the display. These "stationary scrolling" techniques do not exhibit either of two problems that plague spatial scrolling system: discontinuity in salience and the undermining of the user's spatial memory. We should notice that Antero was a co-author with Randy. In
fact one of the first experiments Antero did in the Lively project was
to play around with zooming worlds instead of our Squeak Project-like
wormholes. I thought we even had some code from this in our
repository, but I don't see it. [I see that we lost Antero's Sun
address from the mail list, so I'll CC him]
I just tried inspecting the world and evaluating
this.scaleBy(1.1) but i see we currently don't use the world's
transform (tsk, tsk).
|
|
|
In reply to this post by Richard Karpinski
There's talk lately of presentations and zooming in LK. Came across
this about a year ago. Implemented in flash I think. Just some fodder of prior art for review:
http://prezi.com/ On Sun, Apr 11, 2010 at 2:46 PM, Dan Ingalls <[hidden email]> wrote:
|
|
|
Phil wrote...
There's talk lately of presentations and zooming in LK. Came across this about a year ago. Implemented in flash I think. Just some fodder of prior art for review: http://prezi.com/ Incredible. I just got a pointer to this same site today
from Joe Armstrong.
Not just fodder. It's an inspiration.
Also in the inspiration department...
I haven't got my iPad yet, but I found an old Windoze
machine here with a touch screen. I managed to get Chrome
running on it and LK on that. For some reason simple one-finger
control is part of the gestalt I envision for Lively. [I had
this same feeling about Fabrik over 20 years ago, too]. So I've
been tapping on the screen with a far away look in my eyes.
Let's see...
Draw approximate shapes in the world to make simple morphs.
Tap in an object to get controls and a menu
Handles and menus need to be bigger for fingers
Second tap goes to parent like Squeak
Pie menus? Still not for novices I think
Gravity, momentum, collision detection
Oh help...
|
|
|
In reply to this post by Richard Karpinski
Hello Richard,
In April you wrote: "What would it take to make it easy to construct and navigate such a zoom world in Lively Kernel?" I hope you can spend some time to answer this question with a drawing:
Philip Weaver On Sun, Apr 11, 2010 at 1:08 PM, Richard Karpinski <[hidden email]> wrote: Hello Dan and everybody, |
|
|
Prezi is indeed pretty rad. Patrick, please reply and explain more about the zebra and paths. I have not looked closely enough.
I think that when a zoomable user interface becomes a priority that some aspects of Lively ought to change. Drag to pan by default. Lasso selection via contexual menu instead. (Keep usual single selection behavior.) Double-click a morph to zoom to fit (similar to opening a window from an icon). Viewports (scrollable viewports) should instead be views to subworlds. Etc. But we need some drawings and need to consider "subtlety" for a zoomable user interface. Thanks for writing Patrick. If you are willing to draw mockups for Lively I need you. http://tinyurl.com/lively-mockups/ If you want to code, the project needs you: talk to Dan, Jens, or Robert. Peace, Philip On Wed, Jun 30, 2010 at 10:54 PM, Patrick Shouse <[hidden email]> wrote: This description of a Phooey reminds me of Prezi: http://prezi.com. Some concepts from Prezi like the zebra and paths might be useful in Lively Kernel. |
|
|
Hello Richard,
In April you wrote: "What would it take to make it easy to construct and navigate such a zoom world in Lively Kernel?" I hope you can spend some time to answer this question with a drawing: Thanks, Philip Weaver On Wed, Jun 30, 2010 at 11:16 PM, Philip Weaver <[hidden email]> wrote: Prezi is indeed pretty rad. Patrick, please reply and explain more about the zebra and paths. I have not looked closely enough. |
|
|
In reply to this post by Philip Weaver
Hi Philip and Richard,
apropos Zooming: we build in Mac like simple zooming a while ago.... http://www.lively-kernel.org/repository/lively-wiki/documentation/Zooming.xhtml Best, Jens Am 01.07.10 05:12, schrieb Philip Weaver: Hello Richard, |
|
|
I do read your journals every week or so since April. I hope everyone at HPI will post to this list about your recent accomplishments.
Zooming in Lively would need more structure to not be cumbersome - e.g. double-click a morph to zoom to fit, to drill in. Be able to zoom to any morph from a columnar system browser. But this means that pickup on mouse down is done / over with / dead / finished, I want single-click "pick up" to go away - it is often just very annoying. Keep the auto-grouping, auto-parenting. I personally think that a polished environment and IDE and more important than zooming. But a zooming mockup still needs to be drawn. Don't discuss - don't code - please draw. A picture is worth a thousand words. If no one is exploring any degree of funding to help propel this project or derivatives then that is a great tragedy. Philip On Thu, Jul 1, 2010 at 4:00 AM, Jens Lincke <[hidden email]> wrote:
|
«
Return to Lively Kernel
|
1 view|%1 views
| Free forum by Nabble | Edit this page |