My 12-yr-old son has been using Emacspeak for about two years now. It was a pain to set up as we are using the IBM ViaVoice TTS engine that was available for Linux for a while (but no more, even back then), meaning that I had to scrounge around and make do with some outdated zip files. The situation was also complicated by the fact that I was trying to use the somewhat broken built-in Via audio; things got much easier when I put in a cheap but authentic soundblaster card. Using a free software TTS engine like Festival wasn't quite ready for prime time back then, but I hope that by now it's a reasonable option.
So far we've just been using Emacspeak & Linux for me to teach him basic programming in Python; it's been great so far. We are just embarking on Emacs Lisp, partly for him to learn a different language, and partly so we can customize emacspeak should we find the need.
FYI, he also uses JAWS on Win98 for web browsing and email (too many $$ to upgrade to the version of JAWS that runs on XP, unfortunately). He dual boots himself; adding a bunch of ctrl-Gs to the LILO prompt string lets him know when he reaches the prompt, and then he just hits 'w' for Windows or 'l' for Linux. I put some ctrl-G's in/etc/issue also so he knows when he gets to the login prompt. His.bash_profile sets some environment variables and runs emacspeak directly. He uses dired to manage his files.
Someday I'd like to get him up with reading email and surfing the web in emacspeak too, but I haven't had the time to get that set up.
BTW, related to another topic, I just heard about Rockbox a few days ago from another Slashdot post, so I bought an Archos Ondio off of Ebay with hopes of getting that working for a Christmas present. (Glad he doesn't read Slashdot himself, though it's probably just a matter of time!)
We have a number of tactile maps from a group called Princeton Braillists. They don't seem to have a web site (shocking!) but their address and phone number are:
76 Leabrook Lane
Princeton, New Jersey 08540
phone: (609) 924-5207
I got this information from this National Federation of the Blind article Tactile Educational Materials: Tips and Resources. There are a number of other sources of materials in there that may be useful.
As the father of an 11-yr-old totally blind boy, I'd strongly recommend having your co-worker get in touch with the NFB. There's probably a local chapter nearby. Getting to know competent blind adults should be a very enlightening and positive experience for your co-worker and his son.
I'd also say that although learning braille may seem like it is emphasizing the fact that his son is different, in the long run being unable to access information will make him more different. Being able to read what others are reading, even if it's in a different medium, may make him fit in more, not less.
FWIW, my son loves his braille edition of PC World. So now he's different because he's a geek...
I'll second this. I am teaching programming to my 11-yr-old totally blind son using Python and Emacspeak on Linux. "Voice-lock" mode is particularly cool: it's like font-lock, but uses different voices rather than different fonts to distinguish keywords/strings/comments etc.
It won't die though because HP need it for the ports of VMS, NonStop and Tru64 (incorporating HP-UX).
Right, this fits into my theory: as long as Intel is obligated by HP to make Itaniums, it can't kill them outright, but it will do what it can to save money in the interim. Even if that goes against the long-term viability of the architecture.
I think this is a sign of Intel cutting its losses... this will certainly save money, but at the cost of being able to differentiate Itanium platforms from Xeon platforms. One of the real distinctions between a high-end platform and a run-of-the-mill machine is memory bandwidth; if both types of chips plug into the same socket, they're going to have access to the same bandwidth, and Itanium will lose one of its few remaining advantages over Xeon.
Intel would only do this if saving money was more important to them than giving people a credible reason to buy Itaniums instead of Xeons... and I do believe that's the situation at this point in time. Not a good sign for the future of Itanium.
The parent is mistaken... the i860 and i960 were pretty much unrelated. The only thing they had in common was that they were both general-purpose CPUs that could have displaced the x86 line but for various reasons never did.
The reason the i860 might be remembered as a graphics processor is that it did have special graphics instructions (think MMX) and very high peak floating-point performance, and after it became clear that it would never cut it as a true general-purpose CPU it lived on for a while as a dedicated CPU on high-end graphics boards.
For those who think the i860 looks pretty nice, consider this: it had a "push" pipeline for floating point, which meant that the pipeline would only advance when new instructions were issued. So if it took four stages to do an operation, the result of the first operation came out when you issued the fourth one. Just to keep things exciting, I believe the result of the first instruction was stored in the register specified by the fourth instruction (so the pipeline didn't have to bother remembering the destination register specified by the first instruction).
It all depends on what you consider makes a processor "64 bit". I think most people would agree that a true 64-bit architecture has 64-bit registers and 64-bit addresses (in the architecture, if not fully present in the implementation).
While it's true that the Cray-1 supported 64-bit integers in 1976, it only had 24-bit addresses (though they were addresses of 64-bit words, so it supported 2^27 bytes of memory). The IBM 360 started out with 24-bit byte addresses, later extended to 31 bits, and extended again to 64 bits only in 2000 (with the ""z/Architecture").
The bottom line is that 4GB of memory was not affordable by anyone until fairly recently in computing history (particularly on Crays, where main memory was typically SRAM!), so 32-bit addresses were not a huge limitation.
I don't know of any machine that had 64-bit addresses before the R4000. Specifically, an architecture where a single user process could address a flat 2^64-byte memory space (or even more than 2^32 bytes). I'm sure if I'm wrong someone will point that out!
I agree with your general lament, but I think you're misapplying it here.
Actually the first 64-bit microprocessor available to the general public was the MIPS R4000, which came out in 1992 (see nice picture). So 64-bit computing is more like 12 years old.
I recall when the R4000 came out that most people thought 64-bit computing was premature (DRAM was quite a bit more expensive then per bit!), but the MIPS philosophy was that people would need time to develop software, and would need hardware to develop software on... so by adding 64 bits "early" then maybe the software would be there by the time people really needed it.
You'd think those blind people would get it through their heads that there are some things they just don't need to do on their own. It's like the blacks in the 60s... they should have just been content with the way things were, and not gone whining to the government to get that overbearing Civil Rights Act passed.
One of the guiding philosophies of the Nat'l Federation of the Blind (nfb.org) is that blindness itself is merely a nuisance; it's the attitudes of sighted people that make it difficult to be blind. I'm sorry to say that most of the posts on this thread strongly reinforce that position.
Slashdot Mirror
My 12-yr-old son has been using Emacspeak for about two years now. It was a pain to set up as we are using the IBM ViaVoice TTS engine that was available for Linux for a while (but no more, even back then), meaning that I had to scrounge around and make do with some outdated zip files. The situation was also complicated by the fact that I was trying to use the somewhat broken built-in Via audio; things got much easier when I put in a cheap but authentic soundblaster card. Using a free software TTS engine like Festival wasn't quite ready for prime time back then, but I hope that by now it's a reasonable option.
So far we've just been using Emacspeak & Linux for me to teach him basic programming in Python; it's been great so far. We are just embarking on Emacs Lisp, partly for him to learn a different language, and partly so we can customize emacspeak should we find the need. FYI, he also uses JAWS on Win98 for web browsing and email (too many $$ to upgrade to the version of JAWS that runs on XP, unfortunately). He dual boots himself; adding a bunch of ctrl-Gs to the LILO prompt string lets him know when he reaches the prompt, and then he just hits 'w' for Windows or 'l' for Linux. I put some ctrl-G's in /etc/issue also so he knows when he gets to the login prompt. His .bash_profile sets some environment variables and runs emacspeak directly. He uses dired to manage his files.
Someday I'd like to get him up with reading email and surfing the web in emacspeak too, but I haven't had the time to get that set up.
BTW, related to another topic, I just heard about Rockbox a few days ago from another Slashdot post, so I bought an Archos Ondio off of Ebay with hopes of getting that working for a Christmas present. (Glad he doesn't read Slashdot himself, though it's probably just a matter of time!)
As the father of an 11-yr-old totally blind boy, I'd strongly recommend having your co-worker get in touch with the NFB. There's probably a local chapter nearby. Getting to know competent blind adults should be a very enlightening and positive experience for your co-worker and his son.
I'd also say that although learning braille may seem like it is emphasizing the fact that his son is different, in the long run being unable to access information will make him more different. Being able to read what others are reading, even if it's in a different medium, may make him fit in more, not less.
FWIW, my son loves his braille edition of PC World. So now he's different because he's a geek...
I'll second this. I am teaching programming to my 11-yr-old totally blind son using Python and Emacspeak on Linux. "Voice-lock" mode is particularly cool: it's like font-lock, but uses different voices rather than different fonts to distinguish keywords/strings/comments etc.
Right, this fits into my theory: as long as Intel is obligated by HP to make Itaniums, it can't kill them outright, but it will do what it can to save money in the interim. Even if that goes against the long-term viability of the architecture.
Intel would only do this if saving money was more important to them than giving people a credible reason to buy Itaniums instead of Xeons... and I do believe that's the situation at this point in time. Not a good sign for the future of Itanium.
The reason the i860 might be remembered as a graphics processor is that it did have special graphics instructions (think MMX) and very high peak floating-point performance, and after it became clear that it would never cut it as a true general-purpose CPU it lived on for a while as a dedicated CPU on high-end graphics boards.
There's a pretty good overview here.
For those who think the i860 looks pretty nice, consider this: it had a "push" pipeline for floating point, which meant that the pipeline would only advance when new instructions were issued. So if it took four stages to do an operation, the result of the first operation came out when you issued the fourth one. Just to keep things exciting, I believe the result of the first instruction was stored in the register specified by the fourth instruction (so the pipeline didn't have to bother remembering the destination register specified by the first instruction).
I don't know for a fact, but my guess is that it's Prescott, OR... not too far from Hillsboro where the Prescott CPU was designed.
While it's true that the Cray-1 supported 64-bit integers in 1976, it only had 24-bit addresses (though they were addresses of 64-bit words, so it supported 2^27 bytes of memory). The IBM 360 started out with 24-bit byte addresses, later extended to 31 bits, and extended again to 64 bits only in 2000 (with the ""z/Architecture").
The bottom line is that 4GB of memory was not affordable by anyone until fairly recently in computing history (particularly on Crays, where main memory was typically SRAM!), so 32-bit addresses were not a huge limitation.
I don't know of any machine that had 64-bit addresses before the R4000. Specifically, an architecture where a single user process could address a flat 2^64-byte memory space (or even more than 2^32 bytes). I'm sure if I'm wrong someone will point that out!
I agree with your general lament, but I think you're misapplying it here.
I recall when the R4000 came out that most people thought 64-bit computing was premature (DRAM was quite a bit more expensive then per bit!), but the MIPS philosophy was that people would need time to develop software, and would need hardware to develop software on... so by adding 64 bits "early" then maybe the software would be there by the time people really needed it.
For the easily offended: there was a "sarcasm off" tag between sentences 2 and 3 that got eaten by the html parser. Should have previewed!
You'd think those blind people would get it through their heads that there are some things they just don't need to do on their own. It's like the blacks in the 60s... they should have just been content with the way things were, and not gone whining to the government to get that overbearing Civil Rights Act passed. One of the guiding philosophies of the Nat'l Federation of the Blind (nfb.org) is that blindness itself is merely a nuisance; it's the attitudes of sighted people that make it difficult to be blind. I'm sorry to say that most of the posts on this thread strongly reinforce that position.