I do research on this (among other things) for a living.
With current component prices, even using the cheapest of the cheapest parts, you can't get there if you're building an identifiable PC using sustainable components -- IE something manufactured for the market rather and an opportunistic purchase that you can't count on.
Cheapest x86 CPUs available (not inventory blowouts, but manufacturered at thin margins) -- OEM price of about $20. Cheapest chip sets, $15. Ditto hard disk, $40. Ditto optical drive, $18. You're almost at $100 in cost there, and we've not touched a case or RAM yet, yet alone the motherboard manufacturing costs, support chips, system assembly.
If a $100 PC was do-able, those $200 Walmart PCs would be $100 Walmart PCs.
If you redefine the computer, eliminate any signficant removable or internal storage and use a small amount of flash to boot-load remotely, use a tiny memory footprint, and use a tiny non-x86 integrated CPU/system logic device, like an ARM or something similar... in other words, if you build something that looks a heck of a lot like a PDA without a screen -- then it's marginally do-able. But all you're really doing is moving the high-cost components upstream.
Think back to a sinclair ZX-80 or something similar... that's do-able today, with better performance and way more memory.
A box using new components that runs a fat OS with a disk drive, it's not possible today for $100. A thin linux (or windows CE, for that matter) client booting a remote copy of the OS... on the edge of possible, but it's not a PC.
This is doable, if you're willing to compromise on not being a "complete" graphics solution, IE, not trying to implement hardware accelerated everything.
I'd suggest simply a dumb frame buffer, and doing everything in software. Then your solution is simply a memory controller, plus a CRT controller and an output graphics DAC. Really, if you implemented a VGA controller from the early 90s sans the backware compatible bits for MDA,CGA, EGA, etc, and opened up the line/pixel resolution you'd be there.
If you make the goals modest enough, this could probably be done with a field-programmable gate array, an external graphics DAC, and some RAM. The only tricky part is external DACs are hard to come by these days and aren't cheap. (Prior to their integration in current-era parts, they were down in the $1 range, but since they're integrated now, the only stuff that's commonly available are insanely high end workstation DACs.)
Back in the mid-late 90s, right as 2D acceleration was hitting its peak and 3D acceleration was emerging, there were netlists of VGA designs for sale for as little as $500. So designs of this level aren't hard.
If you're thinking even full 2D acceleration, it gets much harder. and if you're thinking 3D acceleration in an "open" project that would be competitive with even the slowest Nvidia/ATI parts... you're on drugs.
One alternative would be to approach an existing vendor about opening up an "old" product. However, getting a fresh production run of an old product wouldn't be cheap -- you're basically talking a million to get masks made, initial wafer lots, etc. Hence the FPGA suggestion, since that's commonly available hardware that doesn't require any manufacturing specific to the design.
However, some manufacturers may still be in low volume production with a suitable product. Someone who used to be small player in the PC graphics market years ago but isn't now would be a candidate -- perhaps a Silicon Motion or an Avance Logic (part of Realtek now, I think, though I doubt the video products are still active, tho the audio parts are) could be persuaded to open up a part that's in sustained low-volume production.
Seriously, though, if you can't offer significant volumes -- the minimum probably being on the order of 10-20K/quarter, and that's VERY small in this business -- don't expect to get much help from existing vendors.
Where the confusion comes up is in the old days where interlacing was first used -- black and white television. Interlaced TV signals drove black and white CRTs... and by selecting the right phosphor, the displays had persistence, where the image would continue to glow into the next frame even after being drawn. High persistance phosphor DID cut down on flicker, and was necessary because of the interlacing.
If you want to get into really obscure stuff... Radar display tubes (and some oscilloscopes) often had incredible persistence -- often lasting several to even tens of seconds. Surplus radar tubes were popular among ham radio operators doing slow-scan television, as they could maintain the TV display at the insanely low frame rates used at the time (I think it was about 1 frame per 8 seconds or so, memory is hazy here.)
Anyway, think of the displays long ago as a giant analog visual low-pass filter -- you could throw pretty much any signal at them regarless of flicker and get a flicker-free display, so the bandwidth savings from interlacing was an obvious choice. Not so with digital technology today.
Actually, not that powerful, relatively speaking. A typical 121.5 EPIRB puts out 75-125 milliwatts. Keep in mind this is VHF and the satellites are typically NOAA birds in low-earth orbit expecting a line of sight signal, so 100 milliwatts is very workable.
The issue with 121.5 EPIRBs is all they do is send a warble tone -- no ID, location, nothing. All the processing is done by the rest of the infrastructure, and even then the output is basically a position (still no ID) to within a mile or so -- with people using radio-direction finders narrowing it down more.
The newer 406 MHz EPIRBs have specific user data and location information transmitted in their digital packets, so not only do they know the where, but they also know the who -- so when they get a boat beacon originating at someone's house they pretty much already know it's a false alarm.
You don't use HF, don't need it, and don't know any service that does, so screw them.
Pretty shortsighted and impressively incomplete view of this problem. Not all communications are local, which is what every VHF/UHF method you mentioned is. HF is in ONLY radio-based communication system that provides global coverage without needing substantial infrastructure such as satellites, etc.
I realize this is a joke, but driving in the phoenix/scottsdale area today was insane. The roads and sky were effectively removed, I'm afraid.
As I was driving south in the upper northern area of scottsdale (adjacent to Phoenix) this morning, I saw police cars on EVERY possible right-turn option for more than two miles (streets, parking lots, you name it). I noticed many of the cars had trunks partially opened with barricades inside. On my return trip two hours later I made a point of taking a route that would avoid this area -- barely.
Good thing.
They had completely blocked off the road (Scottsale Road) in both directions, immediately before the freeway on-ramp. This road is effectively the only exit draining traffic for an area for about 50 sq. miles, and traffic was backed up more than three miles, and the road was closed with no warning or signage. I spoke with friends and there were similar events elsewhere in town. If you're in an area where they do a 'surprise' closure, you're screwed and won't be able to move anywhere for minimum of 20-30 minutes.
It's staggering how much the level of interference with daily life a visit from the president has taken on. About six years ago when the president was in town, there was a modest 'convoy' and perhaps a few closures of small streets, and that was about it.
Today, entire parts of a major metropolitan city are shut down, and there's a 25-mile radius no fly zone wherever the president is.
I'd argue that this really isn't needed, directly.
While it's an independent project, XFCE is sort of a philosophical fork of Gnome. It's a very simple, clean design, like the old Gnome. It uses GTK+, like Gnome. But it uses a nice unix-like design methodology (small, additive modules).
And you can tell from the postings here a lot of the old gnome fans are moving there to begin with.
XFCE could use a few more tweaks (I don't like their toolbar setup as it stands) but it certainly seems to me to be the right place to start if you're a fan of the old gnome 1.4.
I'm pretty much a gnome fan, but in reading the thread linked to in the story, I have to agree: every gnome since 1.4 has just felt "off" to me.
And having dealt with the hell of compiling gnome on slack, I can't blame Pat a bit.
Funny thing is, although I still use gnome, I've got one box running XFCE and it feels much more like gnome 1.4 did -- I'll probably migrate there as long as I can count on a few GTK+ apps (mostly gnumeric, gvim, and I'll toy with giving up evolution if needed.)
KDE has just never done it for me. I can't put a finger on it, it just doesn't feel right or "open" (yes, I see the irony here.)
The main things that originally attracted me to gnome were a few well-done apps and the clean simplicity of 1.4 -- if only the gnomesters would go back to this root.
Whatever the case, I'd like to echo sentiments here (and on the forum linked to in the article) -- it'd be great if Pat would include a well-integrated Dropline package with slackware, and if Dropline would consider a second 'standard' slackware i486 distro, as this can be counted to run on practically all platforms (the i686 won't.)
Not quite "micro" fluidics, but automatic transmissions have genuine fluid logic. If you've ever seen the inside of the control area, it even looks like a PCB/Chip layout.
So this stuff does get used in real life, only it's much bigger. Of course the total gate complexity is in the single digits, however.
On one of the old soviet space missions (I think it was the Salyut 7 space stations in the early 80s) one of the mix-gender crews requested privacy curtains, and the implication of sex was there though the women claimed their behavior was stictly professional.
NASA pretty much has said it's never happened on one of their missions, even with the best possibility being a 1992 shuttle mission with a husband and wife on the same crew, but they had opposite shifts and reports were also that nothing happened.
Anyway, I'd bet the answer is yes, and that it was the old-era Soviets who did it first.
I've seen other suggestions for this, I'll give you my exact configuration. I'm running a DNS, web, mail and firewall services using my setup (off of static IP on ISDN no less, 24/7!)
I use a VIA CL6000 (this is a dual lan motherboard with a 600 MHz fanless "Eden" processor) with slackware, 256MB of memory, and a 40GB laptop hard disk (complete overkill, 8GB would be plenty). Total cost of the system was well under $400. Power consumption is about 25 watts, and the box is completely silent. I omit the optical drive since I just "borrowed" one to do the initial install, everything has been via the network since. Uptime's been great.
I've been tempted by the Soekis stuff as well, but cost wise it looks like it'd be a near wash, maybe just a bit cheaper. The ITX stuff is a "real" PC, so you just fire it up and go, no CF config, console emulation via serial port, etc. (I had previously used a CF card on an earlier VIA server, it works if you make sure you put the right things into a RAM disk first.)
As others have pointed out, a cheap laptop would work, however I found the fact that I wanted firewall service (two E-net ports needed) made things a bit odd, as all the used cheap LTs I had included no network adapters, so it would have been dual PCMCIA or USB ethernet, and it just felt and looked really kludgey when I played with it.
I posted in a reply that doesn't appear to be getting modded up, so:
The figures for Intel's total share are worldwide, not US. (I should know, my company is the source cited in the link.) Meanwhile the AMD weekly share data (from another company) is for US Retail system sales. So the two data points really aren't comparable on any basis.
I know the figures I cite are exclusive to x86 CPUs. Someone mentioned PowerPC in this thread, and Apple provides sales figures as part of their financials -- based on Q2 data, PowerPCs in Apples comprise about 1.8% of the market if you included them in the calculations.
I wonder if they're bring Bea Aurthur back for this one, too?
On the bright side, given how low the bar was set for the Star Wars Christmas Special, we're pretty much assured that anything that's done for TV by George is going to be better. (OK, bright side in the brown dwarf star sense.)
I've followed project censored for many years, and even when the left was in power, they were left-leaning. So it's an editorial bias on their part, not due simply to who is in power. You can easily find the lists of censored stories from the peak of the clinton era to see for yourself if this is true.
That said, the slant doesn't necessarily invalidate the stories covered. But it would be nice to see some of the under-reported issues that would embarrass the left get similar coverage by them (discounting the coverage of such stories by histrionic radio talk show hosts, of course.)
Let me preface this with the fact that I'm a diehard slackware fan. If I used the Knoppix-based myth distro this all would have been close to plug and play.
My first attempt didn't go well. Unfortunately MythTV has a lot of dependencies and was a tweaking nightmare as it took a couple days of off-and-on effort for me to get to the point where I could just compile. Then work got busier and I set the project aside.
I finally ended up using the excellent walkthrough located here. I still ended up with some tweaking, but that was primarily the result of some bad permissions on a shared MySQL file.
The other thing I did that was a bit unusual is that I have a 4DTV analog and digital C-Band satellite receiver. I wrote a small python script that translates DirectTV program codes to tune the same channel on C-band, so the on-screen guide works (4DTV is yet to be supported by the OSG data supplier, hence this workaround.)
Anyway, it's been solid once I got it up and running. My next project is to attach the 7" touchscreen LCD display I got to use as a remote control head (mostly for the stereo so I don't have to fire up the plasma TV in another room just to listen to music.)
As far as how it compares to Tivo -- all my friends that have visited and seen the setup like the MythTV a lot better, especially the music and weather options. The main downside is that the best output on the PVR-350 is s-video, so you're not really driving the display at HDTV quality.
I can only echo the advice for using a PVR-250 (actually I suggest the PVR-350, which is encode and decode).
I've got a MythTV box running on a VIA mini-ITX using the slowest CPU offered, a fanless 533 MHz C3 "Eden." With the PVR-350 it loafs along at sub 5% CPU loading. The only downside to this slow of a CPU is occasionally there's a lag in menu selections or screens that need to do a lot of database work to update, but given the power savings and the fact that the box can be fanless, it's worth it.
Even cooler, I was able to use MythMusic to rip all my CDs that had been living in a 200 CD changer -- now I've got on screen menu, random access to the collection with complete artist/title info, instead of having to look up that a favorite CD is #153 in the changer. And I can create any playlist I want with minimal effort.
Tivo doesn't do anything like this -- one more advantage for MythTV.
Start adding up the cost of consumer electronic devices a Myth box can potentially replace, and it doesn't take long to make a compelling argument based on expense alone.
You can have Pentium M desktops now -- I'm posting from one. They're not cheap, however.
My primary PC uses a Lippert "Thunderbird" mini-ITX Pentium M motherboard. Unfortunately this is not by any means cheap, as it's ~$1K for the mobo w/ CPU. However, you can make a very fast, very low power, very small, near-silent PC using it, and those extra features were worth the cost [for me]. I can't imagine a mainstream desktop manufacturer would accept the cost premium associated with this particular technology (yes, they could get a motherboard MUCH cheaper, but the CPUs still carry a signficant price premium over their desktop counterparts.)
Secondly, it's highly doubtful Alviso is delayed due to some desktop dream. Alviso is rumored to support a faster CPU bus interface -- which is rumored to show up on CPUs that won't be coming out until early next year. So the delay is more of a packaging issue -- the delays could be due to either the new CPUs or the chip set, but you need both for a viable platform.
You're correct. Intel commonly uses a 6T cell structure for the SRAMs they use as process testing chips. They used to use DRAMs as a process driver up until the mid-1980s, but then switched to SRAMs after they left the DRAM business. Since most CPUs today are half SRAM (cache) anyway, this makes sense.
The manufacturers have the choice of using multichip module packaging (common in notebook graphics controllers, for example) or single die, however it is my current understanding we're talking a single die.
They very likely WILL disable the dud and sell them as single core CPUs. This is how the "value" brands (Celeron, ex-Duron, and now Sempron) are typically created -- when there's a defect in the processor cache (which is a very large area of the die, and thus more likely to have a defect), the faulty bank(s) are turned off via fusing, creating a CPU with a smaller cache.
This is all pretty standard yield management.
Also, your calculations are very close to being correct, while the manufacturers closely guard their yield information, you're in the ballpark -- and it's interesting to note according to my estimates Intel's Celeron volumes approximately mirror your computed single-core yield percentage... meaning it will likely be business as usual in our dual core future.
BTW, if you're interested in computing yield values there's an excellent model to be had in one fo the chapters in Henessy and Paterson's _Computer Architecture, a Quantitative Approach_
I needed to replace my Seagate TR5 tape drive with something faster (and cheaper, as it turns out.)
Portable USB enclosures for 2.5" disks are CHEAP. Some are sub $15, and $30 will get you Firewire and USB. I went through a few vendors looking for something that was well supported under linux, ended up with something using Prolific chips. Enclosure + 40GB 9.5mm hard disk was less than $100, all it took was a screwdriver included with with enclosure to put it together. Newegg had everything I needed.
Nicest part -- if you pick lower power drives like Fujitsus, the thing is easily powered by a powered USB hub, no extra cables/adapters required. Makes backups a breeze.
I'm reasonably sure you can do it with an 80GB drive for under $150.
Cherry usually makes excellent keyboards, but this key-for-everything mindset just mystifies me. Logitech seems to have the same disease. Perhaps it's the typical consumer more-is-better mindset, but you'd think Cherry would realize the average linux user is FAR from the typical PC consumer.
I wish they'd make something like the Happy Hacking keyboard -- which I am pretty happy with but it appears to only have ~ 2-3 key rollover which is problematic at high typing speeds. Cherry usually has N-key rollover, which would be very attractive to me in the same form factor/layout as the happy hacking KB.
Questionable MCS issues aside, I've definitely encountered a few electronic items that set me off, and I'm perfectly healthy save for a the usual pollen allergies.
The one thing they had in common was that I could smell phenols. Most printed circuit boards and certain classes of inductors seem to be loaded with them. (Chloroseptic sore-throat spray uses phenol as an active ingredient if you need a taste/smell reference.) Anyone who has worked around a PCB plant will probably confirm there's some seriously nasty stuff used in their manufacture, some of which you buy in a new mobo.
What may actually being going on is the phenols are what you can smell, but some other chemical in PCBs might be the real offender, though phenols don't exactly have a clean bill of health -- but they're detectable by your nose at really low PPMs so they're easy to point to.
Usually, however, the the phenol smell (and presumably other emissions) fades pretty quickly with time and are proportional to heat, so if I have a really bad device I put in the warmest room of my house running 24/7 with a window open for a while. After the stuff bakes out it generally isn't much of a problem.
I see your point. The dynamic at work here is (I think) that Intel generally banks on having the ability to scale up the on-die cache size as needed, so that likely has biased their design decisions against integrated memory controllers (along with the bad taste from Timna) for a while now. Obviously the decisions impact choices far beyond the chip design, as is BTX is an example of.
Rephrasing your point, they're committed to the non-integrated path by their choices. As AMD is to theirs -- so this should be a particularly interesting competition because as AMD grows the competition appears to be evolving into more of an engineering contest between the two divergent approaches.
Um, Intel is quite capable of on board memory controllers.
They did it many years ago as part of the ill-fated "Timna" project, which integrated a rambus memory controller on the processor. It was a cool and pretty performance oriented design, however it was intended for the value segment of the market -- and it was using a memory that was far more costly than SDRAM, so Intel killed it.
Whatever their reasons, the memory controller isn't on their current processors because it was design choice, not due to lack of ability to do so. AMD obviously made a different design choice here -- so this is more a question of decisions made rather than ability.
Slashdot Mirror
I do research on this (among other things) for a living.
With current component prices, even using the cheapest of the cheapest parts, you can't get there if you're building an identifiable PC using sustainable components -- IE something manufactured for the market rather and an opportunistic purchase that you can't count on.
Cheapest x86 CPUs available (not inventory blowouts, but manufacturered at thin margins) -- OEM price of about $20. Cheapest chip sets, $15. Ditto hard disk, $40. Ditto optical drive, $18. You're almost at $100 in cost there, and we've not touched a case or RAM yet, yet alone the motherboard manufacturing costs, support chips, system assembly.
If a $100 PC was do-able, those $200 Walmart PCs would be $100 Walmart PCs.
If you redefine the computer, eliminate any signficant removable or internal storage and use a small amount of flash to boot-load remotely, use a tiny memory footprint, and use a tiny non-x86 integrated CPU/system logic device, like an ARM or something similar... in other words, if you build something that looks a heck of a lot like a PDA without a screen -- then it's marginally do-able. But all you're really doing is moving the high-cost components upstream.
Think back to a sinclair ZX-80 or something similar... that's do-able today, with better performance and way more memory.
A box using new components that runs a fat OS with a disk drive, it's not possible today for $100. A thin linux (or windows CE, for that matter) client booting a remote copy of the OS... on the edge of possible, but it's not a PC.
This is doable, if you're willing to compromise on not being a "complete" graphics solution, IE, not trying to implement hardware accelerated everything.
I'd suggest simply a dumb frame buffer, and doing everything in software. Then your solution is simply a memory controller, plus a CRT controller and an output graphics DAC. Really, if you implemented a VGA controller from the early 90s sans the backware compatible bits for MDA,CGA, EGA, etc, and opened up the line/pixel resolution you'd be there.
If you make the goals modest enough, this could probably be done with a field-programmable gate array, an external graphics DAC, and some RAM. The only tricky part is external DACs are hard to come by these days and aren't cheap. (Prior to their integration in current-era parts, they were down in the $1 range, but since they're integrated now, the only stuff that's commonly available are insanely high end workstation DACs.)
Back in the mid-late 90s, right as 2D acceleration was hitting its peak and 3D acceleration was emerging, there were netlists of VGA designs for sale for as little as $500. So designs of this level aren't hard.
If you're thinking even full 2D acceleration, it gets much harder. and if you're thinking 3D acceleration in an "open" project that would be competitive with even the slowest Nvidia/ATI parts... you're on drugs.
One alternative would be to approach an existing vendor about opening up an "old" product. However, getting a fresh production run of an old product wouldn't be cheap -- you're basically talking a million to get masks made, initial wafer lots, etc. Hence the FPGA suggestion, since that's commonly available hardware that doesn't require any manufacturing specific to the design.
However, some manufacturers may still be in low volume production with a suitable product. Someone who used to be small player in the PC graphics market years ago but isn't now would be a candidate -- perhaps a Silicon Motion or an Avance Logic (part of Realtek now, I think, though I doubt the video products are still active, tho the audio parts are) could be persuaded to open up a part that's in sustained low-volume production.
Seriously, though, if you can't offer significant volumes -- the minimum probably being on the order of 10-20K/quarter, and that's VERY small in this business -- don't expect to get much help from existing vendors.
Parent reply post is on the mark here.
Where the confusion comes up is in the old days where interlacing was first used -- black and white television. Interlaced TV signals drove black and white CRTs... and by selecting the right phosphor, the displays had persistence, where the image would continue to glow into the next frame even after being drawn. High persistance phosphor DID cut down on flicker, and was necessary because of the interlacing.
If you want to get into really obscure stuff... Radar display tubes (and some oscilloscopes) often had incredible persistence -- often lasting several to even tens of seconds. Surplus radar tubes were popular among ham radio operators doing slow-scan television, as they could maintain the TV display at the insanely low frame rates used at the time (I think it was about 1 frame per 8 seconds or so, memory is hazy here.)
Anyway, think of the displays long ago as a giant analog visual low-pass filter -- you could throw pretty much any signal at them regarless of flicker and get a flicker-free display, so the bandwidth savings from interlacing was an obvious choice. Not so with digital technology today.
Actually, not that powerful, relatively speaking. A typical 121.5 EPIRB puts out 75-125 milliwatts. Keep in mind this is VHF and the satellites are typically NOAA birds in low-earth orbit expecting a line of sight signal, so 100 milliwatts is very workable.
The issue with 121.5 EPIRBs is all they do is send a warble tone -- no ID, location, nothing. All the processing is done by the rest of the infrastructure, and even then the output is basically a position (still no ID) to within a mile or so -- with people using radio-direction finders narrowing it down more.
The newer 406 MHz EPIRBs have specific user data and location information transmitted in their digital packets, so not only do they know the where, but they also know the who -- so when they get a boat beacon originating at someone's house they pretty much already know it's a false alarm.
So, basically in a nutshell:
You don't use HF, don't need it, and don't know any service that does, so screw them.
Pretty shortsighted and impressively incomplete view of this problem. Not all communications are local, which is what every VHF/UHF method you mentioned is. HF is in ONLY radio-based communication system that provides global coverage without needing substantial infrastructure such as satellites, etc.
I realize this is a joke, but driving in the phoenix/scottsdale area today was insane. The roads and sky were effectively removed, I'm afraid.
As I was driving south in the upper northern area of scottsdale (adjacent to Phoenix) this morning, I saw police cars on EVERY possible right-turn option for more than two miles (streets, parking lots, you name it). I noticed many of the cars had trunks partially opened with barricades inside. On my return trip two hours later I made a point of taking a route that would avoid this area -- barely.
Good thing.
They had completely blocked off the road (Scottsale Road) in both directions, immediately before the freeway on-ramp. This road is effectively the only exit draining traffic for an area for about 50 sq. miles, and traffic was backed up more than three miles, and the road was closed with no warning or signage. I spoke with friends and there were similar events elsewhere in town. If you're in an area where they do a 'surprise' closure, you're screwed and won't be able to move anywhere for minimum of 20-30 minutes.
It's staggering how much the level of interference with daily life a visit from the president has taken on. About six years ago when the president was in town, there was a modest 'convoy' and perhaps a few closures of small streets, and that was about it.
Today, entire parts of a major metropolitan city are shut down, and there's a 25-mile radius no fly zone wherever the president is.
I'd argue that this really isn't needed, directly.
While it's an independent project, XFCE is sort of a philosophical fork of Gnome. It's a very simple, clean design, like the old Gnome. It uses GTK+, like Gnome. But it uses a nice unix-like design methodology (small, additive modules).
And you can tell from the postings here a lot of the old gnome fans are moving there to begin with.
XFCE could use a few more tweaks (I don't like their toolbar setup as it stands) but it certainly
seems to me to be the right place to start if you're a fan of the old gnome 1.4.
I'm pretty much a gnome fan, but in reading the thread linked to in the story, I have to agree: every gnome since 1.4 has just felt "off" to me.
And having dealt with the hell of compiling gnome on slack, I can't blame Pat a bit.
Funny thing is, although I still use gnome, I've got one box running XFCE and it feels much more like gnome 1.4 did -- I'll probably migrate there as long as I can count on a few GTK+ apps (mostly gnumeric, gvim, and I'll toy with giving up evolution if needed.)
KDE has just never done it for me. I can't put a finger on it, it just doesn't feel right or "open" (yes, I see the irony here.)
The main things that originally attracted me to gnome were a few well-done apps and the clean simplicity of 1.4 -- if only the gnomesters would go back to this root.
Whatever the case, I'd like to echo sentiments here (and on the forum linked to in the article) -- it'd be great if Pat would include a well-integrated Dropline package with slackware, and if Dropline would consider a second 'standard' slackware i486 distro, as this can be counted to run on practically all platforms (the i686 won't.)
Not quite "micro" fluidics, but automatic transmissions have genuine fluid logic. If you've ever seen the inside of the control area, it even looks like a PCB/Chip layout.
So this stuff does get used in real life, only it's much bigger. Of course the total gate complexity is in the single digits, however.
On one of the old soviet space missions (I think it was the Salyut 7 space stations in the early 80s) one of the mix-gender crews requested privacy curtains, and the implication of sex was there though the women claimed their behavior was stictly professional.
NASA pretty much has said it's never happened on one of their missions, even with the best possibility being a 1992 shuttle mission with a husband and wife on the same crew, but they had opposite shifts and reports were also that nothing happened.
Anyway, I'd bet the answer is yes, and that it was the old-era Soviets who did it first.
I've seen other suggestions for this, I'll give you my exact configuration. I'm running a DNS, web, mail and firewall services using my setup (off of static IP on ISDN no less, 24/7!)
I use a VIA CL6000 (this is a dual lan motherboard with a 600 MHz fanless "Eden" processor) with slackware, 256MB of memory, and a 40GB laptop hard disk (complete overkill, 8GB would be plenty). Total cost of the system was well under $400. Power consumption is about 25 watts, and the box is completely silent. I omit the optical drive since I just "borrowed" one to do the initial install, everything has been via the network since. Uptime's been great.
I've been tempted by the Soekis stuff as well, but cost wise it looks like it'd be a near wash, maybe just a bit cheaper. The ITX stuff is a "real" PC, so you just fire it up and go, no CF config, console emulation via serial port, etc. (I had previously used a CF card on an earlier VIA server, it works if you make sure you put the right things into a RAM disk first.)
As others have pointed out, a cheap laptop would work, however I found the fact that I wanted firewall service (two E-net ports needed) made things a bit odd, as all the used cheap LTs I had included no network adapters, so it would have been dual PCMCIA or USB ethernet, and it just felt and looked really kludgey when I played with it.
I posted in a reply that doesn't appear to be getting modded up, so:
The figures for Intel's total share are worldwide, not US. (I should know, my company is the source cited in the link.) Meanwhile the AMD weekly share data (from another company) is for US Retail system sales. So the two data points really aren't comparable on any basis.
I know the figures I cite are exclusive to x86 CPUs. Someone mentioned PowerPC in this thread, and Apple provides sales figures as part of their financials -- based on Q2 data, PowerPCs in Apples comprise about 1.8% of the market if you included them in the calculations.
Actually, the lead paragraph mistakenly says the 82% figure is US -- it's not, it's ~82% worldwide.
The AMD > 50% figures are specific to US Retail sales, so they are totally uncomparable numbers.
I wonder if they're bring Bea Aurthur back for this one, too? On the bright side, given how low the bar was set for the Star Wars Christmas Special, we're pretty much assured that anything that's done for TV by George is going to be better. (OK, bright side in the brown dwarf star sense.)
I've followed project censored for many years, and even when the left was in power, they were left-leaning. So it's an editorial bias on their part, not due simply to who is in power. You can easily find the lists of censored stories from the peak of the clinton era to see for yourself if this is true.
That said, the slant doesn't necessarily invalidate the stories covered. But it would be nice to see some of the under-reported issues that would embarrass the left get similar coverage by them (discounting the coverage of such stories by histrionic radio talk show hosts, of course.)
Let me preface this with the fact that I'm a diehard slackware fan. If I used the Knoppix-based myth distro this all would have been close to plug and play.
My first attempt didn't go well. Unfortunately MythTV has a lot of dependencies and was a tweaking nightmare as it took a couple days of off-and-on effort for me to get to the point where I could just compile. Then work got busier and I set the project aside.
I finally ended up using the excellent walkthrough located here. I still ended up with some tweaking, but that was primarily the result of some bad permissions on a shared MySQL file.
The other thing I did that was a bit unusual is that I have a 4DTV analog and digital C-Band satellite receiver. I wrote a small python script that translates DirectTV program codes to tune the same channel on C-band, so the on-screen guide works (4DTV is yet to be supported by the OSG data supplier, hence this workaround.)
Anyway, it's been solid once I got it up and running. My next project is to attach the 7" touchscreen LCD display I got to use as a remote control head (mostly for the stereo so I don't have to fire up the plasma TV in another room just to listen to music.)
As far as how it compares to Tivo -- all my friends that have visited and seen the setup like the MythTV a lot better, especially the music and weather options. The main downside is that the best output on the PVR-350 is s-video, so you're not really driving the display at HDTV quality.
I can only echo the advice for using a PVR-250 (actually I suggest the PVR-350, which is encode and decode).
I've got a MythTV box running on a VIA mini-ITX using the slowest CPU offered, a fanless 533 MHz C3 "Eden." With the PVR-350 it loafs along at sub 5% CPU loading. The only downside to this slow of a CPU is occasionally there's a lag in menu selections or screens that need to do a lot of database work to update, but given the power savings and the fact that the box can be fanless, it's worth it.
Even cooler, I was able to use MythMusic to rip all my CDs that had been living in a 200 CD changer -- now I've got on screen menu, random access to the collection with complete artist/title info, instead of having to look up that a favorite CD is #153 in the changer. And I can create any playlist I want with minimal effort.
Tivo doesn't do anything like this -- one more advantage for MythTV.
Start adding up the cost of consumer electronic devices a Myth box can potentially replace, and it doesn't take long to make a compelling argument based on expense alone.
You can have Pentium M desktops now -- I'm posting from one. They're not cheap, however.
My primary PC uses a Lippert "Thunderbird" mini-ITX Pentium M motherboard. Unfortunately this is not by any means cheap, as it's ~$1K for the mobo w/ CPU. However, you can make a very fast, very low power, very small, near-silent PC using it, and those extra features were worth the cost [for me]. I can't imagine a mainstream desktop manufacturer would accept the cost premium associated with this particular technology (yes, they could get a motherboard MUCH cheaper, but the CPUs still carry a signficant price premium over their desktop counterparts.)
Secondly, it's highly doubtful Alviso is delayed due to some desktop dream. Alviso is rumored to support a faster CPU bus interface -- which is rumored to show up on CPUs that won't be coming out until early next year. So the delay is more of a packaging issue -- the delays could be due to either the new CPUs or the chip set, but you need both for a viable platform.
You're correct. Intel commonly uses a 6T cell structure for the SRAMs they use as process testing chips. They used to use DRAMs as a process driver up until the mid-1980s, but then switched to SRAMs after they left the DRAM business. Since most CPUs today are half SRAM (cache) anyway, this makes sense.
The manufacturers have the choice of using multichip module packaging (common in notebook graphics controllers, for example) or single die, however it is my current understanding we're talking a single die.
They very likely WILL disable the dud and sell them as single core CPUs. This is how the "value" brands (Celeron, ex-Duron, and now Sempron) are typically created -- when there's a defect in the processor cache (which is a very large area of the die, and thus more likely to have a defect), the faulty bank(s) are turned off via fusing, creating a CPU with a smaller cache.
This is all pretty standard yield management.
Also, your calculations are very close to being correct, while the manufacturers closely guard their yield information, you're in the ballpark -- and it's interesting to note according to my estimates Intel's Celeron volumes approximately mirror your computed single-core yield percentage... meaning it will likely be business as usual in our dual core future.
BTW, if you're interested in computing yield values there's an excellent model to be had in one fo the chapters in Henessy and Paterson's _Computer Architecture, a Quantitative Approach_
I needed to replace my Seagate TR5 tape drive with something faster (and cheaper, as it turns out.)
Portable USB enclosures for 2.5" disks are CHEAP. Some are sub $15, and $30 will get you Firewire and USB. I went through a few vendors looking for something that was well supported under linux, ended up with something using Prolific chips. Enclosure + 40GB 9.5mm hard disk was less than $100, all it took was a screwdriver included with with enclosure to put it together. Newegg had everything I needed.
Nicest part -- if you pick lower power drives like Fujitsus, the thing is easily powered by a powered USB hub, no extra cables/adapters required. Makes backups a breeze.
I'm reasonably sure you can do it with an 80GB drive for under $150.
Cherry usually makes excellent keyboards, but this key-for-everything mindset just mystifies me. Logitech seems to have the same disease. Perhaps it's the typical consumer more-is-better mindset, but you'd think Cherry would realize the average linux user is FAR from the typical PC consumer.
I wish they'd make something like the Happy Hacking keyboard -- which I am pretty happy with but it appears to only have ~ 2-3 key rollover which is problematic at high typing speeds. Cherry usually has N-key rollover, which would be very attractive to me in the same form factor/layout as the happy hacking KB.
Questionable MCS issues aside, I've definitely encountered a few electronic items that set me off, and I'm perfectly healthy save for a the usual pollen allergies.
The one thing they had in common was that I could smell phenols. Most printed circuit boards and certain classes of inductors seem to be loaded with them. (Chloroseptic sore-throat spray uses phenol as an active ingredient if you need a taste/smell reference.) Anyone who has worked around a PCB plant will probably confirm there's some seriously nasty stuff used in their manufacture, some of which you buy in a new mobo.
What may actually being going on is the phenols are what you can smell, but some other chemical in PCBs might be the real offender, though phenols don't exactly have a clean bill of health -- but they're detectable by your nose at really low PPMs so they're easy to point to.
Usually, however, the the phenol smell (and presumably other emissions) fades pretty quickly with time and are proportional to heat, so if I have a really bad device I put in the warmest room of my house running 24/7 with a window open for a while. After the stuff bakes out it generally isn't much of a problem.
I see your point. The dynamic at work here is (I think) that Intel generally banks on having the ability to scale up the on-die cache size as needed, so that likely has biased their design decisions against integrated memory controllers (along with the bad taste from Timna) for a while now. Obviously the decisions impact choices far beyond the chip design, as is BTX is an example of.
Rephrasing your point, they're committed to the non-integrated path by their choices. As AMD is to theirs -- so this should be a particularly interesting competition because as AMD grows the competition appears to be evolving into more of an engineering contest between the two divergent approaches.
Um, Intel is quite capable of on board memory controllers.
They did it many years ago as part of the ill-fated "Timna" project, which integrated a rambus memory controller on the processor. It was a cool and pretty performance oriented design, however it was intended for the value segment of the market -- and it was using a memory that was far more costly than SDRAM, so Intel killed it.
Whatever their reasons, the memory controller isn't on their current processors because it was design choice, not due to lack of ability to do so. AMD obviously made a different design choice here -- so this is more a question of decisions made rather than ability.