I could not believe the amount of stupid this story generated here and in the rest of the media. Only 5 posts here referenced the definitive white paper that explains in gloriously gory detail what Apple did to secure the iPhone 5 and later models.
If Apple implemented this encryption system correctly, as described in that document, it cannot be broken*, even with a custom iOS image, because all key material and control over the internal parameters, preferences, and machine-state of the Secure Enclave are dependent on iOS tossing the correct user PIN/PASSPHRASE over the wall to that chip. Until that is done, the only way to decrypt the storage on that iPhone is to brute force AES-256.
IMO: The FBI is pursuing a Hail Mary and the judge is buying it hook line and sinker**, because they are even more ignorant that the FBI and the rest of of us here about how this security system works. If anything this will be used by the Gov. to attempt to stir Legislators to get backdoors mandated. As anyone with half a functioning braincell knows, mandating such is pure, undiluted, stupid! And I think the Gov. knows this. I think they simply don't give a shit.
*without a truly heroic effort from top shelf hackers, who make absolutely zero mistakes in their execution.
**I don't think this judge believes that Apple did what that White Paper claims they did. It will be interesting to see if Apple can prove that the system is implemented correctly, because I think that will be a key factor in how this all settles out, and what comes next.
OTOH, many model aircraft used to use methanol for fuel, so it not totally unreasonable as a drone fuel.
Mixed with a fairly significant proportion of Nitromethane. Methanol is not good enough, by itself, for even model engines. The stuff I use has about 16% Nitromethane in it.
Previous commenters noted that Nitrogen ices don't behave the same way water ices do. It is more like a thick fluid. Since it is denser than water ice, the water ices floats on the Nitrogen slushy,
Sorry, KGIII, reading a few articles about the crazy shit that goes on in EVE doesn't give you any notion of what the game is really like. While there are some notorious Alliances and Corps that are out and out trolls the vast majority of players just like to play rough. The way EVE is structured all of that rough and tumble play, the meta-gaming, the scheming and theory crafting make the game a hell of a lot more fun than just about any other game I have played.
I don't know of any other game where literally thousands of players are interacting - live in realtime - on the same server cluster, engaged in coordinated operations that can and do have profound influences on the entire game, and how it is played. CCP (EVE's creators) have had to learn along the way too that emergent game play does not come from restricting play styles and beefing up the TOS.
The scamming, backstabbing and other interesting stories that come out of EVE are mostly operating within the games rules: There are really only a few things that gets CCP swinging the ban-hammer: Real Money trading, hacking the client - server protocols, exploiting bugs in the game logic, and illegal harassment of players (DOXING etc).
EVE's core players take extra joy in making cheaters miserable once their MO is discovered and communicated through the galactic grapevine.
I'm actually developing a system that returns to the fundamentals because I'd like to see kids get experience developing code and hardware that they can actually get their heads around.
32 bit core... could be ARM or MIPs doesn't matter, simple core is best NO fancy SoCs... Simple I/O subsystem to ease understanding and enhance hacking. 1 FPGA to support the custom 'OG' video system that allows [Vic-20/C-64/ATARI/Amiga]-like hacking of 1080P with cheap(ish) hardware. Tokenized BASIC, VB-style, with support for external functions a full 2-pass macro assembler (for said external functions) CLI with disk/system operations. An old school expansion bus using modern low cost connectors. Prototype version will replace the motherboard in a TRS-80 Mod 100, TRS-80 Mod 102, or NEC PC8201A Eventually it will have its own case and form factor.
I have two fully functional mod 100's and a NEC PC-8201A. While they are not daily drivers, I am in the process of reengineering them to be more capable, and still support the old firmware. The goal is to create a system that replaces the MOBO and the antiquated I/O with modern equivalents and still supports all the core features of the system, as well as enabling more modern interfaces.
The Mod 100 is one of the toughest laptop designs out there... I used one for taking notes in Olympic National Park during my stint as a volunteer back country ranger in the mid-90's. It easily handled running for 10 days on one set of batteries (4 AA) and storing all of my daily field notes. It was light and happily survived getting banged around in my field kit. It also easily interfaced with the aging PC XT that the ranger station used for storing our field notes. Simple serial cable interface to DOS on the host PCXT allowed me to stream my notes into a text file and then copy them into the logging system that the park service used to record rangerly activities. The other rangers had a rough time comprehending my madness at first. It took less than a half hour to finish my tour by filing my reports in the station. It took the others hours to do it by typing in their hand-written field notes. I think some of them started to see the light.
The keyboard alone is an amazingly beast that few modern keyboards can beat for feel or reliability even under harsh conditions... Anyone who has spent 6 months touring the back country in the Olympics for 10 days out of 14 at a time knows what I am talking about.
It is easy to argue that all of the ideas in the article are not new.
What I found interesting is that rather than reinventing the wheel with custom built functions this author took advantage of library element that are already in the game engine. Using the physics engine is only a heavyweight solution if you don't have a preexisting API for handling physics. Since all modern game engines come with these tools it makes sense to adapt the crusty old process of generating mazes to the tools that are right in front of your nose.
And all you fellow graybeards kvetching about this kid's solution... get stuffed. He's doing something and sharing his solutions./rant
The first 600 miles of highway 5/1/101 in California is over 330 years old. It has been continuously maintained the whole time. There are even sections of it that have been bypassed to preserve it as the Spanish Mission Chain left it.
The splice would be visible once the decoy cuts were patched. If the fiber crew is doing their job, they will see a new blip of attenuation and reelection loss right where your tap-splice is. It is good practice to maintain records of OTDR for every fiber run so that changes in the run can be spotted. Now, maintaining these records and actually looking at them when something weird happens to the network, is a different issue. Detecting a tap is trivial for someone who knows what they are doing.
Apple ][+ had an OS that was 4KB, it was a boot loader too. The default behavior was to launch the 8KB AppleSoft interpreter. If you happened to have a Disk ][ unit installed it would digress into loading 6KB of AppleDOS into RAM from a 2KB ROM on the DISK controller before seeding the low level globals of AppleSoft to patch interpreter's command dispatcher to handle DOS commands from the CLI, and then calling into the interpreters main loop. Kids these days.... No appreciation for how much quality work a few KB of assembly language code can do.
The largest Atari 2600 games were less than 10KB, most were 4KB, until bank swapping became popular and that included ALL game assets! Most NES games were less than 16KB of executable code.
Integer BASIC on the Apple ][ (not the Apple ][+) was written from scratch by Woz. It is a mutant implementation because it used what we now call a VM embodied in Sweet16, a 16 bit virtual processor. Most of Integer BASIC's core functions were written in this interpreted 16 bit VM. Sweet16 was designed to allow a programmer to easily blend Sweet16 instructions with native assembly code with almost no overhead.
Integer BASIC was an order of magnitude faster than AppleSoft BASIC. It was often used as the executive in video games of the era on Apple ][+ and later machines with Apple Disk ][ drives. Integer BASIC could be loaded into High-Memory or into the Microsoft designed 16K Extended Memory Card that had the ability to bank out the System ROMs. This allowed games developers to leverage a lot of the code in Integer BASIC for their games. It also made reverse engineering such games difficult as calls into the Sweet16 VM would completely confuse most disassemblers of the era. A key reason that Integer BASIC was used this way was that XREF source code for Integer BASIC and Sweet16 were published in early Apple ][ Reference manuals. The XREF source for AppleSoft was not readily available until much later, and even when it was, leveraging it as a runtime library was very difficult.
The Apple ][+ and later Apple//e,//c,//gs models used AppleSoft BASIC, (Floats for all numeric processing) which was written by Microsoft and licensed to Apple. Microsoft later used the license for AppleSoft BASIC as a bludgeon when Apple tried to develop a BASIC interpeter for the Macintosh that would have supported building stand-alone Apps. Apple was threatened with losing all rights to AppleSoft BASIC if they continued development. MacintoshBASIC was quietly suffocated in its crib, and Microsoft quickly announced Microsoft BASIC for Macintosh, an early implementation of what eventually evolved into Visual Basic (BASIC with labels instead of line numbers).
Reference: I was an in-house contractor at Apple SQA as a test engineer from mid-1984 until 1988, and then again from 1990 through early 1992.
Are extremely efficient for their mass and volume. The key issue they must overcome is cooling. If the motor is in the liquid stream, (and this is a likely assumption) cooling is damn near free. As for the power supply, I am fairly sure that the reduced mass, complexity of the pump, plumbing and associated benefits with reducing the volume of the pump system that must survive extreme temperature and pressure, more than make up for the battery mass.
Those that doubt the tech will scale up to larger designs don't really understand how flexible BLDC motor configurations can be. Comparing a custom engineered BLDC motor application with a stock industrial motor, even a stock BLDC, is simply a waste of effort. Modern BLDC hobby motors can easily produce 16 HP/kg with air cooling for intervals similar to what these turbine pumps need to do at efficiencies up to about 95%. I'm sure they do much better when liquid cooled. The battery tech is not quite as good at scaling yet, so some kind of fuel cell might be needed to produce the required electricity, but I could see this approach working for pump designs that scale to what Falcon 1 can do now with gas powered turbines.
TL;DR: Reducing the pump to an integrated Impeller/motor result in enough reduction in mass and complexity that using primary batteries to power the pumps is a net gain in lift capacity compared to a traditional gas turbine pump design.
The other key for VHS adoption: it was a consortium standard, not a Sony exclusive. All the consortium partners cross-liscensed their relevant patents. That led to a diversity of VHS compliant and compatible systems that trumped Sony's exclusive offering. The only real technical failure of Beta was the 1 hour recording limit.
Hey... maybe you just need a math class
on
Go R, Young Man
·
· Score: 1
How about after teaching Algebra, but before heading off into Trig and Calculus spend a quarter or two on Discrete Math. Right there in one fell swoop is the foundation of all computer science in one fairly easy to grasp branch of mathematics. Tie that in with a bit of PASCAL or Python and there is your wade into programming without forcing students into a CIS track. For budding programmers, that class would tie their foundation together, and with others it just gives them some insight into how a computer or even just a calculator really works.
Orienting the study around languages only works with people who already have some idea of how to program.
If you weren't so cheap, you could have been buying computers not covered in crap for years. Apple has never sold computers with crap like that on it.
The problem is, you want to pay $100 for a $2000 device and ignore the consequences.
Lenovo hasn't actually done this yet, and when they do, they won't be the first.
[emphasis added] This is simply not true. Apple included 3rd party crapware on their computers just as often as the PC OEMs did until the second coming of Jobs.
This goes back to some ideas Neal Stephenson was scratching at in Anathema.
Humans know that the symbols their minds operate on MEAN something. Our minds operate on the meaning of symbols not the symbols themselves.
No machine she have ever built has anything close to such a capacity. In Stephenson's parlance such machines are only capable of parsing syntax to operate on it, and transform it into more syntax.
We have a long way to go before machines can understand their instructions, and thus operate on meaning, rather than syntax.
They are focused on biochemistry, biology, medicine, audio engineering, geology, architecture, marine biology, genetic engineering, truck mechanics, electrical technicians, etc. I have only met a few that were hip to CS in any real sense. Those few were either UI designers/developers (yes they wrote code), or Psychology majors with an interest in UX (didn't write code for development, but did write code for processing their study data).
Stop the pigeon-holing! Many disciplines encourage their practitioners to learn coding, indirectly! HOWEVER, most of the specific personages I reference indirectly above were coders because their disciplinary goals required that they learn to write code on their own, NOT because their discipline requires it, but because their projects indirectly require it! Having a strong coding background myself I was in a position to offer useful advice on where they could refine their skills to achieve their disciplinary goals with code. It really had very little to do with pure CIS... It had to do with getting a pie-eating-contest off their plate so they could move on to the rest of their research!
Code development is not an end. It is a means to accomplish a disciplinary goal that cannot be solved by any other method, and all of the academic and professional women I have met dealing with that particular challenge grok that. They don't love computers, but they know what such machines can do for them, and they are more than willing to roll up their sleeves to learn it, and keep their hands dirty in it, only so long as that effort is required to satisfy that portion of their larger goal.
SO:
In K-12 teach Discrete Math, Data Structures, Algorithms and call it good. If you must get feed back, teach them a core language like ANSI C. But let it go after that. Make it a required section of every math class, scale it to fit the level expected for that student. By the end of k-12 every student should know what an integer and a float is, what a pointer is, what arrays and strings are, what control statements are, and why they are important for computer processing. This should be true even if they never want to see a line of code again. Put it where it belongs. Programming is applied math. Treat it as such. Don't bury women in a discipline that does not interest them. Teach them how they can apply Discrete Math and automatic computation in their own path.
Coda: Women generally won't spend 20 years studying one species of dragonfly -- guys have, and do similar things all the time. That is a guy thing.
Women, if they focus down on something like fruit fly larva, it is generally for a greater purpose that tends to broaden their efforts not narrow them. That is a key difference between guys and gals. Guys are content, for a lot of bad, or ill-defined reasons, to narrow down to a laser fine focus on shit that most women just can't be bothered with. It is not that women can't do it, or won't do it! I think it is more that they recognize it is usually self-destructive and self-limiting. I think the feminine approach is healthier, for a lot of reasons.
First time I saw blind programmer was a professional C/PM App developer back around 1982. He used a minimalist text editor and a Votrax speech speech synth as the print device. I was in high school at the time and said programmer was the father of a fellow hacker in my Pod of computer, theater, and gamer nerds. Seeing (and hearing) this kids father program in C and Fortran was fascinating.
You want to accommodate blind developers? Give them CLI access to every element of the system, with consistent shortcuts and high quality voice//auditory feedback for navigation in screen-format editors.
The only reason this particular developer had a CRT is because it was an integrated part of the serial terminal he used to communicate with the host system.
Slashdot Mirror
I could not believe the amount of stupid this story generated here and in the rest of the media. Only 5 posts here referenced the definitive white paper that explains in gloriously gory detail what Apple did to secure the iPhone 5 and later models.
If Apple implemented this encryption system correctly, as described in that document, it cannot be broken*, even with a custom iOS image, because all key material and control over the internal parameters, preferences, and machine-state of the Secure Enclave are dependent on iOS tossing the correct user PIN/PASSPHRASE over the wall to that chip. Until that is done, the only way to decrypt the storage on that iPhone is to brute force AES-256.
IMO: The FBI is pursuing a Hail Mary and the judge is buying it hook line and sinker**, because they are even more ignorant that the FBI and the rest of of us here about how this security system works. If anything this will be used by the Gov. to attempt to stir Legislators to get backdoors mandated. As anyone with half a functioning braincell knows, mandating such is pure, undiluted, stupid! And I think the Gov. knows this. I think they simply don't give a shit.
*without a truly heroic effort from top shelf hackers, who make absolutely zero mistakes in their execution.
**I don't think this judge believes that Apple did what that White Paper claims they did. It will be interesting to see if Apple can prove that the system is implemented correctly, because I think that will be a key factor in how this all settles out, and what comes next.
OTOH, many model aircraft used to use methanol for fuel, so it not totally unreasonable as a drone fuel.
Mixed with a fairly significant proportion of Nitromethane. Methanol is not good enough, by itself, for even model engines. The stuff I use has about 16% Nitromethane in it.
Previous commenters noted that Nitrogen ices don't behave the same way water ices do.
It is more like a thick fluid. Since it is denser than water ice, the water ices floats on the Nitrogen slushy,
the vast majority of EVE players do not live in iceland and have never been there.
Sorry, KGIII, reading a few articles about the crazy shit that goes on in EVE doesn't give you any notion of what the game is really like.
While there are some notorious Alliances and Corps that are out and out trolls the vast majority of players just like to play rough. The way EVE is structured all of that rough and tumble play, the meta-gaming, the scheming and theory crafting make the game a hell of a lot more fun than just about any other game I have played.
I don't know of any other game where literally thousands of players are interacting - live in realtime - on the same server cluster, engaged in coordinated operations that can and do have profound influences on the entire game, and how it is played. CCP (EVE's creators) have had to learn along the way too that emergent game play does not come from restricting play styles and beefing up the TOS.
The scamming, backstabbing and other interesting stories that come out of EVE are mostly operating within the games rules:
There are really only a few things that gets CCP swinging the ban-hammer: Real Money trading, hacking the client - server protocols, exploiting bugs in the game logic, and illegal harassment of players (DOXING etc).
EVE's core players take extra joy in making cheaters miserable once their MO is discovered and communicated through the galactic grapevine.
o7
I'm actually developing a system that returns to the fundamentals because I'd like to see kids get experience developing code and hardware that they can actually get their heads around.
32 bit core... could be ARM or MIPs doesn't matter, simple core is best NO fancy SoCs...
Simple I/O subsystem to ease understanding and enhance hacking.
1 FPGA to support the custom 'OG' video system that allows [Vic-20/C-64/ATARI/Amiga]-like hacking of 1080P with cheap(ish) hardware.
Tokenized BASIC, VB-style, with support for external functions a full 2-pass macro assembler (for said external functions)
CLI with disk/system operations.
An old school expansion bus using modern low cost connectors.
Prototype version will replace the motherboard in a TRS-80 Mod 100, TRS-80 Mod 102, or NEC PC8201A
Eventually it will have its own case and form factor.
All of it as open source as possible.
I just learned it by rote... there was no mnemonic in my basic electronics class in high school.
Bk Br R O Y Gn Bl P Gy W (Gd / S)
And the other common color sequence learned by rote:
o/w O g/w BL bl/w G br/w BR.
Wizardry: Proving Grounds of the Mad Overlord by Sir-Tech, was written in Apple UCSD PASCAL.
I have two fully functional mod 100's and a NEC PC-8201A. While they are not daily drivers, I am in the process of reengineering them to be more capable, and still support the old firmware. The goal is to create a system that replaces the MOBO and the antiquated I/O with modern equivalents and still supports all the core features of the system, as well as enabling more modern interfaces.
The Mod 100 is one of the toughest laptop designs out there... I used one for taking notes in Olympic National Park during my stint as a volunteer back country ranger in the mid-90's. It easily handled running for 10 days on one set of batteries (4 AA) and storing all of my daily field notes. It was light and happily survived getting banged around in my field kit. It also easily interfaced with the aging PC XT that the ranger station used for storing our field notes. Simple serial cable interface to DOS on the host PCXT allowed me to stream my notes into a text file and then copy them into the logging system that the park service used to record rangerly activities. The other rangers had a rough time comprehending my madness at first. It took less than a half hour to finish my tour by filing my reports in the station. It took the others hours to do it by typing in their hand-written field notes. I think some of them started to see the light.
The keyboard alone is an amazingly beast that few modern keyboards can beat for feel or reliability even under harsh conditions... Anyone who has spent 6 months touring the back country in the Olympics for 10 days out of 14 at a time knows what I am talking about.
To add to the collusion fever:
Notice that the feature was added to the very bottom of the Cellular Data option list.
After all the apps.
It is the feature that Apple's Telecom partners didn't want you to see.
It is easy to argue that all of the ideas in the article are not new.
What I found interesting is that rather than reinventing the wheel with custom built functions this author took advantage of
library element that are already in the game engine. Using the physics engine is only a heavyweight solution if you don't have a
preexisting API for handling physics. Since all modern game engines come with these tools it makes sense to adapt the crusty
old process of generating mazes to the tools that are right in front of your nose.
And all you fellow graybeards kvetching about this kid's solution... get stuffed. He's doing something and sharing his solutions. /rant
The first 600 miles of highway 5/1/101 in California is over 330 years old. It has been continuously maintained the whole time. There are even sections of it that have been bypassed to preserve it as the Spanish Mission Chain left it.
The splice would be visible once the decoy cuts were patched. If the fiber crew is doing their job, they will see a new blip of attenuation and reelection loss right where your tap-splice is. It is good practice to maintain records of OTDR for every fiber run so that changes in the run can be spotted. Now, maintaining these records and actually looking at them when something weird happens to the network, is a different issue. Detecting a tap is trivial for someone who knows what they are doing.
it is probably a boot loader.
Apple ][+ had an OS that was 4KB, it was a boot loader too. The default behavior was to launch the 8KB AppleSoft interpreter. If you happened to have a Disk ][ unit installed it would digress into loading 6KB of AppleDOS into RAM from a 2KB ROM on the DISK controller before seeding the low level globals of AppleSoft to patch interpreter's command dispatcher to handle DOS commands from the CLI, and then calling into the interpreters main loop.
Kids these days.... No appreciation for how much quality work a few KB of assembly language code can do.
The largest Atari 2600 games were less than 10KB, most were 4KB, until bank swapping became popular and that included ALL game assets! Most NES games were less than 16KB of executable code.
Integer BASIC on the Apple ][ (not the Apple ][+) was written from scratch by Woz. It is a mutant implementation because it used what we now call a VM embodied in Sweet16, a 16 bit virtual processor. Most of Integer BASIC's core functions were written in this interpreted 16 bit VM. Sweet16 was designed to allow a programmer to easily blend Sweet16 instructions with native assembly code with almost no overhead.
Integer BASIC was an order of magnitude faster than AppleSoft BASIC. It was often used as the executive in video games of the era on Apple ][+ and later machines with Apple Disk ][ drives. Integer BASIC could be loaded into High-Memory or into the Microsoft designed 16K Extended Memory Card that had the ability to bank out the System ROMs. This allowed games developers to leverage a lot of the code in Integer BASIC for their games. It also made reverse engineering such games difficult as calls into the Sweet16 VM would completely confuse most disassemblers of the era. A key reason that Integer BASIC was used this way was that XREF source code for Integer BASIC and Sweet16 were published in early Apple ][ Reference manuals. The XREF source for AppleSoft was not readily available until much later, and even when it was, leveraging it as a runtime library was very difficult.
The Apple ][+ and later Apple //e, //c, //gs models used AppleSoft BASIC, (Floats for all numeric processing) which was written by Microsoft and licensed to Apple. Microsoft later used the license for AppleSoft BASIC as a bludgeon when Apple tried to develop a BASIC interpeter for the Macintosh that would have supported building stand-alone Apps. Apple was threatened with losing all rights to AppleSoft BASIC if they continued development. MacintoshBASIC was quietly suffocated in its crib, and Microsoft quickly announced Microsoft BASIC for Macintosh, an early implementation of what eventually evolved into Visual Basic (BASIC with labels instead of line numbers).
Reference: I was an in-house contractor at Apple SQA as a test engineer from mid-1984 until 1988, and then again from 1990 through early 1992.
Rarely have I ever seen this much stupid in a single thread on /.
Good job on that.
Are extremely efficient for their mass and volume. The key issue they must overcome is cooling. If the motor is in the liquid stream, (and this is a likely assumption) cooling is damn near free. As for the power supply, I am fairly sure that the reduced mass, complexity of the pump, plumbing and associated benefits with reducing the volume of the pump system that must survive extreme temperature and pressure, more than make up for the battery mass.
Those that doubt the tech will scale up to larger designs don't really understand how flexible BLDC motor configurations can be. Comparing a custom engineered BLDC motor application with a stock industrial motor, even a stock BLDC, is simply a waste of effort. Modern BLDC hobby motors can easily produce 16 HP/kg with air cooling for intervals similar to what these turbine pumps need to do at efficiencies up to about 95%. I'm sure they do much better when liquid cooled. The battery tech is not quite as good at scaling yet, so some kind of fuel cell might be needed to produce the required electricity, but I could see this approach working for pump designs that scale to what Falcon 1 can do now with gas powered turbines.
TL;DR: Reducing the pump to an integrated Impeller/motor result in enough reduction in mass and complexity that using primary batteries to power the pumps is a net gain in lift capacity compared to a traditional gas turbine pump design.
The other key for VHS adoption: it was a consortium standard, not a Sony exclusive. All the consortium partners cross-liscensed their relevant patents. That led to a diversity of VHS compliant and compatible systems that trumped Sony's exclusive offering. The only real technical failure of Beta was the 1 hour recording limit.
How about after teaching Algebra, but before heading off into Trig and Calculus spend a quarter or two on Discrete Math. Right there in one fell swoop is the foundation of all computer science in one fairly easy to grasp branch of mathematics. Tie that in with a bit of PASCAL or Python and there is your wade into programming without forcing students into a CIS track. For budding programmers, that class would tie their foundation together, and with others it just gives them some insight into how a computer or even just a calculator really works.
Orienting the study around languages only works with people who already have some idea of how to program.
The main issue I have with this guys article is that he wants to make a laundry robot out of straw.
If you weren't so cheap, you could have been buying computers not covered in crap for years. Apple has never sold computers with crap like that on it.
The problem is, you want to pay $100 for a $2000 device and ignore the consequences.
Lenovo hasn't actually done this yet, and when they do, they won't be the first.
[emphasis added]
This is simply not true.
Apple included 3rd party crapware on their computers just as often as the PC OEMs did until the second coming of Jobs.
Shaftoe! Shut up about the goddamed lizard!
This goes back to some ideas Neal Stephenson was scratching at in Anathema.
Humans know that the symbols their minds operate on MEAN something. Our minds operate on the meaning of symbols not the symbols themselves.
No machine she have ever built has anything close to such a capacity. In Stephenson's parlance such machines are only capable of parsing syntax to operate on it, and transform it into more syntax.
We have a long way to go before machines can understand their instructions, and thus operate on meaning, rather than syntax.
They are focused on biochemistry, biology, medicine, audio engineering, geology, architecture, marine biology, genetic engineering, truck mechanics, electrical technicians, etc. I have only met a few that were hip to CS in any real sense. Those few were either UI designers/developers (yes they wrote code), or Psychology majors with an interest in UX (didn't write code for development, but did write code for processing their study data).
Stop the pigeon-holing! Many disciplines encourage their practitioners to learn coding, indirectly! HOWEVER, most of the specific personages I reference indirectly above were coders because their disciplinary goals required that they learn to write code on their own, NOT because their discipline requires it, but because their projects indirectly require it! Having a strong coding background myself I was in a position to offer useful advice on where they could refine their skills to achieve their disciplinary goals with code. It really had very little to do with pure CIS... It had to do with getting a pie-eating-contest off their plate so they could move on to the rest of their research!
Code development is not an end. It is a means to accomplish a disciplinary goal that cannot be solved by any other method, and all of the academic and professional women I have met dealing with that particular challenge grok that. They don't love computers, but they know what such machines can do for them, and they are more than willing to roll up their sleeves to learn it, and keep their hands dirty in it, only so long as that effort is required to satisfy that portion of their larger goal.
SO:
In K-12 teach Discrete Math, Data Structures, Algorithms and call it good. If you must get feed back, teach them a core language like ANSI C. But let it go after that. Make it a required section of every math class, scale it to fit the level expected for that student. By the end of k-12 every student should know what an integer and a float is, what a pointer is, what arrays and strings are, what control statements are, and why they are important for computer processing. This should be true even if they never want to see a line of code again. Put it where it belongs. Programming is applied math. Treat it as such. Don't bury women in a discipline that does not interest them. Teach them how they can apply Discrete Math and automatic computation in their own path.
Coda: Women generally won't spend 20 years studying one species of dragonfly -- guys have, and do similar things all the time. That is a guy thing.
Women, if they focus down on something like fruit fly larva, it is generally for a greater purpose that tends to broaden their efforts not narrow them. That is a key difference between guys and gals. Guys are content, for a lot of bad, or ill-defined reasons, to narrow down to a laser fine focus on shit that most women just can't be bothered with. It is not that women can't do it, or won't do it! I think it is more that they recognize it is usually self-destructive and self-limiting. I think the feminine approach is healthier, for a lot of reasons.
First time I saw blind programmer was a professional C/PM App developer back around 1982. He used a minimalist text editor and a Votrax speech speech synth as the print device. I was in high school at the time and said programmer was the father of a fellow hacker in my Pod of computer, theater, and gamer nerds. Seeing (and hearing) this kids father program in C and Fortran was fascinating.
You want to accommodate blind developers? Give them CLI access to every element of the system, with consistent shortcuts and high quality voice//auditory feedback for navigation in screen-format editors.
The only reason this particular developer had a CRT is because it was an integrated part of the serial terminal he used to communicate with the host system.