Welcome to the sixth-sixth episode of Access On, the National Federation of the Blind's Technology podcast.
Episode
Listen to the sixth-sixth episode of the Access On podcast (Browser).
Or listen on your preferred podcast platform.
Timestamps
On this week's episode of Access On:
- Introduction 0:00
- Looking ahead to next week's CSUN recap 0:53
- Join us for our Access On seminar on Windows screen readers 1:56
- NVDA past, present, and future 3:08
- Closing and contact info 1:09:26
Transcript
Quentin Christensen:
Live the life you want.
Jonathan Mosen:
Welcome to Access On, the Technology podcast of the National Federation of the Blind. On this episode, an extensive interview with three key people from NV Access, the charitable foundation responsible for the NVDA screen reader.
We are already at episode 66 of Access On. Get your kicks on episode 66. Catchy, isn't it? And I'm Jonathan Mosen, the Executive Director for Accessibility Excellence at the National Federation of the Blind, coming to you from the Jernigan Institute in Baltimore, Maryland, where I believe I can say we're getting a wee bit of a sense of spring. So that's a wonderful thing.
A couple of notes before we get into our major feature today. If you are coming home from CSUN as you listen to this, I mean, I hope you're not listening to this at CSUN because you want to be getting out there and networking, right?
So if you're coming home from CSUN, we are going to be doing a feature for next week's Access On in which we talk to some people who have attended CSUN and we recap some of the sessions that we went to and the things that we saw at the exhibit hall.
Just in case there are people who don't know, CSUN is the California State University Assistive Technology Conference. It's been going for over 40 years, and it is a pretty significant event on the Access Technology calendar. So we will talk about this next week.
If you went to CSUN and you have some thoughts on what you saw, you are very welcome to get in touch and we'll try and include some of these emails on next week's episode. You can attach an audio clip or just write an email down and send it into AccessOn@nfb.org. That's AccessOn@nfb.org.
Next, let's look ahead to our Access On Seminar. Now, when we call them a seminar, that means it's the long one. It's a four-hour presentation. We'll be doing this on March the 18th, so that's next Wednesday as this podcast is published. And we're looking at the screen reader world, specifically the Windows screen reader world, and we'll be giving an extensive look at Narrator, NVDA, Dolphin ScreenReader, and JAWS.
We're not comparing all four, but we will go through the features, what perhaps makes them stand out, what their philosophy is, some of those sorts of things, and how you can learn them, how you can get help. If you want to install one of these things that you've not used before, how do you get up and running? How do you get going?
It should be an interesting four hours and there will be time for Q&A. If you would like to register for this, just head over to nfb.org/cena. That's nfb.org/cena, which is short for our Center of Excellence and Non-visual Accessibility here at the National Federation of the Blind.
You'll be very welcome. And if you register but can't attend, registration ensures that you'll get a link to a full recording of the seminar. So it's worth doing, nfb.org/cena.
Any blind person who thinks that the system is too big and established to change need only look at the story of NVDA for inspiration. Born in 2006, out of one man's frustration with the cost of access technology, NVDA has grown into the most widely used screen reader on the planet with more than 250,000 users across 175 countries, which is quite extraordinary.
And this year, 2026, NVDA turns 20 years old. To mark that milestone and to dig into where NVDA has been, where it's going and what users are really asking for, I'm delighted to welcome three people from NV Access to Access On today.
Mick Curran is the founder and technical consultant for NV Access. He started writing NVDA at the age of 22, shortly after losing his sight because he believed that blind people deserved access to computers for no more cost than their sited peers. Dr. Gerald Hartig is the chief technology officer at NV Access, the person who now steers NVDA's technical direction.
Gerald brings over a decade of expertise in software engineering and product management. And Quentin Christensen, who's no stranger to these podcasts, is NV Access's training and support manager, the person who helps ensure that once people get NVDA, they can actually use it to its full potential.
Gentlemen, welcome to Access On. It's a real pleasure to have you all here. Mick, what problem were you trying to solve when you began NVDA in 2006, and what felt broken about the status quo to you?
Mick Curran:
I think it's fair to say that I was fairly lucky here in Australia with the kind of support that I got as a blind person going through education. I started off life vision impaired, quite severely vision impaired, but not totally blind. And I was able to get around pretty much without using screen reading software.
I was more relying on Braille and large print, but then I lost my vision at 15. But right at that time, then it was identified that I would require screen reading software and other kind of technology.
But in order for us to get that kind of technology, we had to rely on charitable funding and we were very lucky to get that. But right from the get go, right from that point, I thought that was a little bit strange that, as I say, even for us in our situation, my family still had to rely on extra funding outside of the system in order to get the technology that I required simply just so that I could continue my school.
So that's in the back of my head for quite a while, even though I was only 15 at that time. And I continued on with my schooling, went to university.
But yeah, I kept on thinking, "This is a bit strange that I still have to rely on all of this extra stuff because what if I wasn't so lucky? What would I be doing and where would I be positioned if I wasn't able to be the one successful child in my town or whatever to get that kind of charitable funding?"
And so then partway through university, I just kept on thinking about it more and more thinking, "Well, look, I use all this other kind of open source software and Linux and other free software, so why isn't there something here for Windows?" I mean, I don't know how difficult it is to do. I have no idea, but could at least try.
And that's really how it started, I guess. So it was partly my own need because I didn't want to or couldn't pay for the constant updates, because okay, you might get charity funding to get the initial technology, but you've got to keep it up to date.
So yeah, it was partly my need, but it was also partly a curiosity too, something to do in life. I had no idea how big the project would be or anything, but just, "Hey, let's give it a go."
Jonathan Mosen:
Hypothetical question. In 2006, Australia didn't have the comprehensive and quite respected national disability insurance scheme that it now has. Do you think you would've done NVDA had the NDIS existed then?
Mick Curran:
That's a really tricky question. I think I probably would have, knowing me simply out of interest and curiosity as a fun project, because I was into a lot of different projects to do with audio engineering and scripting for other screen readers and all that kind of stuff.
So I was really embedded in that kind of thing anyway. So in some ways, it was just something that interested me. As to whether how much energy I put into it and whether other people around the world would've, or even in Australia would've taken interest, I don't know. But I think I still would've tinkered with it.
Jonathan Mosen:
And was the fact that NVDA is open source a matter of philosophical principle from day one or did that kind of evolve as the project evolved?
Mick Curran:
It was definitely a core principle from day one, because I'd sort of grown up, or at least from my late teenage years, very much embedded in the open source movement with using Linux and setting up servers and all that kind of stuff, and I could really see the potential there. And no, from the outset, it was very, very clear to me that we needed to make this open source.
Not just because I felt it was the right thing to do, but also because I knew that I didn't have all the answers. And there was a whole community out there of blind people who probably had way better knowledge and skills than me around the world and hopefully might be able to contribute to the project.
And I think there was also a bit of a sense of just giving us something to do as well, something fun, because I knew and I was in the same position, but I knew there were a lot of blind people around the world who, especially in developed countries, even in Australia, there's a bit of a trend to sit at home on your blind pension, everything, the world's good and not be too motivated to look for a job.
And even at 22, that sort of annoyed the crap out of me. I'm like, "No, there has to be more to life than this." And I wanted to contribute back somehow.
And so I, maybe a bit of a righteous point of view, but thought, well, maybe this might be something fun for us to do and to, I don't know, change our destiny a little bit, take the problem into our own hands.
Jonathan Mosen:
If you go all the way back to the invention of Braille itself, we have seen countless examples of groundbreaking technology being developed by blind people for blind people. And it feels like the ability for anyone to contribute to the code has really created this strong feeling of community and ownership in NVDA.
Mick Curran:
Yeah, that's very true. Absolutely.
Jonathan Mosen:
I imagine there are many examples by now of blind people who learned how to code precisely because they wanted first and foremost to contribute to this product. They wanted to use it, they wanted to shape it, and have gone on then to software development. Have you seen that where people have kind of cut their teeth on NVDA and then made a career out of software development?
Mick Curran:
Yeah, we've definitely seen people like that. Certainly, I can think of, say, someone who started around 15 or 16 contributing to NVDA as a teenager and then have gone on to computer science degrees and then gone on to work for big companies.
I mean, I can think of in particular one person who's gone on to have a great career at Google. There was another one who works for Microsoft. There are lots and lots of cases where people have joined our email list as I say, around 16 or 17 or so, and we've watched them grow over the years.
And they've learnt through our project, I guess the truths, I suppose, and the etiquette of how to run an open source project, what to do and what not to do, how to interact with the community, the work it does take to polish your code so that it will be accepted by a big project like ours, how to work in a friendly way with other community members. And yeah, it's really exciting to watch them grow and then hear that they've gone on and started these wonderful careers with the big techs in part because of NVDA.
Jonathan Mosen:
When you look at all that's been achieved, the usage, the impact that it's had, you must feel incredibly proud.
Mick Curran:
Yeah, I do. I do. Yeah, I'm pretty happy to wake up each morning and know what we've all achieved together. It's not just me at all. I wrote the first line of code, but my good friend and colleague, Jamie The, also jumped on the project very, very early with me within months.
And then of course, as I say, the whole community has jumped on board and NVDA is much, much, much more than us. I'm sure we'll get into this later, but I mean, at least 40, 50% of the code is community contributed.
Jonathan Mosen:
And Gerald, that is the vibrant environment that you are now responsible for. And of course the add-ons adding value.
Gerald Hartig:
Oh, I was going to say, of course, the add-ons as well, one of the unique things about NVDA is that there's two parallel paths for development of getting code into the hands of blind people. One is, of course, the traditional way that Mick was talking about with core development work, and the other is of course, add-ons, which represent a fairly low risk way of getting your code out there.
So that lowers the barrier of entry that to get code as an add-on to the hands of people, it may not need to pass the same very, very rigorous standards that we have for getting code into core. It's all about opportunities.
Jonathan Mosen:
We have a mixed audience here. There'll be some people who will be wanting me to get into the weeds because they're using NVDA every day, and we'll see if we have time for that. But I'm also mindful that there will be people who haven't installed it at all at this point. So let me start with that. NVDA has become increasingly capable.
If we assume for a moment that someone can get funding for JAWS if they want to, how does NV Access now position NVDA? Are there use cases where you still think that JAWS is the better fit, or do you believe that NVDA has evolved to the point that it can replace JAWS for pretty much any screen reader user?
Mick Curran:
There are many people across the world who choose to use it full-time in their work or at home. I suppose it's really just another option in the mix. And I probably don't really want to get into what we do better or JAWS does better, because frankly, I don't really know.
I haven't used JAWS for 20 years or more, so I don't know. Sorry, not 20. Yeah, 20. Yes, 20. So I can't really answer that question. All I can really say is that, yeah, we do have hundreds of thousands of people around the world who choose to use NVDA, the majority of them full-time. And so clearly they're getting their work done, so yeah.
Jonathan Mosen:
So I appreciate that you're not in a position to give product comparisons, and we may well do that. That's something that we're looking at here at our Center of Excellence and Non-visual Accessibility. But can I put it this way? If there's somebody who's listening to this who has another screen reader installed and they haven't tried NVDA yet, what would you say to them about why they might like to give it a shot?
Mick Curran:
I think there's nothing to lose. I mean, it's a free product. So if you are starting out, you haven't got a screen reader yet, or you are finding that your other screen reader, you may be having some issues in learning how to do something or something's not working, well, then give NVDA a go, because as I say, it's completely free and just try it out. If it works for you, that's absolutely great.
Gerald Hartig:
We do tend to find that when we do a survey of screen reader users, that there is a very strong tendency for people to use both screen readers. And the reasons for that, of course, they're quite complex as Mick was referring to.
But again, this speaks to providing blind and visually impaired people with more options that once people start using NVDA, they continue to use it, but they also continue to use their old screen reader as well because they solve different problems for them. And I think those problems are fairly specific to individual people. We just want to make sure that we can give people those options.
Jonathan Mosen:
Let's say I get offered a job somewhere that goes beyond the beaten path of Microsoft Office or Google Workspace, and my employees using some sort of proprietary application that would benefit from customization to improve efficiency. Can NVDA be viable in a use case like that?
Quentin Christensen:
Quentin here. Yeah, absolutely. In terms of using it, yep, there's no problem with that. And indeed, we've had a lot of corporate users who have found that the licensing and the ability to just install NVDA on any computer across their network makes it a much easier option to use from the IT department's point of view and from an end user's point of view.
I mean, broadly, you asked before about a feature comparison. I mean, broadly, if you step back and summarize what we do and what other screen readers do in a handful of paragraphs, they broadly do the same thing. They approach the same problems.
We might do it in a slightly different way, which is probably a lot like how many of us have two different browsers installed on our computer, because sometimes one works better than the other. And yeah, that can be the case here too.
In terms of corporate usage, we do offer support. So we have telephone support, which is one thing that many organizations, it's their first question, "What kind of support do we have?" So we do have telephone support. And training material for users as well, so they can get all of their users up to speed and proficient screen reader users.
Can NVDA work with that proprietary written entirely for this organization piece of software? It's hard to give a definitive answer obviously to the hypothetical, but for many software, and certainly if it's been written vaguely adhering to any kind of standards, then probably yes. If it hasn't, can things be customized to work with it? Yes, that's certainly possible.
So NVDA has, as we mentioned before, a rich add-on ecosystem. Just in the official NVDA add-on store, we've got nearly 300 add-ons at the moment, or actually nearly 300, which are compatible with the latest version of NVDA. I didn't check the older ones.
And there are organizations that go off and make things for specific use cases. There was one add-on which helped kickstart the ability to use NVDA with things like Citrix or RDP or a lot of the remote services that organizations use. There's now also a free add-on that will do that as well.
Jonathan Mosen:
So for example, let's say that I'm working in a call center environment with a customer relationship management system that is not typical, it's around. It would be possible then to write an NVDA add-on so that when I pressed a key, I could hear who's calling and when I pressed another key, I could hear bits of their contact record, that kind of thing.
Quentin Christensen:
Theoretically, yes. Obviously, yeah, it's hard to answer the specific on a vague like that. Yeah, theoretically, yeah, that could certainly be possible. We've also introduced features into NVDA, for example, sound split. So in that call center situation, for instance, some users prefer to have the caller that they're talking to in one ear and NVDA in the other ear, and you can set that up. There's an option in NVDA if you prefer to do that.
Jonathan Mosen:
You mentioned installing NVDA on various computers, and it is a very simple process. But of course, many of us who have worked in environments where they're very security conscious have encountered the grumpy IT person, story to stereotype there if you're an IT person person. Sometimes it's very tough because they're reluctant to install any screen reader at all.
And I wonder whether you've encountered any issues relating to how you explain security posture to IT departments that are nervous about, say, add-ons or remote features or configuration drift. Some of these installations, particularly in high secure environments like the federal government are quite locked down. How does NVDA deal with that and trying to crack that to get into those workplace environments?
Quentin Christensen:
Absolutely. So the first thing I would say is if you are an IT manager in that situation or you are employed and wanting to get your IT department to approve NVDA, if anyone has any questions, do please reach out. We are more than happy to answer those.
We have a number of features in NVDA to ensure security and that it will work in those environments and can be approved in those environments. So things like there's a secure mode which locks down a lot of the features which may be concerning to an organization like governments, banks, other organizations concerned about security.
We have a page on our website which lists a lot of the security things that people might be concerned about and how to ensure that we meet those.
Things like NVDA, the executable is signed so that you can check the executable and ensure that it's exactly the same binary program that we've uploaded to our website, for instance.
That secure mode, as I talked about, definitely makes it so that there's very little that can be changed, which IT departments can be quite happy about. And we're also looking at ways of improving that experience for corporate users as well.
Gerald Hartig:
And of course, touching on something that, Jonathan, you already mentioned earlier, the open source ethos that NVDA has. There's some groups of IT people that think that open source is less secure. Whereas whenever there's been a scientific study on the matter, they've seen that the reverse is often the case, that open source is the more secure option.
And part of that is that the code is available, it allows for continuous public auditing. But some IT managers would push back and say, "Well, we don't have the resource to do that and we can't trust the community."
But now that AI is coming on board, this now for the truly paranoid IT manager would allow someone to carry out a white box analysis using AI of our code. You can literally put all of our source code, let's say 100,000 lines of Python through an AI, and verify for yourself that it is a secure option.
And of course, we don't need to talk to the internet for anything. We can live entirely within a fully constrained, secure environment.
Mick Curran:
Not to mention also that we are also used by many of the large corporations, Google, Microsoft, Adobe, et cetera. I can't comment on to how much, how many employees, et cetera, but they're definitely using it within their networks and they are definitely supporting NV Access in an official capacity.
Jonathan Mosen:
We now see Microsoft doing significant work on Narrator and increasing its capabilities. And of course, there's JAWS, which is a strong commercial product in developed countries at least. As Narrator grows in capability, do you have any concerns about NVDA being squeezed by a free product available on every single PC on the one side and a strong commercial operator on the other?
Mick Curran:
Firstly, I think competitorship is healthy. I think that it's really important that people have multiple options, even if they are, more than one of them is free. And I think that's very obvious when we look at historically platforms like Apple, iOS, and Mac, where it's probably fair to say that for some periods of time in the past, some of their options have become a little bit stagnant.
They've done very well, but because there's no competitorship there, sometimes as I say, features can come a bit stagnant. There's less innovation or things can tend to break a little bit.
So at least on Windows, the other options are keeping us on our toes, hopefully. In fact, we're all keeping each other moving forward. I'm sure that all three of the screen readers, us included, could mention situations where the others have done something or whatever, and we've felt, or the other ones have felt the push to do so as well, or do it in a different way.
So I welcome it. I think it's absolutely great. Also, I want to be transparent here that we work very closely with Microsoft. We do receive funding from them. We provide consulting to them. A lot of the consulting we provide does actually directly help Narrator, and we're proud to do so.
My goal has always been from the very, very beginning that I would be able to walk into a computer store, pick up a laptop, go home and use it. Now, if that was because of Narrator, great. That's probably how things should have been from the beginning. It wasn't due to various reasons, and thus we started NVDA and have been doing so for the last 20 years.
It's really exciting to see Narrator grow into a maturing product, which does work in particular situations. And great. I mean, certainly for installing Windows and probably using quite a few different sort of inbox applications in Windows, Narrator is perfect for that. But the user model is a little bit different.
I think NVDA is a little bit closer to JAWS in terms of its key commands, et cetera, and maybe in some of the choices we make in terms of how information is presented. So people might be a little bit more used to that or et cetera. But on the other hand, someone starting out who's just given Narrator, might be able to cope with that.
I think it's a collaborative environment, all of us working together. And if one day Narrator succeeds in all the ways that matter and we no longer become relevant, I'm perfectly happy with that, because a charity only needs to exist while there is actually a problem to solve. I'm sure that NV Access has a long life yet and NVDA still has an extremely large role.
It will continue to play because although Narrator technically might be doing fairly well on some of the inbox experiences, it's still definitely lacking on third party support, and again, the user model side of things. And probably by far most important is the community side of things.
We definitely have a much stronger community around us, and of course we're open source as well, and I'm sure that's probably what's going to fuel us on for a lot longer.
But as I say, we're not scared of working in an industry with other options. We definitely welcome it. I think people need multiple choices, multiple options. And yeah, I think it's a great thing.
Jonathan Mosen:
Stay with us because there's plenty more from our guests from NV Access as Access Oncontinues.
Love the accessible shopping experience on Innosearch AI? Now your next order can do even more. Innosearch AI has teamed up with the National Federation of the Blind to introduce the new Roundup program.
It's simple. When you check out, just select the option to round up your purchase to the nearest dollar. That small change is donated directly to the NFB, helping to fund vital advocacy, education, and support programs for blind Americans. Turn your sense into independence. Enable Roundup at checkout on Innosearch AI today.
Speaker 5:
We've heard promises about accessibility before. What makes this different is the clock. ADA Title II sets a real deadline, April 2026. By then, state and local governments are expected to make their digital documents accessible. That includes the information blind people rely on every day. If you're encountering barriers now, pay attention. This is the moment to advocate locally. Ask questions, document problems, hold agencies accountable.
Time is running out. Many agencies are scrambling to fix years of inaccessible content. Using its augmented media remediation platform, Pneuma Solutions has remediated millions of documents in weeks, not years. This is real movement and it affects you directly. To Understand what's coming and how it affects you, visit www.title2.info. Accessibility is not a privilege. It's a right.
Jonathan Mosen:
This is Access On from the National Federation of the Blind with our guests from NV Access today.
I have seen this one on email lists and social media over the years, and I'd really welcome the opportunity for you to respond to this on the record. On the one hand, there are so many blind people around the world who welcome NVDA because it's free.
And it's certainly in many countries where robust funding just doesn't exist, or even in developed countries where too many blind people fall through the crack, so people celebrate that. It's a game changer.
But there have been some people who've expressed concern over the years that commercial companies like Vispero invest heavily in research and development, and they come up with an idea only to see it copied within a few months by an open source community and given away for free. And actually, Quentin, you mentioned the sound split. This is an example of that.
A long time ago now, Vispero tried software patents, and regardless of their legal enforceability, it was a PR disaster for them and they stopped doing that. But does the presence of NVDA risk stifling commercial R&D and with it significant innovation?
Mick Curran:
That's a really hard one. I mean, because you've got to weigh that up with, I can guarantee you that we don't deliberately go out there and copy JAWS' features. Certainly, on my part, NV Access's part, we do not do that. Never have.
Of course, as NVDA gets bigger and the community gets involved, then of course, then people are putting a lot of hours themselves into some of these features, and we have to think very carefully about whether we should or should not put these features into NVDA. And generally, if someone has done all the hard work and it meets our coding standards and it seems to be it's going to help people, then I think it would be disappointing to throw code away.
But yeah, look, it's really hard because you're weighing up... well, as we've just said, there are hundreds of thousands of people around the world, probably a lot more. When we really look into developing countries, it doesn't matter what JAWS does, they're not going to get access to it. There has to be a fallback. And so yeah, look, I don't know if someone else wants to have a go at answering that, but...
Gerald Hartig:
Yeah, I've got some thoughts on that one. I think we've really touched on this one already. And it was a very nice segue, Jonathan, from the competitors having multiple players in this space. And really what you've touched on is the core philosophical issue around innovation. There's an organization innovating, it gets released to the market, eventually the market catches up.
This isn't unique to the screen reader area. This isn't unique to the blind and visually impaired. It isn't even unique to software. This happens everywhere. Without alternatives, organizations have very little incentive to innovate. They rest on their laurels.
So in our role as an alternative, even if our role is just to keep the pressure on other screen readers, just by existing and giving the community a means to get these features into the hands of, as Mick mentioned, third world countries developing countries, we are doing a social good.
And the commercial alternatives still benefit from this. Let's say our commercial competitors, every quarter come out with a new feature. Well, there will be a group of people that want early access to those features and will be willing to pay license fees to get access to those features ahead of everyone else. We saw the same thing with the AI boom. Everyone flocked to ChatGPT because they were a little bit ahead.
This is a good thing for the ecosystem, and it is a common thing. It is a natural thing.
Jonathan Mosen:
Do you think that the argument changes any when you talk, for example, about a third party JAWS script developer who might be a single person operation, but who makes their living from making those JAWS scripts, who comes up with innovation to customize a product, and then they find that the very thing that keeps this one individual employed is being copied by an NVDA add-on that's in being given away for free to customize that same application?
Mick Curran:
I'm not quite sure how I'm going to answer this one because that was actually the reason why I started NVDA. Yeah, I'm not quite sure how to answer this because I have a very strong opinion and I don't...on this.
Quentin Christensen:
I think it's also important to note that it doesn't only go one way. I'm quite sure there are features that we've developed that are now present in other screen readers as well. We certainly don't develop by looking at what others are doing and going, "Oh, we're going to copy that."
And in fact, we quite often get requests for features and the main justification is "You should implement this because JAWS has it." We will always push back on that and go, "No, that is not a reason to implement it. What actual problem does it solve? What use does it bring?" And then we might consider whether to implement it.
And even things like AI mentioned before, people have asked us, "Oh, JAWS has got these new AI features that have just come out. What's NVDA doing with AI?" And in actual fact, I can turn around and go, "Well, NVDA has had add-ons which give you that AI functionality for seven or eight years now," which I think was before other screen readers.
So yeah, it certainly goes both ways. And I know from our point of view, what we're trying to do is, as Mick said, ensure that blind users can access all of the features of the computer, and that includes new features as they come out.
And as Microsoft adds new things to Windows, other things come along like new AI improvements. And naturally there is going to be some overlap between ChatGPT comes out with something new, we're going to try and ensure that we support that, or our ad-on developers will, as will other screen readers.
Jonathan Mosen:
So let's talk about how easy it is to get up to speed with NVDA. And Quentin, you mentioned earlier the training material that's available, which is one way that NVDA is funded. And from our perspective, that training material is exceptionally comprehensive and very well written. What training material is available and also what's on board NVDA itself in terms of help that you can quickly access so you can find out what you're doing as you get up to speed?
Quentin Christensen:
Absolutely. So one of the things that we brought in a few years ago now, in the user guide for NVDA, which you can get to from the help menu, it's also online, one of the first parts of that is a quick start guide, which gives you a lot of the things that you might need on day one, the keystrokes, the functions, how to just get it to read something or stop reading something. So that is a really good starting point.
There is then the training material. And while it is a paid offering and something that we sell that does bring in income, we do try and make sure it is accessible to the widest range of people as possible. So basic training for NVDA is the name of our main module.
And despite that name, it actually covers pretty much everything from using NVDA for the first time through to some of the advanced functionality of NVDA, object navigation, the review cursor, configuration profiles, and so on.
And the cost of that for the electronic text version is 32 dollars Australian, which works out to just over 20 dollars U.S. So it is quite reasonably priced, we think. And it is very much designed so that no matter your level and your ability, you can work through it as slowly or as quickly as you need to.
And it will hold your hand with press this key, press that key, press another key. And also then go, okay, do this task more independently so you can review and ensure that you're up to speed as you move along.
And then following on from that, we have a series of modules for Microsoft Office, which go through Word, Excel, Outlook, and PowerPoint. And we also have NVDA expert certification, an exam, an online exam that you can sit. And that exam is free for anyone.
Anyone can sit that exam and test your skills with NVDA to acquire the certification and the actual certificate and listing on the website afterwards. There is a cost for that if you choose, but yet that's another sort of value.
Jonathan Mosen:
So you get told that you would be certified and then you can apply for the certificate. Is that how it works?
Quentin Christensen:
That's right. Yes.
Jonathan Mosen:
Can you talk me through object navigation, which is a very innovative and helpful at times, very helpful way to navigate the screen. And for those who haven't used NVDA before, it might be quite an interesting concept to explore what object navigation is and what its advantages are.
Quentin Christensen:
Sure. So object navigation basically lets you move through the various components that make up the screen. So for instance, you've got the menu bar or the ribbon and the actual body of say the Word document, the status bar at the bottom and other elements. So you can move between those objects. I won't go into detail of every single keystroke here, but you can move between those objects.
And then for instance, on the ribbon in office, you can go between the home and insert and page layout and so on, and then you can go into that ribbon. So I can go into the home ribbon, I've got Excel open in front of me here. I can go into the clipboard section and the font section and the alignment section. And then from there, I can go into and navigate through the different elements in that section.
One feature that we recently introduced in NVDA is the ability to move through a flattened view of that. So instead of having to go in and out of objects, you can just go next object, next object, next object, all the way through the whole screen if you want. Object navigation can be a really useful way of getting to parts of the screen, parts of a program which aren't already accessible just with the regular keyboard functions themselves.
One thing that we look at in the training material, for instance, with the review cursor particularly, but object navigation as well is the up-out Windows box, the Windows version. So if you press the Windows key and type winver, it tells you what version of Windows you've got, but if you were to press tab or the arrow keys to try and move around that, you're on the okay button and I think the link to the license.
So that's a fairly simple example where you can use object navigation to move through those sections to find out what version of Windows you've got and what build. You can also use, so the review cursor is kind of tied to object navigation.
Where object navigation lets you move between the main sections and the main parts, the review cursor will let you look at the text in the current object. So you can, for instance, select and copy the build of Windows that you've got if you need to give that to somebody for troubleshooting purposes.
Mick Curran:
I think a great analogy for object navigation really is imagining the entire Windows operating system as a tree view. I think most people should be familiar with it with the tree view control. I mean, the keyboard commands are slightly different, but the model is the same. It's about drilling down from the top of the operating system.
So the top being the desktop, and then you have all the, at the next level, all the applications. And then inside each application, you might have the main areas, as Quentin said, like the ribbon and the document and the status bar.
But inside each of those, each individual control, and maybe even inside those individual controls, there might be little subcontrols. And so essentially, yeah, it's a tree structure of the entire operating system.
Jonathan Mosen:
I'm thinking of people who might be listening to this and wondering, so what's the difference between object navigation in NVDA and say the way that I would interact with different items on my Mac? Are there similarities?
Mick Curran:
Very, very similar. Yes. Object navigation came from a much older concept. I think the first screen reader to do it was like Virgo 4, which was an old open source Linux screen reader from like 2002 or something. So yes, it's very similar.
I think the Mac one is probably a little bit more limited in some of the objects it can get into. I'm not as familiar anymore, but yes, it very much does feel like the Mac. And then as Quentin said, though our flattened view as well, where you don't have to keep on going in and out of objects, that's a lot more like how you would flick left and right on iOS.
Jonathan Mosen:
Purchasing training material and the certification are all helping to fund the NV Access process and you get something back for that, which is wonderful. I know that a lot of people also donate to NVDA kind of on the concept that people give what they can for those perhaps who aren't in a position to give so everybody has access to quality screen reading. How much of your funding is individual donations like that versus corporate or government support?
Mick Curran:
It's fair to say that we don't really have any kind of government support whatsoever, never really have. We do receive corporate support. I can't be sure on exact numbers, but I mean, we're really talking around, I think it's like 40 to 50% or something like that is corporate funding from the big techs.
And then there's other philanthropic funding that is provided to us in Australia and abroad. And then yeah, of course we have the donations from individuals. And to be fair, the very small amount we get from selling some training material.
Jonathan Mosen:
It seems like you do have some very generous donors, right? There are some people who really believe in this. They believe in what NVDA stands for, and they're currently in a position to donate quite generously and do.
Mick Curran:
Yeah, we're extremely appreciative of that. That is very true. Some people are able to help us out quite a lot, and we are very thankful for that.
Jonathan Mosen:
Transparency is a big deal for NV Access. To the extent that you've started publishing an NVDA roadmap for 2026 with short, medium, and long-term priorities, and anybody can go to the website and look that up, why did you decide to take that approach?
Gerald Hartig:
Well, that was in part so that we could bring the community on with the journey, and also to get some early feedback from the community. As we've touched on, we're very much a community-driven project, but where NV Access's role fills the largest gap is really the long-term plan for NVDA.
So our focus really is on making sure that NVDA is a successful, popular, and useful screen reader, not just this year, but in five years, in 10 years, and hopefully in 20 years. But roadmap is part of that process. So people can start to get excited about the features that are coming, they can give us some feedback on it.
And to allow that conversation to be more of a two-way conversation so that people understand, well, what do NV Access actually do? What are they thinking about? Where do they see the screen reader going in the future?
Jonathan Mosen:
Do you consider that roadmap very much binding, or would there be circumstances where something came up that might say shift your short-term goals?
Gerald Hartig:
Very much the latter. We try to be as responsive as possible to the community. So for example, if a significant security issue pops up, then of course we need to move our resources into addressing that as quickly as possible, which then shift everything else around.
And in some cases, the feedback that we get from the community is a little bit, "Why are you focusing on that? Why can't you focus on this over here?" And so it is very much a living document. It drives directly from our internal planning processes where we've got our own internal roadmap, which goes into a fair bit more detail than what we provide.
That's probably about a quarter of what we have planned at the moment that is on the public roadmap, because a lot of the other things are also internally facing.
We've recently rebuilt our entire server infrastructure to make it more resilient both to cyber attacks, but also to enable it to scale and to allow us to be able to develop on it a lot faster. And things like that don't make it onto the roadmap, but they do sometimes cause us to deviate from that.
So when it comes to the roadmap, we tend to update that about every six months to say, "Well, okay, given the current context that we're operating in, given what we need to get out and given what the community's interested in, these are now our new short-term, medium, and long-term goals." And of course, it gives us an opportunity to tick off all the things that we've delivered in the meantime.
Jonathan Mosen:
What's the best way for people to have input into those goals?
Gerald Hartig:
If someone does have a specific issue with an NVDA feature that's been delivered, of course, the best way to raise that, particularly if it's a bug or a feature request, is to go via our GitHub.
But also we have a very active user group that we can provide the link to. And NV Access staff are on that user group every single day responding to requests from the community and clarifying issues. Quentin is very active there as are I think most of the team.
Jonathan Mosen:
Braille is on the roadmap as well, and there's a real Braille shift going on right now with multi-line devices coming on the market. And also NVDA is on HumanWare's new BrailleNote Evolve. How would you describe the state of Braille in NVDA at the moment, and what are your immediate priorities in terms of developing Braille support?
Mick Curran:
I think it's fair to say that we have very robust but basic support for Braille. You can definitely get the job done if you rely on either Braille in speech or just Braille alone. We will be transparent and say it doesn't have all the different Braille modes and things that perhaps you might be used to with some of the other screen readers, and we definitely get a lot of feedback about that.
But really though, we definitely have many, many people who rely entirely on Braille who successfully use NVDA, but yeah, it might be slightly different to some of the other screen readers.
And we're very excited to see what's coming out on the market hardware-wise in regards to multi-line Braille, et cetera. And we definitely want to see what we can offer there, but at the same time, these multi-line devices are still extremely new and very expensive.
And so we want to be careful as a project, certainly as NV Access, not to put all our time and resources directly into that and then sacrifice all the work that we could be doing that would help the vast majority of our users.
That's not at all to say that we don't care about multi-line Braille. We absolutely do, and it's a really, really exciting time. Even for me personally, I'm very interested in it and have already done a lot of prototyping and we can already see support for some of these multi-line displays in NVDA to a certain extent.
But yeah, we do need to be very careful with the resources that we have. And not everyone is going to get access to these multi-line devices. Again, they are very expensive.
So yeah, we just need to be realists here as well. It's really exciting where the industry, where the blindness community, where the possibilities are going, and we want to keep our finger on the pulse there, but we're not going to drop everything either just because these multi-line displays come out either.
Jonathan Mosen:
But I presume if there's a multi-line Braille display user who can code and is really passionate about making NVDA work to the max with a multi-line device, they could write the code and submit that.
Mick Curran:
Absolutely. Absolutely. And we really, really encourage that, and that's already being done. And in fact, we want more of that. We want more Braille coders. That would be something really great to the project to see. Yes.
Jonathan Mosen:
Speaking of coders, what impact, if any, do you think tools like Claude Code will have on the speed of NVDA development?
Mick Curran:
Already internally in NV Access, we use coding tools like GitHub Copilot within coding environments like VS Code. And we've been using these tools, certainly all of our developers have been using these tools for several years now.
That really has sped up a lot of our process in terms of just writing a lot of the little fiddly bits. We know what changes we need to make, but there are constant changes we need to make to files, et cetera. That's really sped up the process.
But I do want to, right from the outset, say, and this is just a general sort of AI thing, that just because we do use AI doesn't mean we just accept the code that it gives us. Not at all. Anytime that you use AI for coding, unless you can understand each and every line of code, it shouldn't go into the product. And that's a very strong policy of ours.
You need to see AI as an assistant, I guess someone sitting with you, it's a bit like paired programming I suppose, but at the end of the day, you need to understand it all. You need to review it like you would any other type of code. Don't assume the AI is correct.
And there are many instances where it might get it wrong or there might be a typo or something like that. And so yeah, you've got to read it each and every line. So it does speed you up in some ways, but it's not absolute magic either.
That is, of course, getting it better and better. And as you mentioned, Claude code, et cetera. I haven't personally tried some of this. Some of the really newer stuff. It's very exciting. And certainly outside of NV Access, we're now starting to see a very rapid increase of people writing NVDA add-ons who perhaps before AI did not have the coding skills or did not want to start learning the coding skills, I guess, to do so.
And they're now coming out with some pretty spectacular add-ons to solve their own needs, some of them small, some of them large. It's really exciting to see what's going on there in the add-on space for NVDA with this sort of vibe coding.
I mean, it personally doesn't fit too well with me as a coder from a philosophical sense, I guess, but it's something that I'm just going to have to get used to. I feel a bit old when I look at this kind of stuff.
But at the same time, if it's solving problems, we are now hearing of people, as I say, they may not have had coding skills, but they're writing add-ons, which in one case I can think of a very, very concrete case where they were able to access some call center software of their own, bespoke call center software within their organization, and they were able to solve their need by themselves, and it worked. And that's really, really exciting.
Jonathan Mosen:
Let's talk a bit more about add-ons because it is such a useful and fun part of the product. For those who haven't used NVDA, you can browse the add-on store.
I guess it's similar to the app store or the Google Play Store in some ways, and you can download and install various things that give NVDA different, new, modified functionality. And it really creates a sense of community and extensibility.
Now, sometimes they break after an NVDA update and some users will choose not to disable add-ons that are flagged as incompatible, so that could contribute to instability. Why do add-ons break after some NVDA releases? And is that likely to improve or is that something users just have to get used to?
Mick Curran:
I think I'll talk about the old bit first or the history of this. When we created add-ons in NVDA way back in 2012 or something, we essentially allowed add-ons to have absolute full access to the internals of NVDA. From a coding perspective, essentially your add-on code is running within the same space with all the rest of the internal code.
So if you really know what you're doing, you can tweak any part of the internals of NVDA. But with power it comes great responsibility, it's very, very easy to break things with that kind of power. But at the same time, anything is possible. If you really know what you're doing, you can make NVDA do anything.
But it doesn't protect the user very well. We'll certainly admit that. And as NVDA has become much more popular, especially in corporate environments, et cetera, we are seriously putting a lot of effort into looking at how we can restructure our add-on system so that it becomes much more robust and has a surface that will be much less prone to breakage, because you asked about why add-ons break.
Well, yeah, the reason for that really is because add-ons have full access to the internals of NVDA. And those internals will and must change as releases go on, because in order to fix a bug, you've got to change some code, got to move some code. You can't avoid that.
But the problem is that those add-ons have not been coded against a surface that we can promise will never change. And so that's something that we are, as I said, putting in a lot of effort into coming up with a completely new solution where that surface that add-ons are coded against will not break over releases.
Jonathan Mosen:
So it sounds like a form of sandboxing in some ways, but would that potentially make add-ons less powerful?
Mick Curran:
Yes, it will make them less powerful, but I want to be clear here and say, but in some ways they were too powerful. I think it's really about then as we develop this sandbox and develop this sort of more limited API about understanding what the real needs of the add-ons are.
And obviously we want to be on par or greater than other screen reader offerings, of course, as in we want to make sure that people can get their job job done. But at the same time, in order to do that, I don't think they need the keys to the office, which they don't need absolutely to be able to access every tiny little piece of code running in machines.
Jonathan Mosen:
Should we expect a bit of a bumpy ride as NVDA transitions to 64 bit in terms of perhaps quite a few add-ons that aren't being actively developed, they might be abandonware but they've been working okay, but there might be a little bit of a rocky road for a while, right?
Mick Curran:
That's true. Certainly, with the move to 64 bit, which I will admit has been a very strong ask from the community that we move to 64 bit, and so we have finally done that. But yeah, one consequence of that is some add-ons may break, especially those that rely on third party DLLs or things like that, which do need to be themselves converted to 64 bit or may not even be available to do so.
But the beauty of add-ons, because add-ons themselves, at least are mostly open source. And so if something is abandonware or assuming that the license is okay, there are other people in the community who can fork it or take it over or move it in a different direction, update it for 64 bit. And we are definitely seeing that happen at a very rapid rate.
Quentin Christensen:
One of our community contributors a few months ago, actually, it was October, he wrote this and I haven't got it in front of me so I'm working off my head here, one of our community contributors analyzed the available add-ons, and yeah, there are close to 300 just in our store, 280 odd in our store.
And found that only, I think it was 13% of them had code which was not compatible with 64 bit, so they certainly would need to be updated. And that was in October, and that's an impressively low number for such a fundamental change to the code base.
So I think it's easy when we put out beta 1 of the 0.1 release. So one important point actually is that when we break that add-on compatibility, it's for the 0.1 release of each year, the first release of each year. That's really the only one that you need to be worried about.
When we go from 0.1 to 0.2 to 0.3, et cetera, any add-ons that you've got that are working with 2026.1 will work in 2026.2 and 2026.3. So it's just making them compatible with the first release.
And usually by the time that final release comes out, most add-ons are updated by then. We always see people who will download the first beta and the first couple of beta, "Oh, none of my add-ons work." No, we know. We expect that.
If you absolutely rely on your add-ons and you cannot use NVDA at all without them, then don't use the beta. Or download the beta and try it as a portable copy or the temporary copy to test out what's new, but know that the add-ons are yet to be updated.
Gerald Hartig:
And the final thing I'd like to add to that is that that sandboxing will in future decrease that breakage. So any add-on that is written to run in that sandbox will not, moving forward, suffer from API breaking releases.
Jonathan Mosen:
So when do you expect that new architecture to be in place?
Gerald Hartig:
Well, some of it's already going to be in 2026.1. It's a soft rollout. We're focusing on speed synthesis and Braille, particularly on secure screens. And as the year progresses, our plan is to add more and more features into that sandbox, more and more APIs that the add-ons can access, until we reach the point where for the vast majority of add-on developers developing for the secure sandbox is the preferred approach because they won't have to worry about API breakages moving forward.
They won't have to worry about making NVDA unstable from poor code. And in fact, we think that the timing for this has come at a great time with the huge surge of interest in people vibe coding their own add-ons, something running in the sandbox will not be able to break NVDA. And with the potentially low quality of code or at least code that hasn't been as rigorously analyzed by a human, this could be a really good combination.
Jonathan Mosen:
You've all been incredibly forthcoming with your answers and your time. And I know that people will be feeling most informed after this conversation. I have a couple of questions.
One final thing for Mick, and then we'll just talk about how people can get in touch with NV Access and maybe download it for the first time if they haven't done that. Mick, I'm curious if you've ever thought if NVDA didn't exist, what would blind access to Windows look like today globally?
Mick Curran:
Look, it was a really interesting time and place back in 2006 or so. I think a lot of things aligned in regards to want and need and interest in technology. And even the few years after that, the willingness for companies, corporates to fund and go out on a limb and blue sky some things.
It was really exciting times back then, and I think that's sort of how NVDA was born within that particular time and place. And I don't think we would've got NVDA off the ground if that hadn't have occurred. So yeah, look, it's really hard to say. I mean, I think we could just be happy where we are today that there are multiple solutions and people have choice.
Jonathan Mosen:
I did want to give any of you, maybe it's a Quentin thing, a chance to just inform people about how they can get NVDA on their computer and try it. It doesn't mean that they need to abandon what technology they're already using.
It can be an extra tool in the toolbox and that can be very useful and people may find that they gravitate over time. So how do people get acquainted with NVDA?
Quentin Christensen:
Firstly, you can download NVDA from our website, which is nvaccess.org, and you can download it from there. It will install and run quite happily alongside whatever other screen readers that you're using. Obviously, don't have multiple screen readers running at once and speaking, but certainly it will work fine with others installed.
One of the strengths of NVDA is you can have it as a portable copy. So without needing to install it at all, you could put it on a memory stick and take it to somewhere that you have access to a computer but not access to installed software. And you can run it there and it will have almost all of the same ability to work.
So nvaccess.org, one of our great strengths, I think, is your ability to reach us. So you can write to us at info@nvaccess.org. You can reach us on social media, we're on all the main platforms.
At the bottom of our homepage is a link to all our various social platforms. We're quite active on Mastodon because our community is quite active there. We're also on Facebook, LinkedIn, Twitter or X.
Jonathan Mosen:
You are quite patient on Mastodon, I have to say.
Quentin Christensen:
Thank you. Thank you. We're here to work with the community. We have a wonderful community. So yeah, it's fantastic to interact with people. And we were talking before about vibe coding and AI coding and things like that, and obviously how that is starting to explode now with NVDA add-ons.
And I realized that's one other thing that sets NVDA apart, is that because NVDA is open source and all of the code is there and available, how do AI models learn? They learn by scooping up all of the open source and freely available information, so they have access to our code.
So I don't know whether that means that it's easier to vibe code an add-on for NVDA than it might be to do a script for another screen reader, but that's one thing that's there.
We have the newsletter, the blog that Gerald mentioned earlier. We put out usually every fortnight. There's a link on the website, but you can also subscribe to that via email and there's a link to that, but you can also go to nvaccess.org/newsletter and subscribe that way. You can also find out about or get emailed when we put out a new version from there as well.
Jonathan Mosen:
It's been an absolute pleasure to have the three of you on Access On. Thank you so much, and I look forward to staying in touch.
Quentin Christensen:
Thank you.
Mick Curran:
Thank you.
Jonathan Mosen:
That concludes this episode of Access On, the Technology Podcast of the National Federation of the Blind. To send in a contribution for a future episode, email us, attach an audio clip or just write it down and send it to AccessOn@nfb.org. That's AccessOn@nfb.org. To keep up to date with Access On, follow us on Mastodon. AccessOn@nfb.social. That's AccessOn@nfb.social on Mastodon.
To subscribe to an announcement only email list about upcoming episodes, send a blank message to AccessOn-announce-subscribe@nfbnet.org. That's AccessOn-announce-subscribe@nfbnet.org.
To learn more about the National Federation of the Blind, visit our website, NFB.org or phone us (410) 659-9314. That's (410) 659-9314. And be sure to check out the nation's blind podcast right from where you heard this podcast.