I had heard that the museum was "small but pretty interesting". That ended up definitely being an under-sell.
The Computer History Museum in Mountain View is cool and all, but the Cryptologic Museum struck me on an entirely different level. Instead of the "Here is how computing evolved" theme of the Mountain View museum, I really felt like this was the "Here is why computation is relevant to communications (and warfare)" counterpart. They display voice and data encryption tools of the last five decades, from STE's and STU-III's back to (as other posters mentioned) the mechanically-synchronized SIGSALY machine that used giant turning vinyl records to encrypt the traffic. There is a handset you can pick up to hear pre-recorded messages representing the voice quality of each system. The oldest were barely intelligible, the newest are (obviously) crystal clear.
The Cray XMP and YMP are impressive, and are in almost flawless condition! Rather than the exhibit at Mountain View, it felt like these machines were just recently taken out of service, and could easily be made operational again. They didn't seem like they'd been cobbled back together or had sat in closets neglected and falling apart for years. The density of some of the components on the Thinking Machines CM-5 memory and processor slices is impressive, and the descriptions of the power and cooling apparatus required (think many kilowatts and lots of Fluorinert) were equally amazing -- truly a testament to what can be done when money isn't much of an object, and a machine's value is measured solely in MIPS or MFLOPS.
There is a three-foot-tall full-relief wooden replica of the Great Seal of the U.S. on the wall, which apparently was a gift from Russian schoolchildren to the U.S. embassador in Moscow. After hanging prominently on the wall for years in the embassador's office in Moscow, in 1952 it was discovered that it contained a resonant cavity eavesdropping bug on the inside that was very difficult to detect with sensing equipment of the time, unless it was activated by radio signal (presumably by Soviet spies) from the outside. I met there three (very proud) tourists of Russian descent who chuckled heartily at that one (and who tried to teach me how to say "Medvedev" properly, thanks!)
As everyone else mentioned, the working Enigma machine was fun to encipher a message to a friend with (they have a pad and pencil for you to use), and the displays on the history of the agency and of the Korean and Vietnam conflicts were well put together. The GRAB II and Poppy ELINT satellites were especially interesting to me, and reminded me of the kind of things a senior class at the USAF Academy might build for a project these days (relics of an era when launch considerations and electronics density actually drove simplicity into designs).
If you're an electronics/history/information assurance/security/aerospace/DC trivia fan, you'll almost certainly enjoy the trip, even if the facility is kind of small and out of the way. While you're in the area, go see the Udvar-Hazy center, too! And don't forget to tip your docents...
I've been wondering lately why no common file systems seem to implement error correcting codes (ECC/EDAC).
In hardware, there's often a checksum, ECC/Hamming code, parity bit, Reed-Solomon code, etc. to detect and/or correct for inadvertent bit flips. But, as far as I know, no error correcting information is ever stored within the filesystem itself. Certainly the filesystem tracks how many blocks are dedicated to a particular file, and how many bytes long the file is, and one can always hash the file twelve ways to Sunday to assure that it hasn't changed since it was originally hashed, but none of that helps repair errors to the file should the medium that's being used to store it decay beyond what's already correctable via the medium access hardware.
I can imagine scenarios where, for example, the RAM buffer in a hard drive is upset and perfectly encodes the wrong bit into a file (or even multiple stripes + parity in a RAID). In this case, the medium access hardware is useless (the data was, after all, ecoded perfectly wrong), but ECC in the filesystem would detect and potentially correct the error the next time the file was read back, even if it were decades later. I appreciate that it would add overhead, and thus maybe shouldn't be the default, but I don't see it being even an option anywhere, and some people would pay the performance penalty to get the data integrity benefit.
Especially in instances like encrypted (or compressed, or both) loopback file systems where one bad bit can destroy an entire partition, why don't we have more data assurance layers available? Or have I just not found them?
Whining of which, what was the deal with GNU ecc? Everyone speaks of "oh, yeah, the algorithm was deeply flawed, bummer..." but I don't ever see any details...
For some reason I'd thought that these highly-efficient multi-junction cells made by Spectrolab generally quote peak efficiency statistics as based on the ideal spectrum of incident solar light found in space, i.e. the orbits of the (usually geostationary) communications satellites where these cells are often encountered (hence why Boeing has the interest in Spectrolab). I thought I remembered that when you aactually account for the spectrum of visible light that makes it through the atmosphere, the efficiency goes down quite a bit.
But TFA hints at terrestrial applications, so maybe I'm just pulling this out of thin air, so to speak...
As each CPU is tested and binned especially for power dissipation AND maximum clock speed, this margin is low and the gains minimal. And you spend a lot of time to find out what is the lowest safe voltage.
It should probably be mentioned here that the "lowest safe voltage" (if there can be said to be such a thing) is temperature-dependent (a function of the effectiveness of a fan, the density of the air, the load on the regulators, the number of components powered up vs. down...), and can even be data-dependent.
In order for a CPU to work properly, an awful lot of "digital ballet" needs to be happening, billions of times per second, in perfect harmony. Failing a single setup or hold check on a clocked logic gate leads to erroneous bits entering your data and control streams. Quite a few logic blocks, implemented in quite a few different transistor-level CMOS techniques, have different timings depending on what data is flowing through them at the time -- i.e. whether certain bits are zeroes and others ones, or whether the bit transitions from a one to a zero at the same time as an adjacent wire is switching, etc.
What I'm saying, of course, is that if you "spend a lot of time to find out what is the lowest safe voltage", and then (oops) you plug in a USB peripheral that you weren't using before, or (oops) your room heats up a little, or (oops) you use a laptop cooler to get it off your lap a little (setup and hold times, as well as clock and data paths are voltage dependent y'know...) or (oops) you access an unusual data pattern in a program you haven't run in a while, or (oops) you use your laptop on an airplane, your happy little CPU (that you just had to eke out that last little bit of thermal margin from!) starts failing. Quietly, sometimes.
If you're lucky, Windows just crashes -- but manages not to trash your hard drive while doing so (remember, this is weeks or months after you ran your little three-day burn-in marathon "torture test"). But in a not-so-pleasant outcome, your data just starts quietly... rotting. Binary file formats start getting corrupted. Programs stop running. Checksums start failing. Spreadsheets start changing quietly behind your back. I grant you that on a Windows-based laptop this sort of thing could generally be considered the norm (due to virii and spyware and so on), so you might not really notice.
But fortunately for the retarded selves of the people "publishing" (I use the term loosely) this crap, there are thousands of engineers at Intel and AMD running detailed and comprehensive timing simulations of all of the tens of millions of transistors in those teeny-tiny CPUs. They can't possibly EDAC- and parity-protect every net in the design, but they do run static and dynamic timing checks, with probabalistically-developed parasitics and 3-dimensional noise parameters. They consider crosstalk and electrical noise, process variability, and electromigration. They run extensive tests on the CPUs at the wafer and package level to detect slight variations in current consumed during carefully-chosen JTAG test vector execution. They speed-bin and scrap parts that don't pass the multi-billion-dollar test regimens that they spent close to a decade perfecting, and use the detailed results to craft application notes and engineering design guidelines that they feed to their OEM partners, who in turn carefully couple the CPUs to circuit boards with traces perfectly matched to one another, as well as to the impedance characteristics of the solder they're using, and attach them to advanced power supplies and controllable oscillators.
And then you go ahead and download a shitty little ninety-nine cent program off teh unternet and procede to pick your own CPU voltage that "seems to work OK".
Yeah, good luck with that. Sounds like a plan to me. There once was this handy expression about fools and money -- I guess "data" is just the modern substitute.
This is actually a point that's interesting to me but seems to keep getting neglected. Can any lawyers, almost-lawyers, once-lawyers, or pseudo-lawyers comment on the following questions?
1) Does downloading a copyrighted work place the downloader in violation of any presently existing U.S. federal statute or common-law precedent? If so, which one(s)? I.e. is downloading actually illegal (I'm almost convinced it is not!)? Is it duplication? For extra credit, make any relevant arguments about state-specific issues. I live in California, if that's interesting to you.
2) Presumably, voluntarily providing a copyrighted work for upload (specifically, one for which you do not own the copyright) violatates federal copyright law if that content is subsequently voluntarily downloaded, because the uploading is considered electronic distribution and duplication (yes?). But are there any mitigating factors that can absolve the uploader from liability? For example, suppose that copyrighted content was removed from the uploader's computer without his or her knowledge or consent. Has duplication taken place? Distribution? Is the uploader still liable? What if the downloading was performed automatically by a computer script, e.g. a virus?
3) Have any threatening lawsuits by copyright holders been issued naming downloaders as defendants (who were not also uploaders)? If such a lawsuit were issued (filed? submitted?), how would jurisdiction and venue be determined? Would such a lawsuit be, on its face, an abuse of process, in the sense that (depending on your answer to 1 above, downloading probably isn't actually a crime), akin to suing someone for being ugly?
I know slashdot is a horrible place for asking legal advice, but I have seen a couple of IAAL's making interesting and relevant points in the past...
> Some 9 years ago I worked on some chip design for Hughes
I work at what's left of one of the old Hughes Space & Comm. digital ASIC groups, with several of the old(er) Hughes Radar fellas.
> Back then, we used 1.2um on 4" (or 6" in the new fab) wafers -
> and everything was built on a sapphire substrate instead of a silicon substrate...
SOS is much less prevalent today. My first design was quarter micron Si CMOS, and many new designs are tenth and sub-tenth micron. It's becoming more difficult to convince foundries to perform space-qualified work. Thinner gate oxides, rising leakage current, clock distribution power dissipation, and shrinking feature sizes in libraries are making good standard cell ASIC development for the space environment more difficult, and the full custom stuff is pricier. Modern COTS hardware is having a more and more difficult time meeting satisfactory (particularly end-of-life) performance requirements. And, to boot, obsolescence issues are continuing to rear their ugly heads. Technology qualification is getting more expensive every year, and the spaceborne commercial telecom market was ravaged by overcapacity.
To top it off, the hardest part is actually finding people who can write rugged, professional, reliable code (and test plans, and device specifications, and interface documents, and...) day after day.
> It was dull, as every single chip had about 12 inches of paperwork from QA.
> That was then. Can anyone let me know how much things have changed?
> I had a buddy who was a gear head and use to race drag
> when we were in high school...
>
> His car? A pinto. The car is so damned light that it
> beats a lot of muscle cars for the 1/4, and nothing, I mean
> nothing, is worth more than the look on the face of
> someone who was just beaten by a car known far as[sic] wide
> for it's[sic] lack of anything worthy.
>
> I get a similar feeling when people realize they were
> just owned by my wife at CS.:)
And, tell us, just how is the look on your wife's face when she checks your slashdot account and sees that she's on the wrong end of that particular Pinto analogy?
I'm leaving aside whether one can even be "owned" at Computer Science at all, of course (I believe you were looking for "5ko0l3d", there).
> Dick Rutan will find a way to have this craft go up once more, a new (modified)
> design will be built which fixes this instability and SpaceShipOne will go to the
> Smithsonian before it hurts anyone.
When Dick figures out a way to have this craft go up once more, do you think he'll be kind
enough to share it with his brother, Burt?
You have conveniently misread the contents of your own link. While I can see how one might misread the awkward Frank Davies/Knight Ridder summary you linked to above, the somewhat more precise PIPA survey data should sort things out for you.
You claimed that "80% of misinformed Americans get thier information from FOX news". The truth is much closer to the following:
Rep. pg. 7: 60% of respondents had at least one of the three listed misperceptions about the Iraq-al Quaeda link, WMD, and world opinion opposing the war
Quest. pg. 5: 80% of poll respondents selected "From TV and radio" as "where [they] tend to get most of [their] news".
Quest. pg. 6: 18% of poll respondents (presumably 18% of those listing TV and radio as their primary source, though it's not specified) selected Fox News as "[their] primary sources[sic] of news"
Rep. pg. 13: 80% of those poll respondents who selected Fox News as their primary source of news also had one or more of the three "misconceptions" listed, namely that "evidence of Iraq-al Qaeda links have been found, WMD have been found[,] and [that] world public opinion favored [the] Iraq war"
I claim the following:
Among 1,362 respondents, 817 are "misinformed", given the criterion of having at least one of the three listed "misconceptions".
Among 1,362 respondents, 1090 selected TV and radio as their principal news source.
Among 1,090 respondents listing TV and radio as their principal news source, 196 listed Fox News as their primary news source.
Among 196 people listing Fox News as their primary news source, 157 had at least one of the given misperceptions.
Among the 817 misinformed respondents, 157 use Fox News as their primary source of news.
Therefore, 157 / 817= 19% of misinformed respondents "got their news from FOX News". 19% != 80%. In fact, one might say it's quite the opposite.
I now direct you to repeat whichever course failed to teach you Bayes' theorem.
I've read a lot of discussions lately about recent evidence for why there must, at one time, have been liquid water on Mars. But, much of that evidence relates to the deposition of sediment, presence of erosion patterns, aftereffects of evaporation, presence of salts, crystallization patterns, and so forth -- none of which (to my knowledge) requiring the liquid in question to be H2O. Some of the evidence, on the other hand, relates to the formation of minerals such as hematite, which presumably form only in or near liquid H2O, and not, say, liquid H2O2, liquid CO2, or liquid N2. The biggest question(s) I have that I've not seen well addressed are:
1. What evidence supports or rules out the presence of liquids other than H2O on the surface of Mars, at one time, in large quantities?
2. How much, if any, of the present evidence could be explained by flows of liquid CO2, nitrogen, methane, ammonia, or some other liquid?
3. Which evidence, if any, points most strongly to the presence of large amounts of H2O as the liquid in question? I know there are currently thought to be large, polar caps of solid H2O, but how much of the current evidence precludes the existence of large seas of some other liquid in the distant geological past?
I apologize if these questions are simple or completely baseless. I am not a geologist, and am legitimately curious.
> Do you think getting a Ph.D. in CompSci or CompEng will improve or
> worsen my career outlook in the industry?
First of all, since when are Computer Science and Computer Engineering interchangeable?
Secondly, of which industry do you speak? You make it sound as though there's only one. Do you want to be the chief tech support monkey? Do you want to be the chief CPU architectural design monkey? Faculty? Algorithmic development and analysis for you? Do you want to manage a semiconductor research group? Do you want to perform cryptanalysis? Do you want to start your own company? Do you want to join a recently-started company, a Fortune 500, are you just indiscriminate?
If you have to ask whether it's for you, I suspect perhaps it isn't. If you have to Ask Slashdot, it makes me all that much more sure.
I hate being terse, especially at this time of morning, but I'm really becoming disillusioned by this "Is Path A likely to make me more employable than Path B?" mentality. The world would be better off if we each just focused more closely on our own interests.
In my limited experience, jobs, prestige, power, fame, and fortune follow those with a passion for what they do, who pay less attention to their resume than their own personal satisfaction. If you want to find a job (people exchanging their money for your time) that you love to do, find something that you love to do which happens to make money for others (ideally, you should make them more money than they pay you, on a time-averaged basis).
Otherwise, it may be easier to use whatever job you can tolerate to support your hobbies and extracurricular personal interests. In this case, if you derive satisfaction from foregoing four years of income to achieve a significant academic milestone, great! If you want candy for job interviews, lead a large group of stubborn people to great financial successes in spite of themselves, and document the process with buzzwords.
Ph.D.s don't make you rich, and they sure don't hold you back. That mushy grey stuff between your ears is what does that.
You know, the only person I remember at Purdue who worked on GPS was a fellow named Jim Garrison. As I recall, he also worked at Goddard at one point, and his research involved measuring reflected GPS signals.
You two should get together and chat. I imagine you'd have a lot to talk about.
First, I have to say, driving past a computerized billboard with a fifty-foot-wide BSOD (or Windows logon screen) hovering prominently on the side of the interstate is a vastly amusing experience. Three weeks ago, a Toyota dealership off of I-405 provided me with this lovely visual gag.
What I'm really wondering about, though, is this. Ever since I first started reading about Boeing's Digital Cinema, I've been curious whether people would now start to use theaters for things other than feature films. Once the medium for displaying visual and auditory effects shifts from film to bitstreams, one could conceivably show the World Series, the State of the Union address, reruns of the Simpsons, or the 2004 Iron Chef Steel Cage Deathmatch Season Finale, in real time or through rebroadcasting. I don't know that these things would necessarily draw large crowds, or that you could get them to cough up much money per person for this, but you'd still be selling plenty of overpriced Ju Ju Bes and Fizzy Sugar Water(TM), and wouldn't be paying for the rights to show a new release instead.
What I'm really asking is: given the interesting things people find to do when the size of their display changes drastically, what new and interesting things would you do with a digital cinema?
Finally, if you were a theater owner, would you choose to get your digital bitstreams off an encrypted copper/fiber network, off of encrypted ROM/DVDs, downlink from SATCOM, etc? How do the relative merits stack up?
I opened up my machine sometime in December in order to
inspect it for this very problem. Because I'd read in
the previous slashdot article that Abit motherboards had
experienced this problem, and my motherboard was (is)
an Abit KT7-Raid (non -A flavor), I was particularly
curious. Sure enough, two capacitors had clearly bulged
open and were leaking paste, and three more were on their
way towards failing. Interestingly, I hadn't had any observable symptoms whatsoever; I just checked the board on a whim.
I consulted the Abit website, and at the time they required
the original 'invoice' from the motherboard if you wanted
to have them repair the problem for free. If you have that
paperwork, RMA'ing the board should not be too troublesome.
I really didn't think I could find my old paperwork for this
board. Abit offers to repair motherboards without original
invoices for a charge of (as I recall) $25.00US. I think
you have to pay shipping one way.
I considered using the services of the guy linked to in
the previous slashdot article, but his prices were about
in line with Abit's. That didn't really help any -- for
the amount of money he wanted, I could just have just had Abit do it. I could also have just as easily replaced the board for $45.00 plus shipping on Ebay, but it probably would have just failed all over again.
I felt I could replace the capacitors myself, and as it
turns out, I was right. Here's my advice to anyone who
wants to try to do this repair themselves:
Obtain a temperature-controlled soldering iron with
a nice pointy tip. I used a Weller brand iron. You can
get away with a constant-power iron if you're good, careful, lucky, or some combination thereof.
Obtain either a solder sucker that you're comfortable
using, or the desoldering braided copper wick that's sold for
this purpose. My experience was that the solder paste that's
used on these PWBs does not wick well. It certainly doesn't
wick the way a higher quality silver solder wicks. I got by
anyway, but it was a bit ugly at times.
It is helpful to have a lighted magnifying lens, a pair
of small pliers, some good solder, and a circuit board
holding jig. You can get away with less. I didn't use a jig.
It goes (almost) without saying that you should by now have removed all the easily-removable goodies from the motherboard (RAM, CPU, etc.), and that the motherboard should be fully removed from its case. You should give some thought to static control and ESD, of course.
My board had three through-holes for each capacitor, only
two of which were occupied by the capacitor's leads. I suspect
this is for interchangeability of capacitor models.
Obtain good new capacitors. I had to search for 'Low-ESR
2200 microfarad 6V radial electrolytics'. A higher voltage rating
is fine, even recommended, but can increase the physical size. I went with 10V-rated
capacitors, which were a touch large, but workable. The capacitors
I found were rated to 85 degrees Celsius, but 105's are available (and also
recommended). I advise against replacing the capacitors with anything other than the previous capacitance rating,
although you could probably get away with it. Some caps are rated
as 'computer-grade' or some such. This is generally good. Digikey
offered good capacitors for $3.00 apiece in small quantities. I
found a small shop in southern California near where I live which
charged 65 cents apiece. Radio Shack and Fry's are unlikely to have acceptable parts in stock, even if either carries them.
An iron temperature around 790 degrees Fahrenheit worked well
for me. Conversion to Kelvin is left as an exercise for the reader.
The power capacitors on my board were a fair distance away
from any delicate CPU traces. That helped ease my conscience a bit.
Hopefully, yours will be too.
Using the hot iron and desoldering braid, gently remove as much
solder as you can from the underside of the capacitor leads, starting on the back face (non-component side) of the board. Remove
the capacitors from the board, and thread the leads of the new ones
through the exposed holes. POLARITY MATTERS! There is likely to be
a polarity indicator on your capacitors, you should match the current
polarity (assuming your board manufacturer didn't screw that up, too.
Some have.) The polarity indicator typically looks like a painted
stripe along one edge of the cap, indicating that the outermost radial
lead is (conventional) negative/ground.
Solder on the new capacitors, making sure to deposit a nice,
adequate but not excessive volume of solder. Make sure not to leave
a cold (badly-formed) solder joint. Make sure that there's little
play in the capacitor afterwards, but that the leads aren't straining
their joints.
Good luck. Don't blame me if you scew something up, burn something out,
or get someone killed. Send it back to the manufacturer if you're not
up to the task, or don't have much experience in such matters. There
is a touch to this sort of repair that comes from practice, I think, and
the only way to develop it is to get your hands dirty. Or burnt. Or
something. Previous to this, I had only a little experience with this sort of rework, so don't be too shy.
Besides, Natalie Portman demands a guy with trained fingers.
There's a good story about the windshield washer pumps that I haven't seem anybody tell yet. Before I tell it, though, I need to say a few quick things. The first is that I've spent a fair amount of time with Harvey Mudd students recently, and they tend to be seriously gifted individuals. I've also spent time with M.I.T., CalTech, and Rose Hulman students (to name a few), and I have to say that the Mudders are seriously underrated on the global undergraduate stage. These folks are personable, inquisitive, dedicated, and enormously talented. And funny. Really funny.
Forgive me if I'm telling it wrong, this is second- and third-hand information.
Not long after the bar monkey became operational, it was being used over the course of an evening to serve cocktails in the lounge. When the students retired for the evening, they deactivated the monkey and left to finish their 'stems homework and whatnot. Later that night, there was a localized power disruption, though I don't know its duration. What I do know is that the monkey wasn't on a UPS of any sort (and for $200, can you really blame them?) and lost power.
When power was restored, and the machine booted back into linux, the parallel port data bits were apparently all lifted high at some point in the boot process. This, unfortunately, meant that all of the liquor-dispensing windshield wiper pumps were briefly activated. Now, as someone who has hooked a large inductive load to a DC power supply can attest, the momentary current draw of having all of these pumps simultaneously active was not negligible. In fact, it overwhelmed the power supply, but only after a shot of Bottle 1 + Bottle 2 had been dispensed onto the floor. Vodka and rum, I believe it was.
Now, when the power supply sensed its overcurrent condition (I believe this is how the story goes), it did the noble thing and: rebooted. You can see where this is going: lather, rinse, repeat. The rebooting webmonkey once again lifted its data pins high, once again tripped the power supply overcurrent sensor, and once again rebooted the monkey (all the while dispensing yet another tasty shot of Bottle 1+Bottle 2 onto the floor).
By the time someone came back to check on the monkey in the morning (a few hours later), the monkey had apparently drenched the floor in vodka, and was still rebooting. I'm told only bottles 1 and 2 were drained, and that the rest of the precious nectar was left intact inside the monkey.
I'm convinced that nobody who witnessed the aftermath of this event will ever again neglect the power cycling / bootstrapping phase of an electronic device. I don't have a link to it, but I seem to remember AT&T having a similar problem with their switching software across the Atlantic seaboard, maybe six or seven years ago. You can draw your own conclusions, but if I were trying to hire a hardware engineer (not that there's really any other kind, you wimpy IT CS MIS MSCE slashdot derelicts), I'd give much greater weight to their ability to explain the projects they've completed (and the associated lessons learned), than I would to a 4.0 GPA or a Cisco certificate of trainability.
Mudd grads (math, science, engineering, and other) are presently worth four times their weight in silver the first year after they graduate -- and that's a bargain. They should be worth at least one twenty-fifth their weight in rhodium.
Slashdot Mirror
I had heard that the museum was "small but pretty interesting". That ended up definitely being an under-sell.
The Computer History Museum in Mountain View is cool and all, but the Cryptologic Museum struck me on an entirely different level. Instead of the "Here is how computing evolved" theme of the Mountain View museum, I really felt like this was the "Here is why computation is relevant to communications (and warfare)" counterpart. They display voice and data encryption tools of the last five decades, from STE's and STU-III's back to (as other posters mentioned) the mechanically-synchronized SIGSALY machine that used giant turning vinyl records to encrypt the traffic. There is a handset you can pick up to hear pre-recorded messages representing the voice quality of each system. The oldest were barely intelligible, the newest are (obviously) crystal clear.
The Cray XMP and YMP are impressive, and are in almost flawless condition! Rather than the exhibit at Mountain View, it felt like these machines were just recently taken out of service, and could easily be made operational again. They didn't seem like they'd been cobbled back together or had sat in closets neglected and falling apart for years. The density of some of the components on the Thinking Machines CM-5 memory and processor slices is impressive, and the descriptions of the power and cooling apparatus required (think many kilowatts and lots of Fluorinert) were equally amazing -- truly a testament to what can be done when money isn't much of an object, and a machine's value is measured solely in MIPS or MFLOPS.
There is a three-foot-tall full-relief wooden replica of the Great Seal of the U.S. on the wall, which apparently was a gift from Russian schoolchildren to the U.S. embassador in Moscow. After hanging prominently on the wall for years in the embassador's office in Moscow, in 1952 it was discovered that it contained a resonant cavity eavesdropping bug on the inside that was very difficult to detect with sensing equipment of the time, unless it was activated by radio signal (presumably by Soviet spies) from the outside. I met there three (very proud) tourists of Russian descent who chuckled heartily at that one (and who tried to teach me how to say "Medvedev" properly, thanks!)
As everyone else mentioned, the working Enigma machine was fun to encipher a message to a friend with (they have a pad and pencil for you to use), and the displays on the history of the agency and of the Korean and Vietnam conflicts were well put together. The GRAB II and Poppy ELINT satellites were especially interesting to me, and reminded me of the kind of things a senior class at the USAF Academy might build for a project these days (relics of an era when launch considerations and electronics density actually drove simplicity into designs).
If you're an electronics/history/information assurance/security/aerospace/DC trivia fan, you'll almost certainly enjoy the trip, even if the facility is kind of small and out of the way. While you're in the area, go see the Udvar-Hazy center, too! And don't forget to tip your docents...
I've been wondering lately why no common file systems seem to implement error correcting codes (ECC/EDAC).
...
In hardware, there's often a checksum, ECC/Hamming code, parity bit, Reed-Solomon code, etc. to detect and/or correct for inadvertent bit flips. But, as far as I know, no error correcting information is ever stored within the filesystem itself. Certainly the filesystem tracks how many blocks are dedicated to a particular file, and how many bytes long the file is, and one can always hash the file twelve ways to Sunday to assure that it hasn't changed since it was originally hashed, but none of that helps repair errors to the file should the medium that's being used to store it decay beyond what's already correctable via the medium access hardware.
I can imagine scenarios where, for example, the RAM buffer in a hard drive is upset and perfectly encodes the wrong bit into a file (or even multiple stripes + parity in a RAID). In this case, the medium access hardware is useless (the data was, after all, ecoded perfectly wrong), but ECC in the filesystem would detect and potentially correct the error the next time the file was read back, even if it were decades later. I appreciate that it would add overhead, and thus maybe shouldn't be the default, but I don't see it being even an option anywhere, and some people would pay the performance penalty to get the data integrity benefit.
Especially in instances like encrypted (or compressed, or both) loopback file systems where one bad bit can destroy an entire partition, why don't we have more data assurance layers available? Or have I just not found them?
Whining of which, what was the deal with GNU ecc? Everyone speaks of "oh, yeah, the algorithm was deeply flawed, bummer..." but I don't ever see any details
For some reason I'd thought that these highly-efficient multi-junction cells made by Spectrolab generally quote peak efficiency statistics as based on the ideal spectrum of incident solar light found in space, i.e. the orbits of the (usually geostationary) communications satellites where these cells are often encountered (hence why Boeing has the interest in Spectrolab). I thought I remembered that when you aactually account for the spectrum of visible light that makes it through the atmosphere, the efficiency goes down quite a bit.
But TFA hints at terrestrial applications, so maybe I'm just pulling this out of thin air, so to speak...
Anybody?
As each CPU is tested and binned especially for power dissipation AND maximum clock speed, this margin is low and the gains minimal. And you spend a lot of time to find out what is the lowest safe voltage.
It should probably be mentioned here that the "lowest safe voltage" (if there can be said to be such a thing) is temperature-dependent (a function of the effectiveness of a fan, the density of the air, the load on the regulators, the number of components powered up vs. down...), and can even be data-dependent.
In order for a CPU to work properly, an awful lot of "digital ballet" needs to be happening, billions of times per second, in perfect harmony. Failing a single setup or hold check on a clocked logic gate leads to erroneous bits entering your data and control streams. Quite a few logic blocks, implemented in quite a few different transistor-level CMOS techniques, have different timings depending on what data is flowing through them at the time -- i.e. whether certain bits are zeroes and others ones, or whether the bit transitions from a one to a zero at the same time as an adjacent wire is switching, etc.
What I'm saying, of course, is that if you "spend a lot of time to find out what is the lowest safe voltage", and then (oops) you plug in a USB peripheral that you weren't using before, or (oops) your room heats up a little, or (oops) you use a laptop cooler to get it off your lap a little (setup and hold times, as well as clock and data paths are voltage dependent y'know...) or (oops) you access an unusual data pattern in a program you haven't run in a while, or (oops) you use your laptop on an airplane, your happy little CPU (that you just had to eke out that last little bit of thermal margin from!) starts failing. Quietly, sometimes.
If you're lucky, Windows just crashes -- but manages not to trash your hard drive while doing so (remember, this is weeks or months after you ran your little three-day burn-in marathon "torture test"). But in a not-so-pleasant outcome, your data just starts quietly... rotting. Binary file formats start getting corrupted. Programs stop running. Checksums start failing. Spreadsheets start changing quietly behind your back. I grant you that on a Windows-based laptop this sort of thing could generally be considered the norm (due to virii and spyware and so on), so you might not really notice.
But fortunately for the retarded selves of the people "publishing" (I use the term loosely) this crap, there are thousands of engineers at Intel and AMD running detailed and comprehensive timing simulations of all of the tens of millions of transistors in those teeny-tiny CPUs. They can't possibly EDAC- and parity-protect every net in the design, but they do run static and dynamic timing checks, with probabalistically-developed parasitics and 3-dimensional noise parameters. They consider crosstalk and electrical noise, process variability, and electromigration. They run extensive tests on the CPUs at the wafer and package level to detect slight variations in current consumed during carefully-chosen JTAG test vector execution. They speed-bin and scrap parts that don't pass the multi-billion-dollar test regimens that they spent close to a decade perfecting, and use the detailed results to craft application notes and engineering design guidelines that they feed to their OEM partners, who in turn carefully couple the CPUs to circuit boards with traces perfectly matched to one another, as well as to the impedance characteristics of the solder they're using, and attach them to advanced power supplies and controllable oscillators.
And then you go ahead and download a shitty little ninety-nine cent program off teh unternet and procede to pick your own CPU voltage that "seems to work OK".
Yeah, good luck with that. Sounds like a plan to me. There once was this handy expression about fools and money -- I guess "data" is just the modern substitute.
This is actually a point that's interesting to me but seems to keep getting neglected. Can any lawyers, almost-lawyers, once-lawyers, or pseudo-lawyers comment on the following questions?
1) Does downloading a copyrighted work place the downloader in violation of any presently existing U.S. federal statute or common-law precedent? If so, which one(s)? I.e. is downloading actually illegal (I'm almost convinced it is not!)? Is it duplication? For extra credit, make any relevant arguments about state-specific issues. I live in California, if that's interesting to you.
2) Presumably, voluntarily providing a copyrighted work for upload (specifically, one for which you do not own the copyright) violatates federal copyright law if that content is subsequently voluntarily downloaded, because the uploading is considered electronic distribution and duplication (yes?). But are there any mitigating factors that can absolve the uploader from liability? For example, suppose that copyrighted content was removed from the uploader's computer without his or her knowledge or consent. Has duplication taken place? Distribution? Is the uploader still liable? What if the downloading was performed automatically by a computer script, e.g. a virus?
3) Have any threatening lawsuits by copyright holders been issued naming downloaders as defendants (who were not also uploaders)? If such a lawsuit were issued (filed? submitted?), how would jurisdiction and venue be determined? Would such a lawsuit be, on its face, an abuse of process, in the sense that (depending on your answer to 1 above, downloading probably isn't actually a crime), akin to suing someone for being ugly?
I know slashdot is a horrible place for asking legal advice, but I have seen a couple of IAAL's making interesting and relevant points in the past...
> Some 9 years ago I worked on some chip design for Hughes
I work at what's left of one of the old Hughes Space & Comm. digital ASIC groups, with several of the old(er) Hughes Radar fellas.
> Back then, we used 1.2um on 4" (or 6" in the new fab) wafers -
> and everything was built on a sapphire substrate instead of a silicon substrate...
SOS is much less prevalent today. My first design was quarter micron Si CMOS, and many new designs are tenth and sub-tenth micron. It's becoming more difficult to convince foundries to perform space-qualified work. Thinner gate oxides, rising leakage current, clock distribution power dissipation, and shrinking feature sizes in libraries are making good standard cell ASIC development for the space environment more difficult, and the full custom stuff is pricier. Modern COTS hardware is having a more and more difficult time meeting satisfactory (particularly end-of-life) performance requirements. And, to boot, obsolescence issues are continuing to rear their ugly heads. Technology qualification is getting more expensive every year, and the spaceborne commercial telecom market was ravaged by overcapacity.
To top it off, the hardest part is actually finding people who can write rugged, professional, reliable code (and test plans, and device specifications, and interface documents, and...) day after day.
> It was dull, as every single chip had about 12 inches of paperwork from QA.
> That was then. Can anyone let me know how much things have changed?
Yeah -- we're up to about 50 inches.
> I had a buddy who was a gear head and use to race drag :)
> when we were in high school...
>
> His car? A pinto. The car is so damned light that it
> beats a lot of muscle cars for the 1/4, and nothing, I mean
> nothing, is worth more than the look on the face of
> someone who was just beaten by a car known far as[sic] wide
> for it's[sic] lack of anything worthy.
>
> I get a similar feeling when people realize they were
> just owned by my wife at CS.
And, tell us, just how is the look on your wife's face when she checks your slashdot account and sees that she's on the wrong end of that particular Pinto analogy?
I'm leaving aside whether one can even be "owned" at Computer Science at all, of course (I believe you were looking for "5ko0l3d", there).
> Dick Rutan will find a way to have this craft go up once more, a new (modified)
> design will be built which fixes this instability and SpaceShipOne will go to the
> Smithsonian before it hurts anyone.
When Dick figures out a way to have this craft go up once more, do you think he'll be kind
enough to share it with his brother, Burt?
- Photons have mass.
Photons are Catholic?!? Well, we're all catholic.def 2
I direct you to the document entitled Misperceptions, the Media and the Iraq War (Rep. below) and its source questionnaire (Quest. below).
You claimed that "80% of misinformed Americans get thier information from FOX news". The truth is much closer to the following:
I claim the following:
Therefore, 157 / 817= 19% of misinformed respondents "got their news from FOX News". 19% != 80%. In fact, one might say it's quite the opposite.
I now direct you to repeat whichever course failed to teach you Bayes' theorem.
I've read a lot of discussions lately about recent evidence for why there must, at one time, have been liquid water on Mars. But, much of that evidence relates to the deposition of sediment, presence of erosion patterns, aftereffects of evaporation, presence of salts, crystallization patterns, and so forth -- none of which (to my knowledge) requiring the liquid in question to be H2O. Some of the evidence, on the other hand, relates to the formation of minerals such as hematite, which presumably form only in or near liquid H2O, and not, say, liquid H2O2, liquid CO2, or liquid N2. The biggest question(s) I have that I've not seen well addressed are:
1. What evidence supports or rules out the presence of liquids other than H2O on the surface of Mars, at one time, in large quantities?
2. How much, if any, of the present evidence could be explained by flows of liquid CO2, nitrogen, methane, ammonia, or some other liquid?
3. Which evidence, if any, points most strongly to the presence of large amounts of H2O as the liquid in question? I know there are currently thought to be large, polar caps of solid H2O, but how much of the current evidence precludes the existence of large seas of some other liquid in the distant geological past?
I apologize if these questions are simple or completely baseless. I am not a geologist, and am legitimately curious.
Cheers,
F00F> Do you think getting a Ph.D. in CompSci or CompEng will improve or
> worsen my career outlook in the industry?
First of all, since when are Computer Science and Computer Engineering interchangeable?
Secondly, of which industry do you speak? You make it sound as though there's only one. Do you want to be the chief tech support monkey? Do you want to be the chief CPU architectural design monkey? Faculty? Algorithmic development and analysis for you? Do you want to manage a semiconductor research group? Do you want to perform cryptanalysis? Do you want to start your own company? Do you want to join a recently-started company, a Fortune 500, are you just indiscriminate?
If you have to ask whether it's for you, I suspect perhaps it isn't. If you have to Ask Slashdot, it makes me all that much more sure.
I hate being terse, especially at this time of morning, but I'm really becoming disillusioned by this "Is Path A likely to make me more employable than Path B?" mentality. The world would be better off if we each just focused more closely on our own interests.
In my limited experience, jobs, prestige, power, fame, and fortune follow those with a passion for what they do, who pay less attention to their resume than their own personal satisfaction. If you want to find a job (people exchanging their money for your time) that you love to do, find something that you love to do which happens to make money for others (ideally, you should make them more money than they pay you, on a time-averaged basis).
Otherwise, it may be easier to use whatever job you can tolerate to support your hobbies and extracurricular personal interests. In this case, if you derive satisfaction from foregoing four years of income to achieve a significant academic milestone, great! If you want candy for job interviews, lead a large group of stubborn people to great financial successes in spite of themselves, and document the process with buzzwords.
Ph.D.s don't make you rich, and they sure don't hold you back. That mushy grey stuff between your ears is what does that.
You know, the only person I remember at Purdue who worked on GPS was a fellow named Jim Garrison. As I recall, he also worked at Goddard at one point, and his research involved measuring reflected GPS signals.
You two should get together and chat. I imagine you'd have a lot to talk about.
First, I have to say, driving past a computerized billboard with a fifty-foot-wide BSOD (or Windows logon screen) hovering prominently on the side of the interstate is a vastly amusing experience. Three weeks ago, a Toyota dealership off of I-405 provided me with this lovely visual gag.
What I'm really wondering about, though, is this. Ever since I first started reading about Boeing's Digital Cinema, I've been curious whether people would now start to use theaters for things other than feature films. Once the medium for displaying visual and auditory effects shifts from film to bitstreams, one could conceivably show the World Series, the State of the Union address, reruns of the Simpsons, or the 2004 Iron Chef Steel Cage Deathmatch Season Finale, in real time or through rebroadcasting. I don't know that these things would necessarily draw large crowds, or that you could get them to cough up much money per person for this, but you'd still be selling plenty of overpriced Ju Ju Bes and Fizzy Sugar Water(TM), and wouldn't be paying for the rights to show a new release instead.
What I'm really asking is: given the interesting things people find to do when the size of their display changes drastically, what new and interesting things would you do with a digital cinema?
Finally, if you were a theater owner, would you choose to get your digital bitstreams off an encrypted copper/fiber network, off of encrypted ROM/DVDs, downlink from SATCOM, etc? How do the relative merits stack up?
I opened up my machine sometime in December in order to inspect it for this very problem. Because I'd read in the previous slashdot article that Abit motherboards had experienced this problem, and my motherboard was (is) an Abit KT7-Raid (non -A flavor), I was particularly curious. Sure enough, two capacitors had clearly bulged open and were leaking paste, and three more were on their way towards failing. Interestingly, I hadn't had any observable symptoms whatsoever; I just checked the board on a whim.
I consulted the Abit website, and at the time they required the original 'invoice' from the motherboard if you wanted to have them repair the problem for free. If you have that paperwork, RMA'ing the board should not be too troublesome. I really didn't think I could find my old paperwork for this board. Abit offers to repair motherboards without original invoices for a charge of (as I recall) $25.00US. I think you have to pay shipping one way.
I considered using the services of the guy linked to in the previous slashdot article, but his prices were about in line with Abit's. That didn't really help any -- for the amount of money he wanted, I could just have just had Abit do it. I could also have just as easily replaced the board for $45.00 plus shipping on Ebay, but it probably would have just failed all over again.
I felt I could replace the capacitors myself, and as it turns out, I was right. Here's my advice to anyone who wants to try to do this repair themselves:
Good luck. Don't blame me if you scew something up, burn something out, or get someone killed. Send it back to the manufacturer if you're not up to the task, or don't have much experience in such matters. There is a touch to this sort of repair that comes from practice, I think, and the only way to develop it is to get your hands dirty. Or burnt. Or something. Previous to this, I had only a little experience with this sort of rework, so don't be too shy.
Besides, Natalie Portman demands a guy with trained fingers.
There's a good story about the windshield washer pumps that I haven't seem anybody tell yet. Before I tell it, though, I need to say a few quick things. The first is that I've spent a fair amount of time with Harvey Mudd students recently, and they tend to be seriously gifted individuals. I've also spent time with M.I.T., CalTech, and Rose Hulman students (to name a few), and I have to say that the Mudders are seriously underrated on the global undergraduate stage. These folks are personable, inquisitive, dedicated, and enormously talented. And funny. Really funny.
Forgive me if I'm telling it wrong, this is second- and third-hand information.
Not long after the bar monkey became operational, it was being used over the course of an evening to serve cocktails in the lounge. When the students retired for the evening, they deactivated the monkey and left to finish their 'stems homework and whatnot. Later that night, there was a localized power disruption, though I don't know its duration. What I do know is that the monkey wasn't on a UPS of any sort (and for $200, can you really blame them?) and lost power.
When power was restored, and the machine booted back into linux, the parallel port data bits were apparently all lifted high at some point in the boot process. This, unfortunately, meant that all of the liquor-dispensing windshield wiper pumps were briefly activated. Now, as someone who has hooked a large inductive load to a DC power supply can attest, the momentary current draw of having all of these pumps simultaneously active was not negligible. In fact, it overwhelmed the power supply, but only after a shot of Bottle 1 + Bottle 2 had been dispensed onto the floor. Vodka and rum, I believe it was.
Now, when the power supply sensed its overcurrent condition (I believe this is how the story goes), it did the noble thing and: rebooted. You can see where this is going: lather, rinse, repeat. The rebooting webmonkey once again lifted its data pins high, once again tripped the power supply overcurrent sensor, and once again rebooted the monkey (all the while dispensing yet another tasty shot of Bottle 1+Bottle 2 onto the floor).
By the time someone came back to check on the monkey in the morning (a few hours later), the monkey had apparently drenched the floor in vodka, and was still rebooting. I'm told only bottles 1 and 2 were drained, and that the rest of the precious nectar was left intact inside the monkey.
I'm convinced that nobody who witnessed the aftermath of this event will ever again neglect the power cycling / bootstrapping phase of an electronic device. I don't have a link to it, but I seem to remember AT&T having a similar problem with their switching software across the Atlantic seaboard, maybe six or seven years ago. You can draw your own conclusions, but if I were trying to hire a hardware engineer (not that there's really any other kind, you wimpy IT CS MIS MSCE slashdot derelicts), I'd give much greater weight to their ability to explain the projects they've completed (and the associated lessons learned), than I would to a 4.0 GPA or a Cisco certificate of trainability.
Mudd grads (math, science, engineering, and other) are presently worth four times their weight in silver the first year after they graduate -- and that's a bargain. They should be worth at least one twenty-fifth their weight in rhodium.