fwiw, the 'blue pill' is the next big thing and intel lost out, entirely. you can pay well over $100 or you can pay $2. I know what I would do;)
I bought 20 of the bluepill boards a few months ago when they were mentioned on Hackaday. I plan to use them for small USB HID device projects, and I already have one working with mbed code (the CPU is equivalent to one on an ST Nucleo board) and an ST-Link v2. I've been working with STM32 since 2010, so it's like bread and butter to me, especially the F103.
I read that this morning, and I'm going to have to agree.
- Documentation was too hard to get, even for people who knew Intel engineers. (apparently the specific example was trying to use DMA with SPI) China gets away with a lot of poor documentation because their stuff is so cheap.
- That damn connector may be amazingly compact, but that it also made it hard to work with. It had a limited selection of base boards unless you had a PCB engineer who could design a custom one, so you usually end up with more than you needed. And it wasn't designed for multiple connects/disconnects like you would want in a maker product. The usefulness of 0.1" pin headers can not be underestimated.
- Its main purpose seemed to be a bull-headed "x86 in everything because x86!" attitude, even in applications with no inherent need for an x86 architecture. Their domination of the consumer market made them over-confident.
- Making boards with all the bells and whistles, and then some, drove the price of entry way up. This cuts out the low-end community support beyond students who get free or heavily-discounted units. And in fact, that was exactly who I saw with Edison-based projects at maker faires. Before Arduino, it was common to cram demo boards with various other chips, but that brought the board cost up. The emergence of a standard "shield" pin-out let those other chips move to daughterboards. Intel was clearly still stuck in the old metaphor. Most micro-controller boards now are in the $10-$20 range, more if it runs Linux, less if you get generic Chinese stuff.
It's like a perfect study in how not to get the maker community interested in your product.
I've learned the hard way that you're not buying a product, you're buying a PLATFORM.
This is one of the reasons that I've stuck with the mbed platform. From the time five-ish years ago that an NXP rep left behind an NXP-1768 MBED at my work (fuck the 1st gen LPCXpressos that he left too, their debug interfaces sucked, and not worth working around), I found a good paradigm of using C++ that I was able to apply to my own embedded coding at work. Best of all, it was system-agnostic (programmed via copying a binary to a USB filesystem), which meant it didn't require Windows, like so many micro-controller development systems did back then, and many still do, so I was able to use it with OS X at home.
Now I have a bunch of different boards that support it, with many different pinouts (some Arduino-compatible), including so-called "bluepill" boards that cost as little as $2 each from China, all running ARM Cortex M, not AVR. I ended up with a couple of boards with AVR/Arduino hardware, but I don't even have the Arduino IDE installed.
But I have to say that though you may be having fun with it, D sounds like a bit of a dead end. It's very hard for a new programming language to gain traction, if only because half its users may jump to the next shiny new language in a couple of years. It's even worse for micro-controllers, where C/C++ is king, with only a few crazy MicroPython users and PIC assembly die-hards skittering around the rafters .
I forgot to mention one other thing. I have been watching (or trying to) a Sunday night show on CBS, that being Elementary. They have an infuriating habit of letting professional sports run overtime (particularly handegg, which runs over the time slot about 90% of the time, and yes I know about the Heidi Incident, this is with normal games so they need to start making realistic time slots), and delaying the rest of the line-up. This means that if you wanted to record the shows on a DVR, then FUCK YOU, because now the published schedule is fucked up, sometimes by a full hour. Even worse, I missed the season finale because they changed the show name in the schedule to "Elementary (Season finale)", and my DVR failed to match it.
Fox finally got a clue last season and started putting a sacrificial rerun or two at the start of prime time so they could go JIT on it and leave the rest of the shows (and the local news) at their correct time.
I barely got to watch it, because it was in the dying days of UPN, and where I lived at the time didn't have a UPN station. Cable TV actually brought in a station from the next market down, but I was strictly antenna and had to download it from Usenet. (this was before Bit Torrent) I watched it for most of two seasons, until my preferred uploader's posts started getting a lot of dropped messages (this was a year or so before par2), then I had to stop. I've heard that seasons 3 and 4 were much better, but I've never seen them, and they haven't appeared on the independent sub-channel networks in my market.
There's this amazing new invention called an "antenna" by which shows are streamed directly to your house, only everyone gets the same stream at the same time with no rewinds, and you get to watch it for free. There are even ads in it, too, just like most streaming!
But Sunday night live-watching for me is owned by Fox (at least if they don't keep wedging live-action stinkers in), so a computer will have to download the antenna stream for me (again, just like on the internets!), and let me watch it later. Not that I'm expecting much more than concentrated SJW builshit from it from what I've heard (Star Trek died for me anyhow even before they blew everything up a couple of movies ago and became millennial trash), so I suspect I won't even be DVR-ing it for more than two or three weeks.
Austin has had all the weird watered out of it the past 20 years or so since it became a center of high-tech. Now most of the weird has been replaced with euroblocks all up and down Lamar and Burnet and other main drags turning them into canyons. I just sold my house (in the far NW) for twice what I paid for it, and moved back to San Antonio.
And a few years from now, when Apple swings its sine curve back toward repairability, Microsoft's cosine curve will still be deep into fuck-you territory.
There's an electronics store five miles away from me (not Fry's) where I can get one, they sell iFixit tool sets. (Except I already have one of the medium-size iFixit tool sets and probably would already have the needed bit.)
Back in the day, I took a regular Torx driver and ground down the plastic handle just enough for it to fit. That had nothing on plastic-welding a keyboard cover to close up a computer.
Don't joke about it. I was reading today about a microwave oven that had a switch, which if put in an "unnatural" position, would short across the mains power. Yes, WTF. I guess they really didn't want teenagers to be able to defeat the door interlock and run it with the door open.
Indeed. OS X years ago changed to using a double-precision float, which gives sub-microsecond precision for the prime epoch, and can even deal with Y10K and beyond.
As long as user software declares time values as time_t, Y2038 will work correctly if the OS can deal with it.
I have AT&T as an ISP, and a few years ago they farmed out their mail to Yahoo. I run my own inbound, but for at least a decade, I have dealt with the outbound port 25 block (even before it got blocked) by forwarding outbound mail through my ISP. When it was just AT&T, they could identify "friendly" mail by being in one of their IP blocks. Now you have to use your account's e-mail/password combo.
The problem is that by default it only accepts mail with a From: address equal to that specific e-mail address. Any others you want (I typically use one user name with two different domains, one or two other household members), have to be specifically added with the Yahoo Mail user interface.
I guess I may just end up having to ask for outbound 25 to be opened up, and re-configure my server to send outbound mail directly. The negative side is that will cause my mail to come from a customer address (anti-spammers are likely to flag the IPs as "dynamic", even though it's in a static block), so I may have to get a hosted server for an outbound relay.
Or I could even get lucky, and AT&T might realize the situation they are now in (e-mail service outsourced to one of their telco competitors) and in-source it again.
You can use Flex Tape! Have you seen that TV commercial? A guy sliced a fucking boat in half with a power saw, and then taped it back together with that shit!
I once had to re-spackle the heat-sink compound in a dual-CPU "Windtunnel" G4. It took me fifteen minutes from start to finish, and that's while it was still on the floor.
It was also the time of the last good mac mini. 2011-2012 had the best models of Mac Pro, MacBook Pro, and mac mini. They are even still relevant, since Intel's CPUs haven't gotten significantly better in that time.
In my opinion, it was not a coincidence that this started after Steve Jobs died. He was no longer there to say "fuck you" to people to prevent this kind of crap from happening.
It's also not the first time that Apple made crap computers. In the '90s, before Steve came back, Apple made some really bad hardware. In particular, the 8100 series, where you had to remove the logic board to upgrade its memory, but they also made a lot of under-powered, under-RAMmed crap, while at the same time half of the OS was still emulated on PowerPC. The fruity iMacs happened just as he got there, but even they had design problems, as the original model didn't have any external connector faster than 10-12 Mbps. (10BaseT and USB).
I think the people now at Apple have forgotten that part of the value that justified their higher price is specifically the ability to upgrade it two or three years later to keep it current.
It's probably using FTDI FIFO-bridge chips in 'bitbang' mode to emulate a legacy parallel port. The (e)eprom programmer itself is probably a parallel-port design hardwired to a USB FIFO bridge(*).
Ugh. But that is entirely likely, since this was (IIRC) their first or second USB programmer. The original parallel port design wouldn't have been bit-banging, except as a communication protocol to the CPU inside the programmer (this wasn't some Willem-tier crap), but it makes sense that they would have taken the easy way out, since USB was still rather new at the time.
I have NetBSD on an old iMac which works OK if you don't need graphics acceleration.:(
Apple made a lot of computers without discrete GPUs once Intel started including half-decent ones, so that may be less of a problem than you think. They didn't even make a mac mini with a discrete GPU after i5/i7, and they made a few MacBook Pro models sans GPU. (And the ones they made with GPUs had problems. NVidia made some crappy chips in 2010-2012, you basically have to replace it with a newer chip by hot-air rework)
Have you got a specific example? They haven't sold a computer without x64 support since 2009. And El Capitan still runs on Core 2 Duo, so that's at least seven fucking years. What's the oldest computer you still use?
And unlike Windows, most Mac software doesn't need the latest OS. I am currently running 10.9 on my "daily driver" and have had no need to upgrade any farther. I have other computers with 10.10 and 10.11, but that's the way I got them. In fact, I have more need to keep a computer or two running 10.6 than I do anything else.
Slashdot Mirror
The raisins are even more fun.
From what I've seen, it correlates with both parents being over a certain age. So single GenX guys, start looking for that perfect 27-year-old!
fwiw, the 'blue pill' is the next big thing and intel lost out, entirely. you can pay well over $100 or you can pay $2. I know what I would do ;)
I bought 20 of the bluepill boards a few months ago when they were mentioned on Hackaday. I plan to use them for small USB HID device projects, and I already have one working with mbed code (the CPU is equivalent to one on an ST Nucleo board) and an ST-Link v2. I've been working with STM32 since 2010, so it's like bread and butter to me, especially the F103.
I read that this morning, and I'm going to have to agree.
- Documentation was too hard to get, even for people who knew Intel engineers. (apparently the specific example was trying to use DMA with SPI) China gets away with a lot of poor documentation because their stuff is so cheap.
- That damn connector may be amazingly compact, but that it also made it hard to work with. It had a limited selection of base boards unless you had a PCB engineer who could design a custom one, so you usually end up with more than you needed. And it wasn't designed for multiple connects/disconnects like you would want in a maker product. The usefulness of 0.1" pin headers can not be underestimated.
- Its main purpose seemed to be a bull-headed "x86 in everything because x86!" attitude, even in applications with no inherent need for an x86 architecture. Their domination of the consumer market made them over-confident.
- Making boards with all the bells and whistles, and then some, drove the price of entry way up. This cuts out the low-end community support beyond students who get free or heavily-discounted units. And in fact, that was exactly who I saw with Edison-based projects at maker faires. Before Arduino, it was common to cram demo boards with various other chips, but that brought the board cost up. The emergence of a standard "shield" pin-out let those other chips move to daughterboards. Intel was clearly still stuck in the old metaphor. Most micro-controller boards now are in the $10-$20 range, more if it runs Linux, less if you get generic Chinese stuff.
It's like a perfect study in how not to get the maker community interested in your product.
I've learned the hard way that you're not buying a product, you're buying a PLATFORM.
This is one of the reasons that I've stuck with the mbed platform. From the time five-ish years ago that an NXP rep left behind an NXP-1768 MBED at my work (fuck the 1st gen LPCXpressos that he left too, their debug interfaces sucked, and not worth working around), I found a good paradigm of using C++ that I was able to apply to my own embedded coding at work. Best of all, it was system-agnostic (programmed via copying a binary to a USB filesystem), which meant it didn't require Windows, like so many micro-controller development systems did back then, and many still do, so I was able to use it with OS X at home.
Now I have a bunch of different boards that support it, with many different pinouts (some Arduino-compatible), including so-called "bluepill" boards that cost as little as $2 each from China, all running ARM Cortex M, not AVR. I ended up with a couple of boards with AVR/Arduino hardware, but I don't even have the Arduino IDE installed.
But I have to say that though you may be having fun with it, D sounds like a bit of a dead end. It's very hard for a new programming language to gain traction, if only because half its users may jump to the next shiny new language in a couple of years. It's even worse for micro-controllers, where C/C++ is king, with only a few crazy MicroPython users and PIC assembly die-hards skittering around the rafters .
I forgot to mention one other thing. I have been watching (or trying to) a Sunday night show on CBS, that being Elementary. They have an infuriating habit of letting professional sports run overtime (particularly handegg, which runs over the time slot about 90% of the time, and yes I know about the Heidi Incident, this is with normal games so they need to start making realistic time slots), and delaying the rest of the line-up. This means that if you wanted to record the shows on a DVR, then FUCK YOU, because now the published schedule is fucked up, sometimes by a full hour. Even worse, I missed the season finale because they changed the show name in the schedule to "Elementary (Season finale)", and my DVR failed to match it.
Fox finally got a clue last season and started putting a sacrificial rerun or two at the start of prime time so they could go JIT on it and leave the rest of the shows (and the local news) at their correct time.
I barely got to watch it, because it was in the dying days of UPN, and where I lived at the time didn't have a UPN station. Cable TV actually brought in a station from the next market down, but I was strictly antenna and had to download it from Usenet. (this was before Bit Torrent) I watched it for most of two seasons, until my preferred uploader's posts started getting a lot of dropped messages (this was a year or so before par2), then I had to stop. I've heard that seasons 3 and 4 were much better, but I've never seen them, and they haven't appeared on the independent sub-channel networks in my market.
There's this amazing new invention called an "antenna" by which shows are streamed directly to your house, only everyone gets the same stream at the same time with no rewinds, and you get to watch it for free. There are even ads in it, too, just like most streaming!
But Sunday night live-watching for me is owned by Fox (at least if they don't keep wedging live-action stinkers in), so a computer will have to download the antenna stream for me (again, just like on the internets!), and let me watch it later. Not that I'm expecting much more than concentrated SJW builshit from it from what I've heard (Star Trek died for me anyhow even before they blew everything up a couple of movies ago and became millennial trash), so I suspect I won't even be DVR-ing it for more than two or three weeks.
Austin has had all the weird watered out of it the past 20 years or so since it became a center of high-tech. Now most of the weird has been replaced with euroblocks all up and down Lamar and Burnet and other main drags turning them into canyons. I just sold my house (in the far NW) for twice what I paid for it, and moved back to San Antonio.
And a few years from now, when Apple swings its sine curve back toward repairability, Microsoft's cosine curve will still be deep into fuck-you territory.
There's an electronics store five miles away from me (not Fry's) where I can get one, they sell iFixit tool sets. (Except I already have one of the medium-size iFixit tool sets and probably would already have the needed bit.)
Back in the day, I took a regular Torx driver and ground down the plastic handle just enough for it to fit. That had nothing on plastic-welding a keyboard cover to close up a computer.
Don't joke about it. I was reading today about a microwave oven that had a switch, which if put in an "unnatural" position, would short across the mains power. Yes, WTF. I guess they really didn't want teenagers to be able to defeat the door interlock and run it with the door open.
Indeed. OS X years ago changed to using a double-precision float, which gives sub-microsecond precision for the prime epoch, and can even deal with Y10K and beyond.
As long as user software declares time values as time_t, Y2038 will work correctly if the OS can deal with it.
I have AT&T as an ISP, and a few years ago they farmed out their mail to Yahoo. I run my own inbound, but for at least a decade, I have dealt with the outbound port 25 block (even before it got blocked) by forwarding outbound mail through my ISP. When it was just AT&T, they could identify "friendly" mail by being in one of their IP blocks. Now you have to use your account's e-mail/password combo.
The problem is that by default it only accepts mail with a From: address equal to that specific e-mail address. Any others you want (I typically use one user name with two different domains, one or two other household members), have to be specifically added with the Yahoo Mail user interface.
I guess I may just end up having to ask for outbound 25 to be opened up, and re-configure my server to send outbound mail directly. The negative side is that will cause my mail to come from a customer address (anti-spammers are likely to flag the IPs as "dynamic", even though it's in a static block), so I may have to get a hosted server for an outbound relay.
Or I could even get lucky, and AT&T might realize the situation they are now in (e-mail service outsourced to one of their telco competitors) and in-source it again.
You can use Flex Tape! Have you seen that TV commercial? A guy sliced a fucking boat in half with a power saw, and then taped it back together with that shit!
I once had to re-spackle the heat-sink compound in a dual-CPU "Windtunnel" G4. It took me fifteen minutes from start to finish, and that's while it was still on the floor.
It was also the time of the last good mac mini. 2011-2012 had the best models of Mac Pro, MacBook Pro, and mac mini. They are even still relevant, since Intel's CPUs haven't gotten significantly better in that time.
In my opinion, it was not a coincidence that this started after Steve Jobs died. He was no longer there to say "fuck you" to people to prevent this kind of crap from happening.
It's also not the first time that Apple made crap computers. In the '90s, before Steve came back, Apple made some really bad hardware. In particular, the 8100 series, where you had to remove the logic board to upgrade its memory, but they also made a lot of under-powered, under-RAMmed crap, while at the same time half of the OS was still emulated on PowerPC. The fruity iMacs happened just as he got there, but even they had design problems, as the original model didn't have any external connector faster than 10-12 Mbps. (10BaseT and USB).
I think the people now at Apple have forgotten that part of the value that justified their higher price is specifically the ability to upgrade it two or three years later to keep it current.
Don't use a color printer to leak shit.
I think it was the 15" i5 in 2011 and all 13" since 2011 that have only used integrated Intel graphics.
It's probably using FTDI FIFO-bridge chips in 'bitbang' mode to emulate a legacy parallel port. The (e)eprom programmer itself is probably a parallel-port design hardwired to a USB FIFO bridge(*).
Ugh. But that is entirely likely, since this was (IIRC) their first or second USB programmer. The original parallel port design wouldn't have been bit-banging, except as a communication protocol to the CPU inside the programmer (this wasn't some Willem-tier crap), but it makes sense that they would have taken the easy way out, since USB was still rather new at the time.
Ah, okay, it's been a long time since I had to care that much about the architecture of either.
I have NetBSD on an old iMac which works OK if you don't need graphics acceleration. :(
Apple made a lot of computers without discrete GPUs once Intel started including half-decent ones, so that may be less of a problem than you think. They didn't even make a mac mini with a discrete GPU after i5/i7, and they made a few MacBook Pro models sans GPU. (And the ones they made with GPUs had problems. NVidia made some crappy chips in 2010-2012, you basically have to replace it with a newer chip by hot-air rework)
Have you got a specific example? They haven't sold a computer without x64 support since 2009. And El Capitan still runs on Core 2 Duo, so that's at least seven fucking years. What's the oldest computer you still use?
And unlike Windows, most Mac software doesn't need the latest OS. I am currently running 10.9 on my "daily driver" and have had no need to upgrade any farther. I have other computers with 10.10 and 10.11, but that's the way I got them. In fact, I have more need to keep a computer or two running 10.6 than I do anything else.