From: Kevin Curry [kmcurry@home.com] Sent: Wednesday, October 03, 2001 2:29 PM To: bowman@vt.edu Cc: 3dui List; Chris Rhoton Subject: Re: text/number input in VEs Doug, That mail was definitely not as well thought out as it should have been. Maybe I can clarify. First, I completely agree that it's a problem to require that people jump in and out of their VE to do "real work" (or just rote tasks, really). I also agree that adding a keyboard to the mix when you may already be needing gloves or a wand is not a good idea. It's a really bad idea if it interferes with those devices. Finally, "best tool for the job" is indeed a poor choice of words - most familiar tool for the job is really what I meant. My overall intention was to question why the keyboard is being totally overlooked in the discussion, so let me speak to that. Maybe I'm guilty of CAVE snobbery here. Are most people on this list working with HMDs? If I'm in a CAVE, or better yet a telecubicle, then I should have the option of typing into a keyboard that sits on a desk - even while wearing a head tracker, and maybe even while wearing gloves (maybe I just turn off the glove input while typing). If you will permit me to ignore the darkness problem for a moment...I'm still not convinced that there will be unsurmountable issues with training transfer or miniaturization. Millions of people are typing into miniature QWERTYs everyday with Blackberry and the like. Even worse, so many kids in Norway are typing with their alpha/numeric phones that physical therapists have identified a new type of repetitive stress syndrome related to it! It seems clear to me that people know how to type even if the keyboard is not a certain size and shape or sitting on a table. In any case, there is certainly more prior knowledge to work with than requiring the user to learn a new alphabet (Palm grafitti is a new alphabet). I should check out the links you provided for the virtual pinch keyboard before I spout off too much more. It seems you are trying to leverage keyboard knowledge somehow, but I don't know how exactly. KMC Doug Bowman wrote: Kevin, First, good to hear from you. Second, I couldn't disagree more! :-) In immersive VEs, sitting down to type is most often not an option. It forces you to remove equipment you're wearing or carrying, to break your sense of presence, and most importantly, to stop the flow of action that's moving towards accomplishing some goal via the interface. I don't think we should think of VEs as simply the place to look at cool 3D images, then go back to "real computers" to do "real work". In fact, that's the whole thrust of my research - doing real work *inside* immersive VEs. I also disagree with your statement that the "keyboard is the best tool for the job". It's the tool that most people are familiar with, which is a big plus. That's why we used it as the model for the pinch keyboard. However, people are familiar with using QWERTY when they are seated, when the keyboard is a certain size and shape, and when it's supported by a table, among other things. Strapping a keyboard to your body is not going to allow much transfer of training. Miniaturizing the QWERTY keyboard and typing with one or two fingers is also not the answer - remember the early PDAs with miniature QWERTY keyboards? Impossible to touch type on - that's why handwritten input is so much more popular now. Add to that an HMD (or even dark conditions in a CAVE) not allowing you to see the device, and I say you wouldn't be able to type at all. What do others say? Am I way off base here? --Doug -- Doug A. Bowman, Ph.D. (540) 231-2058 Assistant Professor (540) 231-6075 (fax) Computer Science bowman@vt.edu Virginia Tech www.cs.vt.edu/~bowman/ > -----Original Message----- > From: owner-3dui@hitl.washington.edu > [mailto:owner-3dui@hitl.washington.edu]On Behalf Of Kevin M Curry > Sent: Saturday, September 29, 2001 4:45 PM > To: Chad Wingrave > Cc: Pablo Figueroa; 3dui List > Subject: Re: text/number input in VEs > > > I'll have to read that thesis Chad... > > The only difficulty with text/number input that I see is figuring > out how to do > it standing up, without a keyboard. If at all possible, I say, > "sit down and > type!" You can try to teach someone "grafitti" or some special > set of pinches > with a glove, but the keyboard is still the best tool for the > job. Why should a > keyboard not be made available? You have to provide some kind of > input device. > Why not a keyboard? Figure out how to strap a lightweight > keyboard to your leg, > or your forearm, or whatever and I guarantee you that people will > be able to > type effectively...even if they can't see the hardware (e.g., > through an HMD). > How about those little text messenger gizmos all the kids have these days? > > KMC > > -- > Kevin M. Curry > http://members.home.net/kmcurry > > Chad Wingrave wrote: > > > One concept I am working on is that everything the user > > understands is based on their perception of the environment. That > > includes the interface. So, what is it in the environment that builds > > their understanding because that will modify how they perceive and thus > > how they interact. I term these effectors "nuances" because it is the > > little nuances that change how we work in VEs. > > > > I would love to see VEs with interfaces that are build purely by telling > > the VE what hardware you have and what the VE and task in the > VE is. This > > requires lots of work to fully build but I think it is a conclusion that > > we all have come to. There is no standard WIMP interface for VEs but > > maybe there is a function that maps interfaces based upon nuances of the > > VE? > > > > link to my thesis: http://www.cc.gatech.edu/~cwingrav/papers/thesis.pdf > > See Chapter 3... > > > > Comments? > > > > -Chadwick Wingrave (cwingrav@cc.gatech.edu) > > > > On Fri, 28 Sep 2001, Pablo Figueroa wrote: > > > > > Hi Doug, > > > > > > I don't remember good references about this topic, but I think it is > > > necessary to have good I/O in VR applications for any type of > information, > > > including text and numbers. If we don't, the environment won't be > > > complete, so a user will have to switch between applications > to complete a > > > task (not nice at all... ;->) > > > > > > By the way, I wonder how you explore the solution space. One > option is to > > > compare between different hardware environments, as I think > you're doing, > > > but this always leave doubts about better environments or better > > > interaction techniques, for example one based on speech > recognition. How > > > can we find the best option for a particular task? I think > it's possible > > > to find a certain hardware/sw configuration that maximizes > > > performance/usability, but maybe such an installation is too > expensive in > > > a particular situation. Other option is to classify environments by a > > > certain metric (one option is price, despite I don't like > it...) and to > > > choose the best one in a particular category. In this case, > is the pinch > > > glove comparable to a speech recognition engine, for this > particular task? > > > Is there a set of common performance/usability parameters > that can be used > > > in the comparison? > > > > > > Best Regards, > > > > > > On Wed, 26 Sep 2001, Doug Bowman wrote: > > > > > > > Greetings everyone! > > > > > > > > It's been awfully quiet on the 3DUI list lately. I wanted to > > > > remind everyone that the list is here and still active. There > > > > are lots of you out there - consider sending a message to the > > > > list updating us on your current research, asking a question, > > > > or raising a topic for discussion! > > > > > > > > Along those lines, let me suggest a discussion topic. I've > > > > been working on methods for entering text and numeric data > > > > in immersive VEs where no standard keyboard is available. You > > > > may have seen the Pinch Keyboard technique we presented at > > > > HCI International 2001. I'm firmly convinced that such techniques > > > > are necessary and useful in many VE applications, but others > > > > apparently don't see the need. What do you think? Is text/ > > > > number entry an important task for immersive VEs? Why or why not? > > > > > > > > On a related note, I'm looking for references in this area. I > > > > know about Ivan's virtual notepad, and several other "virtual > > > > keyboards" (not used in VEs, but could be applied there). > > > > Has anyone used a chord keyboard, soft keyboard, or speech for > > > > text/number entry in an immersive VE? Any usability or performance > > > > results? > > > > > > > > We're currently doing a study comparing the pinch keyboard with > > > > a Twiddler 2 chord keyboard and a soft keyboard using the pen & > > > > tablet technique. > > > > > > > > Thanks, > > > > Doug > > > > > > > > -- > > > > Doug A. Bowman, Ph.D. (540) 231-2058 > > > > Assistant Professor (540) 231-6075 (fax) > > > > Computer Science bowman@vt.edu > > > > Virginia Tech www.cs.vt.edu/~bowman/ > > > > > > > > > > > > > > > > > > -- > > > ------------------------------------------- > > > Pablo Figueroa pfiguero@cs.ualberta.ca > > > PhD Student http://www.cs.ualberta.ca/~pfiguero > > > University of Alberta > > > > > > Relax ... God is in charge > > > ------------------------------------------- > > > > -- Kevin M. Curry http://members.home.net/kmcurry kmcurry@home.com