Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!wuarchive!sdd.hp.com!think.com!yale!bunker!hcap!hnews!369!8!Rick.Alfaro From: Rick.Alfaro@f8.n369.z1.fidonet.org (Rick Alfaro) Newsgroups: misc.handicap Subject: Re: artic vs vocalize Message-ID: <19170@bunker.isc-br.com> Date: 4 May 91 04:18:41 GMT Sender: news@bunker.isc-br.com Reply-To: Rick.Alfaro@f8.n369.z1.fidonet.org Organization: FidoNet node 1:369/8 - The Catwalk BBS, Davie FL Lines: 77 Approved: wtm@bunker.hcap.fidonet.org Index Number: 15382 [This is from the Blink Talk Conference] DA> Mary, I have been using Vocal-Eyes and ASAP both over the past DA> couple weeks. You ask a good question. In part, experience DA> helps you set up a good configuration. It can be hard to get DA> it though. Secondly, most programs, including V-E come with DA> pre-set configurations for common programs like WordPerfedct. DA> Artic does too. You can study them for ideas. Finally, often DA> you just have to go into review and carefully examine the DA> screen. I was initially impressed with ASAP. It works with DA> most things pretty well, but as I started using it for day to DA> day work, I got increasingly annoyed. For some things, it is DA> more chatty then I would like. For some programs with complex DA> screens, it doesn't read things in the right order. It may DA> read the bottom first. I presume that it has to do with the DA> way the screen is updated. Larry may well be on to something, DA> but it could be six months or a year before it is stable and DA> reliable. With it doing the thinking, I don't always know what DA> it is doing or why, so I might miss something and not even know DA> it. Ultimately, the configuration/artifical intelligence thing DA> probably won't be an either or thing, but we are in a DA> transition time. David, I think you brought up some very pertinant and valid points with respect to AI and ASAP. They don't offer a program demo, so I haven't had hands on experience with the program, but I did however listen very carefully to the demo tape. Even on the Demo tape, it was obvious that some lines were getting repeated, and in some cases information was being read that in my opinion didn't need to be. The solution for those cases where it was talking too much or giving info in odd order is to configure the program for that application. As you can see, the user is then back to having to create a configuration anyhow! Like you said, Larry may very well be on to a good idea, but it definately needs maturing. Even when something like this does eventually get perfected somewhat, I don't think there will ever be anything that will take the place of a real good "custom" configuration. I certainly understnd where users are coming from when they express the opinion that they don't want to be bothered with configuring their speech programs, as I know it sometimes can be frustrating. However, having parts of the screen automatically read to you sometimes in unpredictable order seems to me would in many cases cause uneeded confusion. Going into some sort of review mode on the other hand to examine a screen gives you information that can be easily interprated by the user as you know exactly what you are reading and where it is on the screen. Once you determine what exactly it is that you want read automatically, it is just a matter of telling your speech program to do it as part of your configuration. I know this is easier said than done, but the various speech programs are getting better I think with repsect to user friendliness with the exception of a couple of them. I know for a fact that vocal eyes is definately getting friendlier and more transparent to the user without loosing any of its power and flexibility. Hopefully, other speech software packages are doing the same, but artificial intelligence still has a long way to go I believe, and in my mind, will never be a replacement for total configurability and flexibility. The end result of a good configuration is an application that speaks exactly what you want it to speak and when you want it spoken. The key to this is to make this configuration as easy to accomplish as possible and good down to earth documentation. No matter how friendly a program is however, the user has to be willing to put some time into it. I mean, we are talking here about a piece of software that makes your day to day life much simpler and opens up another window to the world, not a text adventure game or something like that. I make it my business to know any speech program I use on a daily basis inside out because it is probably one of the most powerful tools I posess. Guess I got a little long winded here! Sorry! Regards, Rick -- Uucp: ..!{decvax,oliveb}!bunker!hcap!hnews!369!8!Rick.Alfaro Internet: Rick.Alfaro@f8.n369.z1.fidonet.org