We quite appreciate being slashdotted! Some great and insightful comments. I'll start by addressing two main ones: 1. Crypto is not the end-all security answer. We agree. Much of our documentation is designed to educate about this issue. The main security threats to this data are not someone cracking a strong crypto solution, by our attack analysis. It's the bad password, snooped keyboard strokes, torture, etc. We are also pretty upfront that today's strong crypto is the next generation's college coding project. Our crypto makes HR data 99% more secure (maybe 95%). The most important thing is that it gives these groups more control over their information and makes it a lot less likely that it will be lost. 2. The Al Quaida concern. The terrorist groups already use/have access to secure communications for email. Martus is designed for human rights bulletins. While it is conceivable that terrorists could use it, why would they if they have better tools for their needs? I like to use the example of guns/machetes/hoes. All of them can be used in committing genocide. Hoes just happen to be far better for farming than for killing, and it's rare for them to be used as a weapon. Design is a strong signal of intention.
We've had a lot of success at www.bookshare.org with distributed scanning and proofreading with volunteers. We're adding a couple hundred books a month. Our major goal is a very large library of the books people want to read. However, due to copyright law, we can only do this for people with qualifying disabilities in the U.S. when we're looking at books still in copyright.
I was very impressed with Joe's answers on these questions. I was one of the members of the federal advisory committee that drafted the initial take on Section 508's web accessibility regs. His statement about the web accessibility business is right on: this is a temporary blip as we go through a transition to where accessibility just becomes another part of good design. One thing I will take issue with is Joe's statement that accessibility advocates want alternative text sites. I think that approach falls under the separate and unequal approach: tried that and it's been discredited. The goal for almost everyone is that the same site work for everybody and look good. It isn't hard. The issue about people with cognitive impairments is that they have few advocates at the national lobbying table (other disability groups are good at self-advocacy), plus it isn't clear what to do to make sites more accessible other than make them more usable for everybody.
Actually, it is quite exciting for most blind people and many people with speech disabilities. Technically skilled blind people won't be excited for the reasons posted. As a developer of software for the blind, one of the biggest barriers to use of technology by seniors (the majority of blind people) was the machinelike quality of the current voices. This voice improvement brings the technology much closer to human sounding.
It may work pretty well for Stephen Hawking, but a teenager who has no voice would really prefer to have a voice that sounds closer to a human.
The focus in the New York Times piece was on the downside of a human sounding synthetic voice: I would rather see the positives.
Slashdot Mirror
We quite appreciate being slashdotted! Some great and insightful comments. I'll start by addressing two main ones:
1. Crypto is not the end-all security answer.
We agree. Much of our documentation is designed to educate about this issue. The main security threats to this data are not someone cracking a strong crypto solution, by our attack analysis. It's the bad password, snooped keyboard strokes, torture, etc. We are also pretty upfront that today's strong crypto is the next generation's college coding project. Our crypto makes HR data 99% more secure (maybe 95%). The most important thing is that it gives these groups more control over their information and makes it a lot less likely that it will be lost.
2. The Al Quaida concern.
The terrorist groups already use/have access to secure communications for email. Martus is designed for human rights bulletins. While it is conceivable that terrorists could use it, why would they if they have better tools for their needs? I like to use the example of guns/machetes/hoes. All of them can be used in committing genocide. Hoes just happen to be far better for farming than for killing, and it's rare for them to be used as a weapon. Design is a strong signal of intention.
We've had a lot of success at www.bookshare.org with distributed scanning and proofreading with volunteers. We're adding a couple hundred books a month. Our major goal is a very large library of the books people want to read. However, due to copyright law, we can only do this for people with qualifying disabilities in the U.S. when we're looking at books still in copyright.
I was very impressed with Joe's answers on these questions. I was one of the members of the federal advisory committee that drafted the initial take on Section 508's web accessibility regs. His statement about the web accessibility business is right on: this is a temporary blip as we go through a transition to where accessibility just becomes another part of good design.
One thing I will take issue with is Joe's statement that accessibility advocates want alternative text sites. I think that approach falls under the separate and unequal approach: tried that and it's been discredited. The goal for almost everyone is that the same site work for everybody and look good. It isn't hard.
The issue about people with cognitive impairments is that they have few advocates at the national lobbying table (other disability groups are good at self-advocacy), plus it isn't clear what to do to make sites more accessible other than make them more usable for everybody.
Actually, it is quite exciting for most blind people and many people with speech disabilities. Technically skilled blind people won't be excited for the reasons posted. As a developer of software for the blind, one of the biggest barriers to use of technology by seniors (the majority of blind people) was the machinelike quality of the current voices. This voice improvement brings the technology much closer to human sounding. It may work pretty well for Stephen Hawking, but a teenager who has no voice would really prefer to have a voice that sounds closer to a human. The focus in the New York Times piece was on the downside of a human sounding synthetic voice: I would rather see the positives.