Hard Real Time means you presumably give a damned about your system working consistently 100.00000000000000% of the time. It must NEVER break in any way.
Frankly, even if your timing routines are able to provide this real time assurances, I dont think the linux kernel itself is designed with the sort of bulletproof internals Hard Real Time implies. Linux is stable, yes, but it is not thoroughly audited enough to form the basis of a Hard Real Time system. Kernel panics DO happen, and then your hard real time timing is for naught.
pay half as much: fujitsu p1120
on
OQO For Sale
·
· Score: 4, Informative
pay half as much and get the smallest USABLE laptop. The fujitsu p1120. 800mhz crusoe, 4.5h battery life (with extended battery) and touchscreen. keyboard i can code on all day. its a beautiful beautiful system. shell out the cash for the extra ram. and laptops inc/portable one should gladly upgrade your hard drive, if you so request, usually at very reasonable prices.
Fujitsu p1120. Purchase here and forums for it here.
I'm sure we'll see more UDP bound SOAP standards being implemented. UPnP started it a long time ago, in a propriertary sort of way, meant for small devices. WS-* doesnt descriminate over transport: thats good.
You got one thing dead right though: RAM is the fundamental limit to tcp/ip. It is really the only excuse for not being able to do tcp/ip. Interesting that microcontrollers havent advanced ram more quickly. I'm still hacking on a 136 byte controller, for example.
Still seems like if you're going to bother to make a managed device, making the jump to some WS-* shouldnt be oh so difficult.
to be sure, matching your line becomes very exacting. but manufacturers usually have plenty of reference materials on how to lay good traces, what specs your shooting for.
getting paths down doesnt seem THAT hard. i want to work out each individual piece, lay my traces between two points and be done. everything short of 'my software will lay my traces for me'. thats not THAT sinfully bad. getting to spec doesnt look oh so difficult.
crosstalk, on the other hand, scares the crap out of me. you've got a board layed out that you KNOW has good impedences, know to be layed out right, know you have equal trace lenghts and its still not working. what then?
i know it'll end up an excuse to just use MORE pins, but right now halving the pin count seems really really nice. take this inquirer article on fb-dimm. article. look at the difference! pci suffers a similar bus-going-everywhere effect which pcie reduces marginally.
with good solid connectors, good helpful specs, doing this stuff shouldnt be that hard. yes, its higher speed, but an idiot like me migh have a chance of making an uber-high-speed board if i use some liberal spacing.
one of my friends graduated MIT when i was still a sophmore in HS twiddling with microcontrollers. i asked him what he could do with his CompE degree; he told me could build a motherboard. i was amazed, but then again i already knew i would be a CompE anyways. anecdote over.
with more and more cores becoming more and more embedded, sometimes i think there's a small possibility users might be able to again develop their own systems. alright, so i'll probably never build a motherboard for an intel desktop processor, but perhaps some hyper-evolved ARM or powerpc-derivative isnt entirely out of the question.
one of the classic challenges of motherboard design has always been where do you put all these signal paths. we have two solutions in the pipe aimed squarely at this: PCIe and FB-DIMM, both of which drop channel width dramatically.
i'm just a computer engineer (may '05 baby), they dont teach us much about the black art of high speed signalling, but somehow BGA gives me faith. it may mean you have to have your 6 layer board shipped to you, but that seems like an OK tradeoff for being able to design some amazingly high speed hardware without being a signal-foo master.
of course & as usual, the reality is probably highly departed from my little dream world. but its a nice thought, and i still think it might be possible. at least if i can afford some good PCB design software. that, i believe, will be the crux of the issue.
Do NOT, i repeat, DO NOT look to journalists. what you want now is an economists.
the media has been relegated to a game of he said she said; back and forth. the media never bothers to find out of a candidate is actually lying. they simply report what each candidate says.
the job of a journalist is to make stories. the job of a ECONOMIST is to understand fiscal policies, markets, and (growingly important:) social costs.
if you want to understand your government, find an economist who can tack points to the bottom line ($) and reduce it to something you can understand.
Nick Confessore has a nice article talking about hte rise of economist as a source of information, and in particular the Paul Krugman phenomena. I highly recommend these read, parituclarly in conjunction with some Paul Krugman himself for reference.
you can either figure out what a candidate is REALLY supporting, or you can be just another single issue voter.
the greatest crime is that we have done NOTHING to enable network syncronization of audio.
cpu processing, special effects, fancy mixers. all these features are simply pieces of silicon ripe for the molding. but they are nothing novel. they are not signifcant advancements; they are merely luxuries most consumers are too ignorant to demand. they are not advancements!
what we need is a sound system which is actually more functional! one which is ADVANCED, bears features as of yet unseen:
a computer should be able to declare, "play this audio file at 09:00 exactly", and every computer should be able to start playing in perfect tune. AND, even for a 100 hour sample, it should stay in relative harmony. the ONE time i got two computers to start playing a file in harmony, the two systems drifted apart so that by the end of the 4 minute tune they were quite painfully out of phase.
yes, this is a problem computer networks were never meant to tackle. still, i believe there could be potential solutions short of atomic clocks in every computer.
Hrm, I seem to have worked myself into a linguistic conundrum here.
Lets start with the PPC. (Knowing very little about the PPC a week ago,) I was reading Ars history of PPC and i guess i took them too literally: "...and Apple needed a CPU for its personal computers that would be both cutting-edge and backwards compatible with the 68K", before it goes into talking about the Power 601 and its "bridge support" for 68K. It seems there was initially some degree of hardware compatibility with what must have been hardware books to emulate unsupported instructions. 68k emulation was actually sped up in latter revisions, before ultimately recieving the axe in more modern renditions when pure software emulation became feasible. As for the 601 (from the IBM tech brief), "The PowerPC 601 processor provides a bridge between the POWER and PowerPC Architectures by supporting most of the PowerPC and POWER instructions. The PowerPC 601 executes all compiler-generated user-level POWER instructions. The implementation also supports all but a few of the 32-bit PowerPC instructions". PPC was some new (Power) and some old (68k).
As for the ColdFire, it is indeed a direct decendant of the 68k, however it was "optimized" for a more "RISC" nature, as ambiguous as the term is. A number of memory operations were removed, a number of address modes removed. while it is true that yes, x86 has some address mode deficiencies, the more is better philosophy isnt all their either. a lot of the more "expressive" addressing modes for assembly coders were cut.
I suppose architecturally there's no comparison; the PPC has all the ammenities of a modern superwide superscalar, whereas the ColdFire just gained an MMU. but hey, its got an MMU. now it can run a _real_ linux.
Motorola spunoff (all|a large section) of their IC department and thus was born Freescale. They've been making CPU's for apple for decades.
IIRC, PowerPC was engineered to be backwards compatible with 68k. To preserve apple's software. The main dis-advantage of this is that you'd have to support the umpteen billion addressing modes.
There is a RISC'ified alternate side though: The ColdFire processors. They've been a uClinux target for a while.
However, whats truly notable is that the new MFC54xx series has a mmu. No need for uClinux, it runs real linux. Quite well i'd imaging: 133mhz DDR ram, 433 mhz, pci-interface, dual ethernet (100 mbit), usb and onboard crypto accelerator. All with a low advertised power consumption.
Still awaiting the Base Support Package. C'mon Metroworks.
the fujitsu p series is amazing. there is a forum for them here.
most have 10 - 11 hours with extended battery and modular battery. they're "ultra-portables": 10.6 inch screens with 1280x768 resolution. they're available in oldschool crusoe models for "cheap" or centrino platforms.
i have a p1120, which is only 5 hours with an extended battery, but no modular bay. its got a 8.9 inch 1024x600 screen which is touchscreen - the reason i bought it. it is my baby.
Slashdot Mirror
My one criticism of Stewart is that we have to endure at least ten minutes of ads during the Daily Show, leaving a scant ~19 - 20 minutes of show.
Then again, what else does comedy central have to make money? Puppets making prank calls?
The other thing everyone seems to be missing:
Hard Real Time means you presumably give a damned about your system working consistently 100.00000000000000% of the time. It must NEVER break in any way.
Frankly, even if your timing routines are able to provide this real time assurances, I dont think the linux kernel itself is designed with the sort of bulletproof internals Hard Real Time implies. Linux is stable, yes, but it is not thoroughly audited enough to form the basis of a Hard Real Time system. Kernel panics DO happen, and then your hard real time timing is for naught.
remember kids,
fair and balanced!
pay half as much and get the smallest USABLE laptop. The fujitsu p1120. 800mhz crusoe, 4.5h battery life (with extended battery) and touchscreen. keyboard i can code on all day. its a beautiful beautiful system. shell out the cash for the extra ram. and laptops inc/portable one should gladly upgrade your hard drive, if you so request, usually at very reasonable prices.
Fujitsu p1120. Purchase here and forums for it here.
I'm sure we'll see more UDP bound SOAP standards being implemented. UPnP started it a long time ago, in a propriertary sort of way, meant for small devices. WS-* doesnt descriminate over transport: thats good.
You got one thing dead right though: RAM is the fundamental limit to tcp/ip. It is really the only excuse for not being able to do tcp/ip. Interesting that microcontrollers havent advanced ram more quickly. I'm still hacking on a 136 byte controller, for example.
Still seems like if you're going to bother to make a managed device, making the jump to some WS-* shouldnt be oh so difficult.
Myren
Wake me when we have effective position location sensors, indoors and out, and the required beacon deployment to be useful most places.
Until then its all BS.
Data is useless without context. Position is the best context we have any hope of auto-generating.
whats the deal with them having an adjustable backlight on a transflective display?
i thought most transflective displays had no backlight.
control freaks
looks like microsoft finally found an incentive strong enough to keep people upgrading.
for people stupid enough to NOT upgrade to firefox.
i dont mind a stupidity tax. hell, even if it is going to microsoft's pockets.
sure, but just what part do you think most ludicrous?
how bout the fact that MythTV was built to work not on a computer but a network of computers.
mythtv, like everything else in linux, is all about the options
to be sure, matching your line becomes very exacting. but manufacturers usually have plenty of reference materials on how to lay good traces, what specs your shooting for.
getting paths down doesnt seem THAT hard. i want to work out each individual piece, lay my traces between two points and be done. everything short of 'my software will lay my traces for me'. thats not THAT sinfully bad. getting to spec doesnt look oh so difficult.
crosstalk, on the other hand, scares the crap out of me. you've got a board layed out that you KNOW has good impedences, know to be layed out right, know you have equal trace lenghts and its still not working. what then?
i know it'll end up an excuse to just use MORE pins, but right now halving the pin count seems really really nice. take this inquirer article on fb-dimm. article. look at the difference! pci suffers a similar bus-going-everywhere effect which pcie reduces marginally.
with good solid connectors, good helpful specs, doing this stuff shouldnt be that hard. yes, its higher speed, but an idiot like me migh have a chance of making an uber-high-speed board if i use some liberal spacing.
i could be dead wrong.
cant you get 6 layer boards for, what, like $60 a pop? if not, i know what i'm doing after i'm done with school; undercuting greedy ass punks.
sure you cant play with it like you can a breadboard. the tradeoff is you can actually get speed, something former packaging just didnt give you.
you mean like the VRM's on the PPro & Xeon's?
one of my friends graduated MIT when i was still a sophmore in HS twiddling with microcontrollers. i asked him what he could do with his CompE degree; he told me could build a motherboard. i was amazed, but then again i already knew i would be a CompE anyways. anecdote over.
with more and more cores becoming more and more embedded, sometimes i think there's a small possibility users might be able to again develop their own systems. alright, so i'll probably never build a motherboard for an intel desktop processor, but perhaps some hyper-evolved ARM or powerpc-derivative isnt entirely out of the question.
one of the classic challenges of motherboard design has always been where do you put all these signal paths. we have two solutions in the pipe aimed squarely at this: PCIe and FB-DIMM, both of which drop channel width dramatically.
i'm just a computer engineer (may '05 baby), they dont teach us much about the black art of high speed signalling, but somehow BGA gives me faith. it may mean you have to have your 6 layer board shipped to you, but that seems like an OK tradeoff for being able to design some amazingly high speed hardware without being a signal-foo master.
of course & as usual, the reality is probably highly departed from my little dream world. but its a nice thought, and i still think it might be possible. at least if i can afford some good PCB design software. that, i believe, will be the crux of the issue.
myren
there has to be some ramifications to sapping the power out of the wind.
i thought there were 540p cameras already available for this market segment.
540p is higher bandwidth.
Do NOT, i repeat, DO NOT look to journalists. what you want now is an economists.
the media has been relegated to a game of he said she said; back and forth. the media never bothers to find out of a candidate is actually lying. they simply report what each candidate says.
the job of a journalist is to make stories.
the job of a ECONOMIST is to understand fiscal policies, markets, and (growingly important:) social costs.
if you want to understand your government, find an economist who can tack points to the bottom line ($) and reduce it to something you can understand.
Nick Confessore has a nice article talking about hte rise of economist as a source of information, and in particular the Paul Krugman phenomena. I highly recommend these read, parituclarly in conjunction with some Paul Krugman himself for reference.
you can either figure out what a candidate is REALLY supporting, or you can be just another single issue voter.
the greatest crime is that we have done NOTHING to enable network syncronization of audio.
cpu processing, special effects, fancy mixers. all these features are simply pieces of silicon ripe for the molding. but they are nothing novel. they are not signifcant advancements; they are merely luxuries most consumers are too ignorant to demand. they are not advancements!
what we need is a sound system which is actually more functional! one which is ADVANCED, bears features as of yet unseen:
a computer should be able to declare, "play this audio file at 09:00 exactly", and every computer should be able to start playing in perfect tune. AND, even for a 100 hour sample, it should stay in relative harmony. the ONE time i got two computers to start playing a file in harmony, the two systems drifted apart so that by the end of the 4 minute tune they were quite painfully out of phase.
yes, this is a problem computer networks were never meant to tackle. still, i believe there could be potential solutions short of atomic clocks in every computer.
i got bored this summer and did it with VideoLan software.
its a GREAT generic solution for all network AV systems. Very advanced.
I'd like to do some MythTV integration, but i'm rediculously busy with other projects right now.
http://www.microsoft.com/hardware/mouseandkeyboard /images/starck_products_whois.jpg
Who is Phillip Starck, Phillip Starck, Phillip Starck? Did I love the, did I love the, did I love thee, Phillip Starck Phillip Stack Phillip Starck? Starck Phillip Glass Love-thee? Love-thee Glass Phillip Starck?
Phillip Glass
All in the Timing Play, fifth down.
It was just the first thing I saw when I got to the page.
Hrm, I seem to have worked myself into a linguistic conundrum here.
Lets start with the PPC. (Knowing very little about the PPC a week ago,) I was reading Ars history of PPC and i guess i took them too literally: "...and Apple needed a CPU for its personal computers that would be both cutting-edge and backwards compatible with the 68K", before it goes into talking about the Power 601 and its "bridge support" for 68K. It seems there was initially some degree of hardware compatibility with what must have been hardware books to emulate unsupported instructions. 68k emulation was actually sped up in latter revisions, before ultimately recieving the axe in more modern renditions when pure software emulation became feasible. As for the 601 (from the IBM tech brief), "The PowerPC 601 processor provides a bridge between the POWER and PowerPC Architectures by supporting most of the PowerPC and POWER instructions. The PowerPC 601 executes all compiler-generated user-level POWER instructions. The implementation also supports all but a few of the 32-bit PowerPC instructions". PPC was some new (Power) and some old (68k).
As for the ColdFire, it is indeed a direct decendant of the 68k, however it was "optimized" for a more "RISC" nature, as ambiguous as the term is. A number of memory operations were removed, a number of address modes removed. while it is true that yes, x86 has some address mode deficiencies, the more is better philosophy isnt all their either. a lot of the more "expressive" addressing modes for assembly coders were cut.
I suppose architecturally there's no comparison; the PPC has all the ammenities of a modern superwide superscalar, whereas the ColdFire just gained an MMU. but hey, its got an MMU. now it can run a _real_ linux.
Myren
Motorola spunoff (all|a large section) of their IC department and thus was born Freescale. They've been making CPU's for apple for decades.
IIRC, PowerPC was engineered to be backwards compatible with 68k. To preserve apple's software. The main dis-advantage of this is that you'd have to support the umpteen billion addressing modes.
There is a RISC'ified alternate side though: The ColdFire processors. They've been a uClinux target for a while.
However, whats truly notable is that the new MFC54xx series has a mmu. No need for uClinux, it runs real linux. Quite well i'd imaging: 133mhz DDR ram, 433 mhz, pci-interface, dual ethernet (100 mbit), usb and onboard crypto accelerator. All with a low advertised power consumption.
Still awaiting the Base Support Package. C'mon Metroworks.
Myren
the touch screen is just another mouse device.
the one down side is its calibration is a little off if you run in non-native res. makes playing moo2 difficult.
the fujitsu p series is amazing. there is a forum for them here.
most have 10 - 11 hours with extended battery and modular battery. they're "ultra-portables": 10.6 inch screens with 1280x768 resolution. they're available in oldschool crusoe models for "cheap" or centrino platforms.
i have a p1120, which is only 5 hours with an extended battery, but no modular bay. its got a 8.9 inch 1024x600 screen which is touchscreen - the reason i bought it. it is my baby.