Hi Alan,<div><br></div><div>Oh yes, and we shouldn't leave out Schneiderman's lab and Pad++ et al. I'm not saying that Raskin invented zooming at all. But his take with that hospital information system worked for virtually instant learning. That's the main part I want to replicate elsewhere and Lively Kernel is my favorite choice for where. </div>
<div><br></div><div>I still don't know how to get text into boxes and how to edit it there, but I would wager that is not a real problem at all. Indeed, when I have what I want, I'll decorate demo apps with instructional pages all over the place so that they are at hand exactly when and where the student needs them. The zooming aspect of the system lets me do that without using much space in the demo. There is always room to add as much as you want if you just write small, or very very small.</div>
<div><br></div><div>What I don't know is how to make a box and expand the whole world until the box is the only or almost the only thing showing on the screen. If I could then make the contents of the box be a video, or really a video player like YouTube I'd love it.</div>
<div><br></div><div>Similarly, I'd want another box to be some web site, or really a browser like Firefox or Chrome, that would extend the utility of the system. Of course I want another box to just run some arbitrary application. </div>
<div><br></div><div>Now comes the hard part. Dan spoke of Lively Kernel as a laboratory for user interface experiments. I want more, more, MORE. I want a UI lab system where I get to intervene with scripting to revise the UI of anybody's app. </div>
<div><br></div><div>First thing I want to do is eliminate modal dialog boxes without violating anybody's intellectual property rights. When such a thing happens, because such widgets have not yet been universally banned as inhumane thought train breakers, I want my intervention system, in Lively Kernel, to prevent my ever seeing it, record the contents in a log, and push the usually singular button to recover the ability to do anything else with that app. Usually the button reads "OK" which I hate. It might be a bit less obnoxious if it said "Damn", but I still never want to see it nor have my entire computer captured and held hostage by it.</div>
<div><br></div><div>It might be that some human would have to identify each modal dialog box and set up to activate the appropriate failsafe recovery, but then a few hours of work in the UI lab could repair that aspect of yet another poorly built app. </div>
<div><br></div><div>Then I want to make approaching any new complex app easier for real human users. Again it may take some hours to simplify the app, removing all indication of the true complexity available to a power user of the app. Nobody wants the pain and extended confusion of starting to use Photoshop. Nobody. </div>
<div><br></div><div>There may be a hundred sensible ways to begin using that giant app, but more likely millions. If a power user had a good idea of what kind of mini-Photoshop would be good for a student who just wanted to manipulate the contrast of an image, I want her to have the UI lab facilities to create such a simplified app without touching the actual code of the real app. If she can whip up a demo in a few hours, it would be monstrously easier than the effort to get Adobe to do it for her.</div>
<div><br></div><div>I don't care if the basis for such a UI lab system requires two or three virtual machines per app and Unix pipe like connections for the screen image, the mouse, the audio stream, and so forth with coordinated filters to accomplish the transformation. Still I want the UI lab to have the quality Larry Wall claims for Perl, namely whip-up-itude. Lively Kernel has the marks of such a system. But of course I am asking for a stream of miracles without parallel in the known world.</div>
<div><br></div><div>Now, just to make certain that I insult everyone, I want to use Fitts' Law to justify replacing the handles on a Morphic object with an eight way pie menu (did you notice that the mirror image of 314 is PIE?) activated by rollover or click (test? set by preference?) on the much-larger-than-a-handle object itself. The gesture is so easy to make automatic that your fingers will quickly know what to do and the choice will require no conscious attention at all, leaving that to attend to your task, and not the particular mechanism employed.</div>
<div><br></div><div>Richard<br clear="all"><br>-- <br>Richard Karpinski, Nitpicker extraordinaire<br>148 Sequoia Circle, <br>Santa Rosa, CA 95401<br>Home: 707-546-6760 <br><a href="http://nitpicker.pbwiki.com/">http://nitpicker.pbwiki.com/</a> <br>
<br>
</div>