dear friends. after going through all the resources available as speach options for synthesis on linux and possible screen readers, I have come to the following conclusions. 1. with due respect to all the efords for developing a very natural sounding synthesizer (I know it is very hard), no attempts have been so far organised in a way to make it really of high standards. just hear eloquence or dectalk and you will come to know what I am trying to say. these two are the most widely used synthesizers but are non open source. further they are very costly too. we can approach the carnegi melon univercity or edinbaro for assistance. what is primarily needed is to develop a very high quality exceptable tts engine like text assist or decktalk. we also need to make a speach synthesizer which besides indian languages can talk indian english. it should be noted that today's english synths are only british, australian and american english talking. this is not always easy for a blind person to understand. I have been to the states and UK so I know how easy it is to understand but think about those who don't have this opportunity? (and most of them don't have). 2. we need to develop a really really good screen reader. I will fix up a project meet for this and I will demonstrate the leading screen reader on windows platform, JAWS for windows. compare it with orca and/ or gnopernicus and the difference will be obvious. it is not the amount of functionality which is causing a problem, functionality and features can be added over a period of time. the problem is the way in which accessibility needs to be implemented in screen readers. over the 4 years of speach interfaceing and accessibility, one thing I have noticed is that screen readers still read text and not what it simbolises. fortunately gnome desktop has got accessibility built in with the core api so it becomes that much easy to wrok ork it out. k destop has got accessibility too but has not been exploited right now. 3. general awareness amongst the blind people about linux is another major issue. while this also applies equally to sited people, I have found masses of blind people who still feel that windows and ms office are the only two things that ever exist on a computer. for them linux is as good as discovery of life on other planets. we need to make them linux aware and how it particularly benifits them. 4. point number 1 and 2 must be taken up by the same programming unit or at least different departments of a same team. the two points go hand in hand. as an aside to the awareness issue. I am planning a school level campain for teachers and students. the reason why no such attempt is becoming successfull is that we end up takeing all this unfortunate microsoft users into a strange land. let's not shy away from comparison and rather prove it that gnu/linux not only offers every thing that windows gives, but it gives more than that with some learning. I believe jtd, roni, and others will agree that with the advent of ubuntu and the latest debian releases gnu/linux has become absolutely user friendly and flawless enough that even a house wife can use it. there should be no reason why we should not positively compare our selfs against windows and show the windows users what they are missing? Krishnakant.