Dodgy EDIDs (was: What does your OS look like?)

[onlyonemac]

What I'm doing is focusing on the output side of things (and neglecting the input side), where the output is virtually identical (sans speech synth parameters).

For input; it's just events sent to the front end. For example, the front-end might receive an "UP" event, or a "NEXT_CHILD" event, or an "ESCAPE" event. The developer of the front end doesn't have any reason to care if these events are coming from a keyboard or speech recognition or anything else.

The output is not virtually identical. Imagine editing a document with an audio interface along the lines of Siri: I don't want to hear "OK, applying formatting" or "sure, here's a new paragraph" (whatever speed/intonation the voice uses) when editing a document, whereas the lonely truck driver on the road would appreciate a personal touch to the interface to keep him company.

But actually it's not so much about how similar the interfaces are, it's about what subset of those interfaces the majority of users are going to use and what subset of those interfaces blind users absolutely require, and the fact that developers will only implement the subset that the majority of users are going to use and leave blind users behind.

I never said that blind people can't be casual users; what I said is that they don't want to be restricted to being casual users because the advanced features of an application are only available in the graphical interface. I also never said that sighted people can't be regular users of audio interfaces; what I said is that they aren't going to use an audio interface in the same situations that blind people are required to use audio interfaces, so aren't going to place the same demands on the interface that blind users will (e.g. a sighted user isn't going to care if they can't change the font size of their document with the audio interface because they'll do that later when they've got a monitor available, but a blind user who's trying to produce a professional-looking document certainly will care).

There's relevant research into this phenomenon.

I don't see how that's relevant. I think you're refusing to accept that blind people WILL REFUSE to be given an inferior interface, and that developers won't bother to implement things that only blind users will benefit from (i.e. 75% of the audio interface).

If you think blind people don't deserve anything good; then you could just implement a "generic audio front end on top of any visual front end" (a screen reader) yourself instead of using the audio front ends that were designed specifically for each application.

Or how about this idea for the "let's be mean to blind people" camp? Just leave all the work of making anything accessible to the application developer, then look surprised when the application developer doesn't make anything accessible?

[Brendan]

Hi,

What I'm doing is focusing on the output side of things (and neglecting the input side), where the output is virtually identical (sans speech synth parameters).

For input; it's just events sent to the front end. For example, the front-end might receive an "UP" event, or a "NEXT_CHILD" event, or an "ESCAPE" event. The developer of the front end doesn't have any reason to care if these events are coming from a keyboard or speech recognition or anything else.

The output is not virtually identical. Imagine editing a document with an audio interface along the lines of Siri: I don't want to hear "OK, applying formatting" or "sure, here's a new paragraph" (whatever speed/intonation the voice uses) when editing a document, whereas the lonely truck driver on the road would appreciate a personal touch to the interface to keep him company.

Why not try to imagine using a wood saw to drill holes on brick, or using a freezer to cook bread, or anything else that is just as nonsensical as using a "request, reply" interface for something it was never intended for?

It's more fun to imagine a blind person using a keyboard and screen reader to find out today's weather while they are in the bathroom having a shower.

But actually it's not so much about how similar the interfaces are, it's about what subset of those interfaces the majority of users are going to use and what subset of those interfaces blind users absolutely require, and the fact that developers will only implement the subset that the majority of users are going to use and leave blind users behind.

Why would a developer skip something that (sighted and blind) regular users need just because some (sighted and blind) casual users might not use it?

I never said that blind people can't be casual users; what I said is that they don't want to be restricted to being casual users because the advanced features of an application are only available in the graphical interface. I also never said that sighted people can't be regular users of audio interfaces; what I said is that they aren't going to use an audio interface in the same situations that blind people are required to use audio interfaces, so aren't going to place the same demands on the interface that blind users will (e.g. a sighted user isn't going to care if they can't change the font size of their document with the audio interface because they'll do that later when they've got a monitor available, but a blind user who's trying to produce a professional-looking document certainly will care).

There's relevant research into this phenomenon.

I don't see how that's relevant.

It's very relevant, because you're focusing on trivial nonsense like "What if a front end doesn't support feature X that for no sane reason sighted people never use but blind people use frequently while at the same time batman eats a donut made out of goat's cheese?". If something like changing font sizes isn't supported, any (sighted or blind) regular user will complain, just like users complain whenever any useful feature is missing from any application on any OS.

I think you're refusing to accept that blind people WILL REFUSE to be given an inferior interface, and that developers won't bother to implement things that only blind users will benefit from (i.e. 75% of the audio interface).

No; I refuse to accept that blind people are bizarre freaks (and not just normal users that can't see) who need "special unspecified thingamabobs" that other regular users wouldn't want from an audio interface.

If you think blind people don't deserve anything good; then you could just implement a "generic audio front end on top of any visual front end" (a screen reader) yourself instead of using the audio front ends that were designed specifically for each application.

Or how about this idea for the "let's be mean to blind people" camp? Just leave all the work of making anything accessible to the application developer, then look surprised when the application developer doesn't make anything accessible?

I'm treating you exactly the same as I'd treat anyone who says silly nonsense. I'm not being mean to blind people, I'm trying to ensure they don't end up with user interfaces that suck. You are being mean to blind people by trying to ensure they do end up with user interfaces that suck.

I wouldn't be surprised if developers implement the audio interface first (because its easier - 3D graphics is much more complicated than outputting strings of phonetics), and then fork it afterwards so they've got something that works to build graphics on top of.


Cheers,

Brendan

[onlyonemac]

Why would a developer skip something that (sighted and blind) regular users need just because some (sighted and blind) casual users might not use it?

A regular sighted user is not going to use an audio interface, because a regular sighted user will be sitting at a desk with a monitor. It's only for mobile use that a sighted user will use an audio interface, and users (sighted and blind) tend to use mobile devices for casual tasks and keep more advanced tasks for a desktop or laptop. So when we look at the kind of situations where a sighted user will use an audio interface (mobile devices) we can determine the kind of tasks that they will use it for (casual tasks) and thus the range of functionality (basic functionality) and style of interaction (voice commands/two-way discussion with the device) that they will use. If we look at the kind of situations where a blind user will use an audio interface (mobile and desktop devices) we can determine the kind of tasks that they will use it for (casual and advanced tasks) and thus the range of functionality (basic and advanced functionality) and style of interaction (voice command/two-way discussion with the device and keyboard input) that they will use. Of course, considering how many more sighted mobile-device users there are compared to blind desktop users, application developers are only going to bother with the former.

[onlyonemac]

I'm treating you exactly the same as I'd treat anyone who says silly nonsense. I'm not being mean to blind people, I'm trying to ensure they don't end up with user interfaces that suck. You are being mean to blind people by trying to ensure they do end up with user interfaces that suck.

Actually I'm just trying to make sure that (in your fantasy world) blind people get an interface that gives them access to all the features of an application that sighted users have, when we already have that in screenreaders which give blind people access to at least 90% of the features of most applications. You, on the other hand, are depriving blind people of a proper (i.e. not a mobile interface or an interface which provides only a subset of features) by refusing to accept that developers WILL ONLY IMPLEMENT WHAT THE MAJORITY OF PEOPLE WILL USE, and that SIGHTED PEOPLE WILL NOT USE AN AUDIO INTERFACE WHEN ALTERNATIVES ARE AVAILABLE.

[Brendan]

Hi,

I'm treating you exactly the same as I'd treat anyone who says silly nonsense. I'm not being mean to blind people, I'm trying to ensure they don't end up with user interfaces that suck. You are being mean to blind people by trying to ensure they do end up with user interfaces that suck.

Actually I'm just trying to make sure that (in your fantasy world) blind people get an interface that gives them access to all the features of an application that sighted users have, when we already have that in screenreaders which give blind people access to at least 90% of the features of most applications. You, on the other hand, are depriving blind people of a proper (i.e. not a mobile interface or an interface which provides only a subset of features) by refusing to accept that developers WILL ONLY IMPLEMENT WHAT THE MAJORITY OF PEOPLE WILL USE, and that SIGHTED PEOPLE WILL NOT USE AN AUDIO INTERFACE WHEN ALTERNATIVES ARE AVAILABLE.

You're creating a self-fulfilling prophecy.

By ensuring that audio interfaces suck, you're ensuring that only people who don't have a choice use them, which ensures that developers (who are typically driven by economics and don't want to spend time on something that only a small minority use) don't bother, which ensures that audio interfaces suck.

I'm also creating a self-fulfilling prophecy.

By ensuring that audio interfaces are good, I ensure that people who have a choice do choose to use them (where convenient), which ensures that developers (who are typically driven by economics and want to spend time on things that the majority of users use) do bother, which ensures that audio interfaces are good.

Existing OSs prove your self-fulfilling prophecy is very real; but do not prove that my self-fulfilling prophecy can't be a reality. This is what makes the majority of your concerns irrelevant - essentially you're using your experience with existing OSs where developers have no reason to give a crap about blind users as an argument against something radically different where your experience with existing OSs is no longer applicable.

Things like Siri show that audio interfaces are something that huge number of users are interested in (if they don't suck). This alone doesn't prove that my self-fulfilling prophecy can be a reality; but does serve as a strong indicator. Given the complete absence of anything to indicate that my self-fulfilling prophecy can't work I'm willing to take the gamble; mostly because I believe I need to ensure my OS is superior to existing OSs just to have a snowflake's chance of getting anywhere, and that this is one way to achieve "better than existing screen reader that sucks OSs".

The worst possible case is that it's not enough on its own, and I have to add incentives to encourage developers to provide good audio interfaces; which can be anything from providing an easy to use "audio interface development kit" (which is a good idea anyway), to maybe setting up some sort of "code bounty" system (which is a good idea anyway), to blocking/prohibiting apps from being included in the OS's "app repository" if there is no audio interface (which is excessively heavy-handed and hopefully won't ever be necessary).


Cheers,

Brendan

[onlyonemac]

I ensure that people who have a choice do choose to use them (where convenient)

Sighted people will only ever regard audio interfaces as convenient in a situation where using a graphical interface is impractical. My experience with sighted people is that they will never choose anything that is audio-only over something that is visual as well. That applies to everything in life, not just computers. Sighted people that I have spoken to or read about have:

• Refused to listen to just the soundtrack for a YouTube music video when their living-room TV had a cracked screen
• Preferred to fight with watery eyes and bad lighting in the evening when trying to read rather than listening to an audiobook
• Rejected my suggestion of a talking clock when they complained about having to open their eyes to see the time in the middle of the night
• Insisted that I replace the broken power LED on their computer because they can't see if it's on or off, even though they can hear if it's on
• And many more...

Trying to pretend that you can make an audio interface good enough that sighted people will choose it over a graphical one where possible is very, pardon the pun, blind.

[Kazinsal]

As a sighted person with mild vision deficiency (I have corrective lenses for nearsightedness and astigmatism) I don't think I'd be interested in using a speech-based interface with my computer unless my computer is a sophont and I can make small talk in the morning about the weather and how I slept the previous night while it makes my coffee. I want to talk to a computer that can formulate a real, intelligent response, not one that I say "okay google what is the time" a few times until it understands what I'm saying and it goes "the time is blah" in the same monotone, faux-cheery voice that everyone else's phone has. I want to wake up in the middle of the night, roll over, grumble out "what time is it, Artemis" and my house's pleasant AI says "It's currently 2:43." and I say "thanks, push my alarm forward ten minutes, please" and lo and behold my alarm is set forward ten minutes today (but not tomorrow).

My problem with audio interfaces for computers at this point in time is that they are still too primitive in their "intelligence". They aren't human enough. I do not expect this to happen within my lifetime, but if it does, I would be pleasantly proved wrong.

[Brendan]

Hi,

I ensure that people who have a choice do choose to use them (where convenient)

Sighted people will only ever regard audio interfaces as convenient in a situation where using a graphical interface is impractical. My experience with sighted people is that they will never choose anything that is audio-only over something that is visual as well. That applies to everything in life, not just computers. Sighted people that I have spoken to or read about have:

• Refused to listen to just the soundtrack for a YouTube music video when their living-room TV had a cracked screen
• Preferred to fight with watery eyes and bad lighting in the evening when trying to read rather than listening to an audiobook
• Rejected my suggestion of a talking clock when they complained about having to open their eyes to see the time in the middle of the night
• Insisted that I replace the broken power LED on their computer because they can't see if it's on or off, even though they can hear if it's on
• And many more...

I wouldn't want to listen to some awful youtube soundtrack either (although I don't think I'd break my living room TV just to create a lame excuse). If I were in the middle of a good story I wouldn't want to change to a completely different story either. If I didn't want to open my eyes, I wouldn't want to shift my arm/hand to press a button either. If I paid for something that's broken I'd want it fixed too.

Sometimes I do listen to youtube videos (mostly university lecturers) while using the screen for other things (just not soundtracks - I mostly stopped caring about music when I was in my 20s). I would listen to an audiobook in bed before going to sleep if it's a good story and if it was available (I've never seen them anywhere, other than my Grandmother's house - she got them "on loan" from somewhere, not because she can't read a book but more because she preferred not to wear her glasses), and if I had a suitable device that didn't involve earphones (I've been tangled in my sleep before). For computer power LEDs, I've got one computer with a hideous "ultra bright" blue LED that annoyed me so I covered it with sticky tape, and an "always on" server that's noisy enough to make it hard to hear any of the other computers.

Trying to pretend that you can make an audio interface good enough that sighted people will choose it over a graphical one where possible is very, pardon the pun, blind.

That's not what I'm saying. If people can choose between video and audio obviously they'll choose video; but there are a lot of cases where people can't choose video (in bathroom, while driving, while doing work-outs at the gym, while doing household chores, while at work, ...) because video simply isn't an option. For these cases people currently choose "nothing" because currently audio sucks.

For me personally; if I had a small device with earphones and some sort of input (a microphone, or even just 5 buttons on the device for navigation) that was connected via. wifi to a distributed OS (or was powerful enough to be useful on its own), and had good audio interfaces, I'd probably clip on on my jeans in the morning and take it off at night; and could end up using it more than I use my desktop systems because I don't have to be tethered to a desk. When I am at my desk I wouldn't stop using the audio interface, I'd just be using 2 sets of applications at the same time.

For people in general; the potential market is huge. For just one simple example (from this news article); "Australians spent 4.4 hours a week, or 53 minutes a day for a five day working week, travelling to and from work", and "around 78 per cent of Australians travel to work by car". That's about 8 million people that could use audio interfaces daily (but can't use a video interface because they're trapped behind a steering wheel driving). Now consider that (from this web page), "there are currently 357000 people in Australia who are blind or have low vision". Note: I was searching for worldwide statistics, and these were just the first useful search results I got (probably because I am an Australian) - other countries may be different.

Basically; the market for "audio interfaces for sighted people trapped behind steering wheels on their way to work" alone is over 20 times larger than the market for "audio interfaces for blind people".


Cheers,

Brendan

[Brendan]

Hi,

As a sighted person with mild vision deficiency (I have corrective lenses for nearsightedness and astigmatism) I don't think I'd be interested in using a speech-based interface with my computer unless my computer is a sophont and I can make small talk in the morning about the weather and how I slept the previous night while it makes my coffee. I want to talk to a computer that can formulate a real, intelligent response, not one that I say "okay google what is the time" a few times until it understands what I'm saying and it goes "the time is blah" in the same monotone, faux-cheery voice that everyone else's phone has. I want to wake up in the middle of the night, roll over, grumble out "what time is it, Artemis" and my house's pleasant AI says "It's currently 2:43." and I say "thanks, push my alarm forward ten minutes, please" and lo and behold my alarm is set forward ten minutes today (but not tomorrow).

Even if advanced AI did exist; for things like editing text or writing software or reading documentation or creating spreadsheets this sort of interface seems completely useless to me.

My problem with audio interfaces for computers at this point in time is that they are still too primitive in their "intelligence". They aren't human enough. I do not expect this to happen within my lifetime, but if it does, I would be pleasantly proved wrong.

My problem with real humans is that they're too human. You provide clear and unambiguous input, like "What time is it?", and before you know what went wrong you're stuck listening to some rambling story about some guy called Dave (who is a complete stranger) who had a wart removed from elbow and likes pop-tarts, and 20 minutes later you've still got no idea what the time is. The nice thing about computers is they tend to have well defined "causation", in an "if user does X, computer does Y" way.


Cheers,

Brendan

[kzinti]

My problem with real humans is that they're too human.

Wow. Talk about denial and not accepting one's nature. I wish you the best.

[Brendan]

Hi,

My problem with real humans is that they're too human.

Wow. Talk about denial and not accepting one's nature. I wish you the best.

All I'm saying is that I think "computers should be more human" destroys the productivity/efficiency and predictability of computers, which is the very reason people use computers (as tools) in the first place.


Cheers,

Brendan

[onlyonemac]

Sometimes I do listen to youtube videos (mostly university lecturers) while using the screen for other things (just not soundtracks - I mostly stopped caring about music when I was in my 20s). I would listen to an audiobook in bed before going to sleep if it's a good story and if it was available (I've never seen them anywhere, other than my Grandmother's house - she got them "on loan" from somewhere, not because she can't read a book but more because she preferred not to wear her glasses), and if I had a suitable device that didn't involve earphones (I've been tangled in my sleep before). For computer power LEDs, I've got one computer with a hideous "ultra bright" blue LED that annoyed me so I covered it with sticky tape, and an "always on" server that's noisy enough to make it hard to hear any of the other computers.

Personally that's not my experience with sighted people, but to each their own I guess. I still don't envisage audio interfaces becoming that popular, nevertheless, when I'm constantly dealing with people who *insist* on having their router in a place where they can see the lights flashing to indicate the loading of their webpage when the computer doesn't seem to be doing anything. Or like when I said "if you don't want to turn the light on in the middle of the night to go get a glass of water, just put your finger over the edge of the glass so you can feel when it's full" (blind people do this all the time) and they weren't interested, or again when I said "if you're worried about falling down the stairs in the middle of the night, just count the steps" and again they weren't interested.

Basically; the market for "audio interfaces for sighted people trapped behind steering wheels on their way to work" alone is over 20 times larger than the market for "audio interfaces for blind people".

I know that, and that's the problem; you just seem unable to see how that's a problem. It's a problem because that means that application developers will design audio interfaces to be used by people who are trapped behind steering wheels on their way to work, not people sitting at a desk with a full set of computer peripherals but no option other than to use an audio interface.

[Brendan]

Hi,

Personally that's not my experience with sighted people, but to each their own I guess. I still don't envisage audio interfaces becoming that popular, nevertheless, when I'm constantly dealing with people who *insist* on having their router in a place where they can see the lights flashing to indicate the loading of their webpage when the computer doesn't seem to be doing anything. Or like when I said "if you don't want to turn the light on in the middle of the night to go get a glass of water, just put your finger over the edge of the glass so you can feel when it's full" (blind people do this all the time) and they weren't interested, or again when I said "if you're worried about falling down the stairs in the middle of the night, just count the steps" and again they weren't interested.

All of these a cases where it's easy for a sighted user to use their eyes. You're ignoring all the cases where its not easy for a sighted user to use their eyes.

How many people would leave their car at home and use public transport to get to work, so that they can use a laptop? How many people would have a bath instead of a shower? How many would quit their (gardener, house painter, carpenter, truck driver, delivery person, factory assembly line, ....) job so that they can use that time doing their own stuff on a computer?

How many companies would be willing to purchase headsets that allow their (gardener, house painter, carpenter, truck driver, delivery person, factory assembly line, ....) employees to be more productive while they work?

Basically; the market for "audio interfaces for sighted people trapped behind steering wheels on their way to work" alone is over 20 times larger than the market for "audio interfaces for blind people".

I know that, and that's the problem; you just seem unable to see how that's a problem. It's a problem because that means that application developers will design audio interfaces to be used by people who are trapped behind steering wheels on their way to work, not people sitting at a desk with a full set of computer peripherals but no option other than to use an audio interface.

Not being able to use your eyes is 100% identical to not being able to use your eyes, regardless of the reason why you can't use your eyes. "Audio interface designed for sighted people trapped behind steering wheels" is 100% identical to "audio interface designed for blind people".

Trying to pretend that blind users are different to people that can't see is idiotic nonsense.


Cheers,

Brendan

[onlyonemac]

All of these a cases where it's easy for a sighted user to use their eyes. You're ignoring all the cases where its not easy for a sighted user to use their eyes.

How many people would leave their car at home and use public transport to get to work, so that they can use a laptop? How many people would have a bath instead of a shower? How many would quit their (gardener, house painter, carpenter, truck driver, delivery person, factory assembly line, ....) job so that they can use that time doing their own stuff on a computer?

How many companies would be willing to purchase headsets that allow their (gardener, house painter, carpenter, truck driver, delivery person, factory assembly line, ....) employees to be more productive while they work?

I have already agreed that those are all valid use cases for a sighted person to use an audio interface, but I've also repeatedly pointed out that those are all "mobile" situations, not "desktop" situations.

Not being able to use your eyes is 100% identical to not being able to use your eyes, regardless of the reason why you can't use your eyes. "Audio interface designed for sighted people trapped behind steering wheels" is 100% identical to "audio interface designed for blind people".

Trying to pretend that blind users are different to people that can't see is idiotic nonsense.

It's not about the reason why one cannot use their eyes; it's about the way that that reason relates to the situation. In other words, if a sighted person is unable to use their eyes, it's because they're doing something else at the same time so they're probably using a mobile device, whereas a blind person will need to use an audio interface on both mobile and desktop devices. You can't keep insisting that "audio interface designed for sighted people trapped behind steering wheels" and "audio interface designed for blind people" are the same thing, because the person in the former situation is also unable to use their hands and will be simultaneously concentrating on driving their car, whereas the person in the latter situation may well be able to use their hands while sitting comfortably in their office chair.

I really don't know how else to explain this, but your repeated refusal to accept that blind users and sighted users who can't use their eyes are different use cases is idiotic nonsense. Perhaps you would like to give some justification as to why those two use cases are not inherently different.

[Brendan]

Hi,

Trying to pretend that blind users are different to people that can't see is idiotic nonsense.

It's not about the reason why one cannot use their eyes; it's about the way that that reason relates to the situation. In other words, if a sighted person is unable to use their eyes, it's because they're doing something else at the same time so they're probably using a mobile device, whereas a blind person will need to use an audio interface on both mobile and desktop devices.

As I've already explained; a front end receives events, and it makes no difference whatsoever if those events came from speech recognition or keyboard or anything else. Note that this has nothing at all to do with sighted vs. blind and nothing to do with mobile vs. desktop. For example, someone (e.g. a double amputee without any hands) might use speech recognition as their input device on a desktop with a "visual interface".

Basically there's 2 completely different and completely unrelated things. There's "output from OS/app to user", which is one of:

• Video interface
• Audio interface

Then there's "input from user to OS/app" which is one or more of:

• keyboard
• Mouse
• Joystick
• Touchscreen, touchpad
• Speech recognition
• Motion sensors (e.g. Kinect, or head tracking sensors built into a VR helmet, or eye tracking and a web camera, or...)

The OS/drivers convert "input from user to OS/app" into events. An application's front-end only handles events. No front-end has any reason to care what the input device/s are. A front end only cares what the output device is.

If the user uses one type of input device there's 6 possibilities. If the user uses 2 types of input devices there's 6*5 possibilities. If the user uses 3 types of input devices there's 6*5*4 possibilities. If the user uses one or more types of input devices there's 6+6*5+6*5*4+6*5*4*3+6*5*4*3*2+6*5*4*3*2*1 = 6+30+120+360+720+720 = 1956 possibilities.

If there are 2 types of output device, that's 3912 total permutations that are all supported with only 2 front ends.

If someone decides to write a front end for "small screen" and one for "large screen"; then there's 5868 total permutations that are all supported with only 3 font ends.

If 10 separate/unrelated groups of developers write 4 front ends each over the course of 10 years (where some are very different and some are very similar - e.g. "CompanyA's word processor front end for audio" vs. the almost identical "CompanyB's word processor front end for audio") then there's 156480 combinations of "front_end + input device/s".

If another 5 separate/unrelated groups of developers write 2 back ends each over the course of 10 years then you end up with 1564800 combinations of "back_end + front_end + input device/s".

If a user has all 6 types of input devices available, installs all 40 word-processor front ends and installs all 10 word-processor back ends; then they can use any of these 1564800 combinations at any time.

You can't keep insisting that "audio interface designed for sighted people trapped behind steering wheels" and "audio interface designed for blind people" are the same thing, because the person in the former situation is also unable to use their hands and will be simultaneously concentrating on driving their car, whereas the person in the latter situation may well be able to use their hands while sitting comfortably in their office chair.

I can keep insisting that you are completely wrong, because you keep being completely wrong.


Cheers,

Brendan