See this slideset, in particular the one titled "DOS File System" (slide 16.)
For a 256MB flash, it'd be FAT16. FAT16 only limits the root directory size to 512 entries for a disk type media (from here.)
Elsewhere it was said that they (NASA's engineers) weren't able to reproduce the problem exactly. I'm still very curious about this... as I said elsehere I've used DOS FAT16 under an RTOS (Thread-X) with lots of files and didn't see this kind of thing happen. Then again we hammered hell out of the FS and even fixed some problems with it.
See this slideset, in particular the one titled "Memory Allocation and Usage" (slide 18 I believe.) They do all their allocation at init time.
So far I see lots of flamage about how VxWorks malloc() sucks, but no concrete explantion why (if?) the number of files caused the crash. Also, there isn't any indication (so far) that the FAT was hosed.
So... what was the real problem? Was it indeed a malloc() failure that led to the reset loop? If so, what was being malloc()'d?
I'm really curious about this, since a project I worked on uses FAT16 on flash; in my case it was a multimedia terminal running Thread-X in 32MB RAM and 64MB Flash, with third-party DOS FAT and TCP/IP add-ons. We had lots of files on this thing, and it didn't seem to flake out like this.
There is work to do just that: Java on top of RT/Linux in the Rocky 7 platform (basically the testbed for MER1/2). See here (PDF file.)
Now the question I have is... which filesystem did they use for MER1/2? Is it DOS FAT? If so, I could see how the "too many files" problem could happen quite easily.
Not that DOS FAT wouldn't be OK for spaceflight, it is very simple, reasonably robust and quite mature. Just gotta watch for those FAT limits.
I didn't list that option because it seemed a roller+film mechanism would be mechanically fragile and heavy given the size it'd need to be for the entire panel set. It'd also cut the panel efficiency somewhat, partially defeating its purpose. Finally, you'd eventually run out of film.
FWIW, there's a set of Smith Roll-Offs in my gear bag somewhere... so I am familiar with the concept.
Sheesh - the JPL guys ought to talk to some desert racers sometime. They deal with more dust in the course of a Baja 500 than most of us do in an entire lifetime.
Anyway, some simple low-risk ideas:
* A small air blower jet to blow the dust off. After all, there is an atmosphere (albeit a thin Martian one) to work with.
* Tilt the panels and give them a gentle shake to get most of the dust off.
* Use an electrostatic coating to keep the fine dust from sticking.
As far as the batteries, couldn't they be deep-cycled to reduce their memory?
Motorcycles uses chains because they're (1) inexpensive, (2) lightweight, (3) lowest unsprung weight, (4) compact and narrow, (5) immune to the 'shaft jacking' (rear end rises and falls due to torque on the rear hub), (5) able to handle lots of power (ZX-12 anyone?), (6) directly compatable with transverse engine layout and finally (7) they're highly efficient when properly cared for.
Some motorcycles use shafts because (1) they're somewhat lower maintenance than chains. That's the sole reason. Yes, there have been some performance improvements in shafties (Paralever for example) but they still find use primarily on models targeted to riders who aren't so concerned about weight and cost and put a premium on maintenance (yes, that'd be you propeller-heads and Gold Wingers.)
And even then I'd argue the maintenance point: an O-ring chain sprayed every 600-1000 miles with a good lube like Maxima Chain Wax will give trouble-free service for a long time. My literbike's chain and sprockets lasted 30,000 miles with this sort of care. Cost of a new chain and sprocket set? About $185, and a half hour to change them out.
Oh yeah, there's belt drive too. S'ok if you don't mind the extra half-inch or so it adds to the width of the drivetrain, not to mention the limited power that realistic belt widths can handle. Also, don't get gravel in one, it'll mess up the sprockets if not damage the belt (I haven't priced belts and sprockets but I bet they're a lot more than a chain set.) However, they don't tend to stretch and need only a little shot of silicone once in a while.
Seems to me that "FLOP" by itself stands for "Floating Point OPeration". Several generations of computer people understand "FLOP" this way, and a MFLOP was a simply million of those (would you like fries with that?), that is, a quantity, not a rate. In the early-mid 80's we spoke of how many FLOPS it would take to do a graphics task, such as a homogeneous 3-d coordinate transform.
I know that Webopedia and suchlike have MFLOP as "mega floating-point operations per second", but that just doesn't make sense: where's the "S" in MFLOP to mean "second"? It ain't there. Hmm....
What's going on here? Maybe there's been a tendency in the casual press to try to coerce "MFLOP(s)" to be a rate like "MIPS" for symmetry's sake. That seems to be where Webopedia went with it. But, it just doesn't work right (that dang "S" is missing yet they tease 'second' out of "MFLOP".) From where I sit, a good and useful close-to-the-machine acronym - FLOP - got sullied by sloppy copywriters and marketing types into a fluff term like MIPS (apologies in advance to Tom, Ray, and Craig - you know who you are.) Tsk, tsk...
So I think the submitter took the high ground and went with decades of historical usage of the acronym FLOP. In that context it was entirely correct to say "TFLOPs/sec"
Suddenly I feel old... recalling when 20 MFLOPs/s was enough to take a company public...
You're forgetting that there are fixed 'service' costs on the bill irrespective of the kWh used. Given Japan's high labor and infrastructure costs these are probably steep.
Also, their kWh usage was probably low in the first place. This is a reasonable assumption given their small home (only 1100 sq. ft.), smaller appliances, and the comparatively frugal (energy-conscious) Japanese lifestyle.
They say they saved 17% on their electric bill and that 'most' (minimum > 50%) of their kWh came from solar. Since solar only affects the kWh used, you could have a case where at most 34% of their (old) bill was kWh and the rest - at least 66% - was fixed 'overhead' service fees.
Think back to when you started programming - what you wanted most of all was to get the computer to do something - anything - under your control. Once you gotten that far, you're hooked....
There's much wisdom in making 'hello, world' a first goal for novice programmers. By the time they've done this they've run through the tool chain at least once. So, make the tool chain to get there as easy and straightforward as possible to build confidence.
If it's a visual suite, that fine but be prepared to train your young coder on it. Personally I think a simple text editor and a shell (with stdin/stdout/stderr) is easier to master at first. It sidesteps the Byzantine GUI 'window dressing' trap and focuses the young mind on the logic of the program itself.
I also have a strong bias toward C/C++ as a starter language because of its compactness and that it can allow a young coder close to the bare metal if that's their proclivity. Some people make a case for Java; mastering its initial environment (classpaths and such) could be daunting, and in many ways it's too restrictive.
Finally, I think providing a dedicated Linux box for your hacker-in-utero helps better foster programming creativity: the tools, languages and sources are all there to be seen, used, and learnt from. Second choice would be OS X, except the hardware is expensive in these days of $100 EPIA boards.
... is what they hammer on in the courses, for good reason: simply put, you can die or be a cripple for life if you screw up. (My course? NAUI in 1976, taught by a Coast Guardsman at Alameda Naval Air Station. No wussy computers then.)
The main benefit of computers is that they simplify the task of figuring safe repetitive dives. In theory they can also give a more accurate plan based on accumulated nitrogen load, and thus allow more bottom time than a strict tables dive. Some computers even include a planning function which provides a simulation of a dive profile, based on its calculated estimate of your N2 load from prior recent dives.
But... the computers can fail. Think about it: you're relying on a relatively low-cost piece of electronic hardware, immersed in saltwater under pressure. There's lots of direct and indirect failure modes... not only could its seal or battery fail, it could be knocked loose, bitten by something etc.
And don't forget user error: if it's like most gizmos you could have forgotten to push the button to start it. Or, worse, you pushed it at depth and so the reading is too shallow.
God forbid it should have a software error like the early Aladdin evidently did.
To protect yourself, prior to your dive you should already have figured out your basic safe dive profile and bottom time given your nitrogen load. Manual tools like the PADI Wheel are used to do this. Then, you have: a watch, a backup gauge, and a buddy (who functions as a redundant set of gear.)
Once you're in the water your plan, your mechanical depth gauge, your watch and your buddy are your friends: they will save your ass (or at least your dive) if that fancy gizmo on your wrist decides to go tits-up at depth.
Yes, I have and use a computer. But I still do a 'sanity check' of the proposed dive using a manual tool like the PADI Wheel or the tables. Actually, if you're like most recreational 'vacation' divers who dive with divemasters, the divemaster does that check for you in case YOU screw up.
More rant: these newer 'air-integrated' computers scare me even more. I won't use one. I recently saw one fail, and the guy fixed the pressure transponder using a spring from a ball-point pen (bragging about 'MacGyvering' the thing) and then DIVED ON IT! Nuts!
My take on the guys who got bent on the early Aladdin is that they were pushing it, relying solely on the computer and doing 'closely-spaced' (repetitive) dives. Had they used manual backup and a lick of commonsense for time-to-fly they would not have gotten the bends.
As pros they should have known better. Yet to my mind they're not much smarter than those Moskito lobster divers I saw marking time in Roatan's hyperbaric chamber at Anthony's Key. Rather than placing faith in an etherial deity like la Moskita, they placed their unquestioning trust in a piece of software. The result is the same: they came up bent.
(Read more about the Moskito epidemic of decompression sickness here.)
The Winbond chip is pretty specific to its application, that is, making motherboard clocks. There are much better serial programmable devices that can provide a wider range of frequencies. You can get Cypress ones at Digi-Key)
Also for more accuracy, you can stack them and refactor P and Q over multiple dividers. On one project (an MPEG encoder) I did just that to make a low-jitter fully-locked 16.9344 / 12.288 / 18.432 audio reference from 27 MHz video. Each PLL was less than $2, and I used an 8051 to control it.
There are also specialty PLL chips used for cellphones that provide good accuracy using some voodoo in their dividers.
Stained glass windows using the lead-and-solder method (lead came) are very bendy and whippy; most of their strength comes from the frame. Large panels generally need to be braced. Also, lead came panels sag over time, especially if the cement (grouting) comes loose.
Lead came construction works for flat windows. As far as making shapes, forget it. It's way too weak.
Ever lift a Tiffany lamp shade? They're pretty hefty, yet, they retain their shape over time and they're not bendy. Tiffany glass (foil-and-solder) is much stronger and resistant to bending, because:
o the foil has adhesive backing to hold the glass
o the finished joint has no gaps (therefore no cementing needed)
o the solder has alloy metals to make it stiffer (e.g., tin) so it doesn't sag.
o the finished piece is much lighter than lead came
Tiffany style construction is a bit more labor-intensive (mostly you need to dress the cut glass edges) but the results are worth it: you have much more freedom in the design, be it a window, lamp, or computer case.
[Disclaimer: I made lamps and terrariums in the 70's, so sue me.]
In short, treat the whole project like a Tiffany lamp.
This case appears to be built that way, although for Tiffany glass it's fugly. Plus he mounts a fan to a glass panel. Bad idea.
If you work stained glass I don't need to explain the advantages of Tiffany style construction. But for the other folks, it's stronger and prettier than 'lead came' ("H" channel lead) construction commonly used in windows.
So, for the glass I would definitely go that way.
As far as mounting the heavy bits, like a lamp I would use a fixture (frame) and keep the glass non-load bearing.
In keeping with the Tiffany lamp theme, use a pretty material for the frame. I'd look into using brass rod and sheet to make a 'perimeter' frame. Use brass lugs to join the pieces. I say this because brass and bronze are common Tiffany lamp fixture materials so it wouldn't look out of place with the stained glass, especially if you applied a patina finish to the metal.
Continuing with this basic frame, you then have a choice of making a glass 'shade' that slips over the frame, individual glass panels that come off, or perhaps a combination of the two. You could get fancy and make hinges for access panels out of tubing and rod.
As far as EMI compliance, forget it; the glass is little better than air for magnetic shielding.
Making a good ground plane under the mobo and to the rear panel is important to prevent ground loops and reduce static discharge, so splurge on the brass sheet there.
Cooling shoudln't be a huge problem so long as you position the fans close to where the airflow is needed.
I might as well take a crack at this too (and I got a "C" in high school chemistry because I insisted on lighting up th acetylene ballons we used in the Gas Density lab...)
To focus on water's phase changes as a heat sink is to miss the point. Surface water is not a major component of Earth's mass, so its ability to smooth out temperature variations as a 'thermal capacitor' would on the face of it be not very significant when compared to the mass of Earth as a whole.
Another fallacy is that the Earth is a closed system. It's not: Earth not only takes in energy from the Sun but it also radiates energy back into space. I should mention that solar radiation isn't the only heat source either: radioactive decay also heats the Earth's core.
Once you understand this, then you appreciate the very reason greenhouse gases are called such: they trap this radiation like the walls of a greenhouse.
The size of the polar ice caps is a reflection of the average Earth temperature, not the other way around. Its effect on mean sea level is an indicator of mean temperature, but it certainly isn't the only one (how about the mean temperature itself?)
More interesting to me is the potential effect of increased water vapor in the air, itself a greenhouse gas. Would it have a positive reinforcing effect on a temperature rise, whether it be due to increased CO2, solar radiation, or cow flatuence? On first glance this phenomenon seems to be notwellunderstood.
Another idea I don't see tossed about much is biological equilibria having influence on CO2 levels. It seems quite reasonable that increased CO2 levels and temperatures would increase the biomass (especially in plankton) and thus the drawdown of CO2 out of the atmosphere.
Global climate is a complicated thing. Unfortuantely it is complicated further by global politics....
I see you're an Accurate Reloading fan. If so, how could you overlook the Ben and the 577 T.Rex video, widely reposted under another, shall we say non-PC title....
... turns out to be just that. Taking a small, inexpensive motor and attaching it to a bicycle? It's a basically sound idea. That is after all how Soichiro Honda got started in the motorcycle business after the war. Putting the engine in the rim? Hey, another nifty idea, but not original: the Megola did that in 1922.
Okay, sometimes synthesizing old ideas into new ones can yield interesting results. At least it makes you think "hmmm, neat hack". I don't see this as a fine example of this. All that gearing and mechanical complexity of the clutches and gearset strike me as expensive and failure-prone: too many moving parts for what it sets out to do. It would seem simpler is better.
Then there's the question of intrinsic value. Mopeds and motorcycles are cheap for the utility and performance they provide when you compare them to today's overpriced yuppie-toy bicycles. You can find a used motorcycle for $1200 or so, and mopeds even less than that. Mileage? Well, my CBR900RR gets 55mpg, and I don't have to worry about looking good in Spandex. Pollution? Many have pointed out that even 2-stroke motors can be engineered to be quite 'green', such as Aprilia's direct-injected 2-stroke scooter using technology licensed from Orbital Engine Corp.
A small confession: I hate having my time wasted. My personal work ethic and my own desire to have some control over my fate makes me really sensitive to the "GFAR" type of command. I think most bright people feel the same way, unless they're so worn down that they merely accept the status quo and just roll with it (a very dangerous apporach, career-wise.) Intelligence, experience, knowledge and initiative are good things. The question is, how can you use these to your advantage in this type of losing scenario?
I think the number one thing you can do is ask questions. You're not in the military, you're not going to be court-martialled if you start probing the 'why' of a seemingly-stupid assignment.
If you have an actual reasonable person for a boss, the dialog that results from them answering your questions (and vice-versa) not only helps paint a more complete and accurate picture of their problem for you, but also forces them to think more deeply (and creatively) about the problem. What can result at the very least is a better definition of the problem. At best, a better solution can be worked out.
But, the more subtle long-term benefit to the dialog is to build a level of mutual trust with that person. You've shown a genuine interest in helping the business by being interested in that aspect of a technical strategy. Likewise, they understand more fully the technical (and ultimately strategic and financial) aspects of the direction they want to pursure. They become more clueful, and so do you.
What you really have is that you and he/she are both teacher and student at the same time. When this works well, it's magic: you become allies in dodging and reforming ill-informed decisionmaking, becoming reliable resources for each other. Everybody wins: you, the manager, and the business as a whole.
Come to think of it, most important communication in business is by its nature, teaching. But, like any good teacher, prepare to be a good student too: listen, take notes, and don't slack off on the homework.
There's actually a good deal more than 3 in the Bay Area. To wit:
Bay Area:
12 KNTV NBC San Jose Loma Prieta (HD) 19 KBWB WB SF Sutro Tower 24 KGO ABC SF Sutro Tower (HD) 27 KTSF - - SF Mt. San Bruno 29 KPIX CBS SF Sutro Tower (HD) 30 KQED PBS SF Sutro Tower (HD) 39 KCNS - - SF Sutro Tower 45 KBHK UPN SF Sutro Tower 47 KTLN - - Novato Burdell Mtn. 49 KSTS Tele. San Jose Mt. Allison 52 KICU - - San Jose Monument Pk. 56 KTVU FOX Oakland Sutro Tower (SD-16:9) 57 KRON - - SF Sutro Tower
Sacramento:
25 KOVR CBS Sacto Walnut Grove (HD) 35 KCRA NBC Sacto Walnut Grove (HD) 55 KTXL FOX Sacto Walnut Grove (SD-16:9) 61 KXTV ABC Sacto Walnut Grove (HD)
Basically, all the majors and a fair swath of independents are at least on DTV, and all the majors offer at least 1 HD subchannel.
The reason you don't see over-the-air HD in retail stores is the same reason you don't tend to see ANALOG there, either: stores (especially big ones) tend to use house feeds, which better-serves their merchandising. Even if they don't use a house feed for HD, chances are they'll use something off satellite like Discovery HD Theater which looks great (it's all-HD, not upconvert) and doesn't have any swear words;-)
There is also the attendant hassle of setting up a good over-the-air antenna in-store, especially in fringe or heavy-multipath areas (the latter showing up a weakness of 8-VSB). As it is, antenna installation and setup - even for analog - is practically a lost art these days. It's pretty sad when Radio Shack is about the only place that carries this stuff anymore.
I actually own a Corbin product (a motorcycle seat) but I wouldn't want to own a Sparrow (all-electric) or a Merlin (gasoline-powered.)
Every Sparrow made (about 250 of them) has had to be recalled numerous times to fix problems. And, the Sparrow has a nasty tendency to tip over (the jury's still out on the Sparrow II and Merlin.)
As far as inherent safety, they are a simple fiberglass shell on a frame which together offer practically no protection to the occupant. Their collision-avoidance ability is suspect at best, and they are so low-slung and small that they easily get lost in blind spots. A bad combination of factors if you ask me.
At least on my bike I have more options in avoiding collisions or injury. I have better visibility because I sit higher (practically eye-to-eye with pickups and minivans) and present a larger visual target.
Even if you ignore the Sparrow's quality problems and its inherent safety issues, the hassle of recharging an all-electric Sparrow, its limited range and its overall inefficiency adds up to a package that basically is a non-starter.
A hybrid is a better investment - at least it's a real car with actual passenger protection. So is a motorcycle, which is arguably 'greener' (consider the overall impact of electricity generation) and and space-efficient than a Sparrow.
All-electric cars are a boondoggle since they only serve to translocate the energy problem, and they are not as efficient as hybrids. In fact, all-electrics do worse than nothing: the additional generating capacity needed to serve them as a practical matter has to come from fossil fuel (and some of that the dirtiest kind: coal.) It is a very good thing this path is being shelved.
In contrast, fuel cell technology is not only more efficent than batteries, internal combustion or hybrid, but it offers more flexibility in source feedstock. Yes, you can use electricity to make H2 (at low efficiency) but that's not the most practical or promising path.
First of all, as it is today most hydrogen isn't even made using electricity. The most common process uses a high-temperature steam-and-catalyst reforming process to crack methane or other hydrocarbons into hydrogen and (unfortunately) carbon dioxide. The energy for the process comes from burning the carbon monoxide that is an intermediate reaction product (basically redirecting some of the methane's potential energy) with a subsequent reduction in yield. More recently, there is discussion of using this process on other longer-chain biofuels, such as sink-trap grease (PDF link - sorry.)
Even more promising is a new process that uses algae that requires neither fossil fuels nor generates CO2 emissions. As a photosynthetic process, it could actually tend to remove CO2 from the air instead of produce it.
Moving to hydrogen-based distribution is good if you accept global warming as fact. Moving to renewable resources is even better, at least for the reason that as it decentralizes control of energy resources and the economic power that comes with it. These are goals worth striving for, even if they seem far-fetched now. Just the fact that there is a serious effort and visible progress is of great strategic importance.
Slashdot Mirror
this one might suit the purpose...
Or not.
See this slideset, in particular the one titled "DOS File System" (slide 16.)
For a 256MB flash, it'd be FAT16. FAT16 only limits the root directory size to 512 entries for a disk type media (from here.)
Elsewhere it was said that they (NASA's engineers) weren't able to reproduce the problem exactly. I'm still very curious about this... as I said elsehere I've used DOS FAT16 under an RTOS (Thread-X) with lots of files and didn't see this kind of thing happen. Then again we hammered hell out of the FS and even fixed some problems with it.
See this slideset, in particular the one titled "Memory Allocation and Usage" (slide 18 I believe.) They do all their allocation at init time.
So far I see lots of flamage about how VxWorks malloc() sucks, but no concrete explantion why (if?) the number of files caused the crash. Also, there isn't any indication (so far) that the FAT was hosed.
So... what was the real problem? Was it indeed a malloc() failure that led to the reset loop? If so, what was being malloc()'d?
I'm really curious about this, since a project I worked on uses FAT16 on flash; in my case it was a multimedia terminal running Thread-X in 32MB RAM and 64MB Flash, with third-party DOS FAT and TCP/IP add-ons. We had lots of files on this thing, and it didn't seem to flake out like this.
This set of slides hints that they used DOS FAT. Since the Flash is 256 MB, that would imply FAT16.
Still doesn't completely explain the running-out-of-RAM problem, though. Kinda nasty that it reboots - it should degrade more gracefully than that.
By the way, I doubt that they swap to Flash or use VM at all due to the non-deterministic nature of VM (and the fact that it wears out the Flash.)
... do they use on the rovers? Anyone know?
There is work to do just that: Java on top of RT/Linux in the Rocky 7 platform (basically the testbed for MER1/2). See here (PDF file.)
Now the question I have is... which filesystem did they use for MER1/2? Is it DOS FAT? If so, I could see how the "too many files" problem could happen quite easily.
Not that DOS FAT wouldn't be OK for spaceflight, it is very simple, reasonably robust and quite mature. Just gotta watch for those FAT limits.
I didn't list that option because it seemed a roller+film mechanism would be mechanically fragile and heavy given the size it'd need to be for the entire panel set. It'd also cut the panel efficiency somewhat, partially defeating its purpose. Finally, you'd eventually run out of film.
FWIW, there's a set of Smith Roll-Offs in my gear bag somewhere... so I am familiar with the concept.
- dvd_tude
Sheesh - the JPL guys ought to talk to some desert racers sometime. They deal with more dust in the course of a Baja 500 than most of us do in an entire lifetime.
Anyway, some simple low-risk ideas:
* A small air blower jet to blow the dust off. After all, there is an atmosphere (albeit a thin Martian one) to work with.
* Tilt the panels and give them a gentle shake to get most of the dust off.
* Use an electrostatic coating to keep the fine dust from sticking.
As far as the batteries, couldn't they be deep-cycled to reduce their memory?
- dvd_tude
Motorcycles uses chains because they're (1) inexpensive, (2) lightweight, (3) lowest unsprung weight, (4) compact and narrow, (5) immune to the 'shaft jacking' (rear end rises and falls due to torque on the rear hub), (5) able to handle lots of power (ZX-12 anyone?), (6) directly compatable with transverse engine layout and finally (7) they're highly efficient when properly cared for.
Some motorcycles use shafts because (1) they're somewhat lower maintenance than chains. That's the sole reason. Yes, there have been some performance improvements in shafties (Paralever for example) but they still find use primarily on models targeted to riders who aren't so concerned about weight and cost and put a premium on maintenance (yes, that'd be you propeller-heads and Gold Wingers.)
And even then I'd argue the maintenance point: an O-ring chain sprayed every 600-1000 miles with a good lube like Maxima Chain Wax will give trouble-free service for a long time. My literbike's chain and sprockets lasted 30,000 miles with this sort of care. Cost of a new chain and sprocket set? About $185, and a half hour to change them out.
Oh yeah, there's belt drive too. S'ok if you don't mind the extra half-inch or so it adds to the width of the drivetrain, not to mention the limited power that realistic belt widths can handle. Also, don't get gravel in one, it'll mess up the sprockets if not damage the belt (I haven't priced belts and sprockets but I bet they're a lot more than a chain set.) However, they don't tend to stretch and need only a little shot of silicone once in a while.
- dvd_tude
Erm, I dunno about that.
Seems to me that "FLOP" by itself stands for "Floating Point OPeration". Several generations of computer people understand "FLOP" this way, and a MFLOP was a simply million of those (would you like fries with that?), that is, a quantity, not a rate. In the early-mid 80's we spoke of how many FLOPS it would take to do a graphics task, such as a homogeneous 3-d coordinate transform.
I know that Webopedia and suchlike have MFLOP as "mega floating-point operations per second", but that just doesn't make sense: where's the "S" in MFLOP to mean "second"? It ain't there. Hmm....
What's going on here? Maybe there's been a tendency in the casual press to try to coerce "MFLOP(s)" to be a rate like "MIPS" for symmetry's sake. That seems to be where Webopedia went with it. But, it just doesn't work right (that dang "S" is missing yet they tease 'second' out of "MFLOP".) From where I sit, a good and useful close-to-the-machine acronym - FLOP - got sullied by sloppy copywriters and marketing types into a fluff term like MIPS (apologies in advance to Tom, Ray, and Craig - you know who you are.) Tsk, tsk...
So I think the submitter took the high ground and went with decades of historical usage of the acronym FLOP. In that context it was entirely correct to say "TFLOPs/sec"
Suddenly I feel old... recalling when 20 MFLOPs/s was enough to take a company public...
- dvd_tude
Before you get any wild ideas, I meant to title it "17 percent savings could be 'most' of the kWh. derr...
- dvd_tude
You're forgetting that there are fixed 'service' costs on the bill irrespective of the kWh used. Given Japan's high labor and infrastructure costs these are probably steep.
Also, their kWh usage was probably low in the first place. This is a reasonable assumption given their small home (only 1100 sq. ft.), smaller appliances, and the comparatively frugal (energy-conscious) Japanese lifestyle.
They say they saved 17% on their electric bill and that 'most' (minimum > 50%) of their kWh came from solar. Since solar only affects the kWh used, you could have a case where at most 34% of their (old) bill was kWh and the rest - at least 66% - was fixed 'overhead' service fees.
Anyone from Japan care to comment?
- dvd_tude
Think back to when you started programming - what you wanted most of all was to get the computer to do something - anything - under your control. Once you gotten that far, you're hooked....
There's much wisdom in making 'hello, world' a first goal for novice programmers. By the time they've done this they've run through the tool chain at least once. So, make the tool chain to get there as easy and straightforward as possible to build confidence.
If it's a visual suite, that fine but be prepared to train your young coder on it. Personally I think a simple text editor and a shell (with stdin/stdout/stderr) is easier to master at first. It sidesteps the Byzantine GUI 'window dressing' trap and focuses the young mind on the logic of the program itself.
I also have a strong bias toward C/C++ as a starter language because of its compactness and that it can allow a young coder close to the bare metal if that's their proclivity. Some people make a case for Java; mastering its initial environment (classpaths and such) could be daunting, and in many ways it's too restrictive.
Finally, I think providing a dedicated Linux box for your hacker-in-utero helps better foster programming creativity: the tools, languages and sources are all there to be seen, used, and learnt from. Second choice would be OS X, except the hardware is expensive in these days of $100 EPIA boards.
- dvd_tude
... is what they hammer on in the courses, for good reason: simply put, you can die or be a cripple for life if you screw up. (My course? NAUI in 1976, taught by a Coast Guardsman at Alameda Naval Air Station. No wussy computers then.)
The main benefit of computers is that they simplify the task of figuring safe repetitive dives. In theory they can also give a more accurate plan based on accumulated nitrogen load, and thus allow more bottom time than a strict tables dive. Some computers even include a planning function which provides a simulation of a dive profile, based on its calculated estimate of your N2 load from prior recent dives.
But... the computers can fail. Think about it: you're relying on a relatively low-cost piece of electronic hardware, immersed in saltwater under pressure. There's lots of direct and indirect failure modes... not only could its seal or battery fail, it could be knocked loose, bitten by something etc.
And don't forget user error: if it's like most gizmos you could have forgotten to push the button to start it. Or, worse, you pushed it at depth and so the reading is too shallow.
God forbid it should have a software error like the early Aladdin evidently did.
To protect yourself, prior to your dive you should already have figured out your basic safe dive profile and bottom time given your nitrogen load. Manual tools like the PADI Wheel are used to do this. Then, you have: a watch, a backup gauge, and a buddy (who functions as a redundant set of gear.)
Once you're in the water your plan, your mechanical depth gauge, your watch and your buddy are your friends: they will save your ass (or at least your dive) if that fancy gizmo on your wrist decides to go tits-up at depth.
Yes, I have and use a computer. But I still do a 'sanity check' of the proposed dive using a manual tool like the PADI Wheel or the tables. Actually, if you're like most recreational 'vacation' divers who dive with divemasters, the divemaster does that check for you in case YOU screw up.
More rant: these newer 'air-integrated' computers scare me even more. I won't use one. I recently saw one fail, and the guy fixed the pressure transponder using a spring from a ball-point pen (bragging about 'MacGyvering' the thing) and then DIVED ON IT! Nuts!
My take on the guys who got bent on the early Aladdin is that they were pushing it, relying solely on the computer and doing 'closely-spaced' (repetitive) dives. Had they used manual backup and a lick of commonsense for time-to-fly they would not have gotten the bends.
As pros they should have known better. Yet to my mind they're not much smarter than those Moskito lobster divers I saw marking time in Roatan's hyperbaric chamber at Anthony's Key. Rather than placing faith in an etherial deity like la Moskita, they placed their unquestioning trust in a piece of software. The result is the same: they came up bent.
(Read more about the Moskito epidemic of decompression sickness here.)
- dvd_tude
The Winbond chip is pretty specific to its application, that is, making motherboard clocks. There are much better serial programmable devices that can provide a wider range of frequencies. You can get Cypress ones at Digi-Key)
Also for more accuracy, you can stack them and refactor P and Q over multiple dividers. On one project (an MPEG encoder) I did just that to make a low-jitter fully-locked 16.9344 / 12.288 / 18.432 audio reference from 27 MHz video. Each PLL was less than $2, and I used an 8051 to control it.
There are also specialty PLL chips used for cellphones that provide good accuracy using some voodoo in their dividers.
- dvd_tude
Stained glass windows using the lead-and-solder method (lead came) are very bendy and whippy; most of their strength comes from the frame. Large panels generally need to be braced. Also, lead came panels sag over time, especially if the cement (grouting) comes loose.
Lead came construction works for flat windows. As far as making shapes, forget it. It's way too weak.
Ever lift a Tiffany lamp shade? They're pretty hefty, yet, they retain their shape over time and they're not bendy. Tiffany glass (foil-and-solder) is much stronger and resistant to bending, because:
o the foil has adhesive backing to hold the glass
o the finished joint has no gaps (therefore no cementing needed)
o the solder has alloy metals to make it stiffer (e.g., tin) so it doesn't sag.
o the finished piece is much lighter than lead came
Tiffany style construction is a bit more labor-intensive (mostly you need to dress the cut glass edges) but the results are worth it: you have much more freedom in the design, be it a window, lamp, or computer case.
See more about the Tiffany method here.
- dvd_tude
[Disclaimer: I made lamps and terrariums in the 70's, so sue me.]
In short, treat the whole project like a Tiffany lamp.
This case appears to be built that way, although for Tiffany glass it's fugly. Plus he mounts a fan to a glass panel. Bad idea.
If you work stained glass I don't need to explain the advantages of Tiffany style construction. But for the other folks, it's stronger and prettier than 'lead came' ("H" channel lead) construction commonly used in windows.
So, for the glass I would definitely go that way.
As far as mounting the heavy bits, like a lamp I would use a fixture (frame) and keep the glass non-load bearing.
In keeping with the Tiffany lamp theme, use a pretty material for the frame. I'd look into using brass rod and sheet to make a 'perimeter' frame. Use brass lugs to join the pieces. I say this because brass and bronze are common Tiffany lamp fixture materials so it wouldn't look out of place with the stained glass, especially if you applied a patina finish to the metal.
Continuing with this basic frame, you then have a choice of making a glass 'shade' that slips over the frame, individual glass panels that come off, or perhaps a combination of the two. You could get fancy and make hinges for access panels out of tubing and rod.
As far as EMI compliance, forget it; the glass is little better than air for magnetic shielding.
Making a good ground plane under the mobo and to the rear panel is important to prevent ground loops and reduce static discharge, so splurge on the brass sheet there.
Cooling shoudln't be a huge problem so long as you position the fans close to where the airflow is needed.
-dvd_tude
... and a literbike will run in the 10's, bone stock. Heck, some are even scratching the top part of the 9's.
Hey I like using a Sawzall as much as the next guy, but what a waste.
See here and weep...
squidliness is next to codliness...
-dvd_tude
I might as well take a crack at this too (and I got a "C" in high school chemistry because I insisted on lighting up th acetylene ballons we used in the Gas Density lab...)
To focus on water's phase changes as a heat sink is to miss the point. Surface water is not a major component of Earth's mass, so its ability to smooth out temperature variations as a 'thermal capacitor' would on the face of it be not very significant when compared to the mass of Earth as a whole.
Another fallacy is that the Earth is a closed system. It's not: Earth not only takes in energy from the Sun but it also radiates energy back into space. I should mention that solar radiation isn't the only heat source either: radioactive decay also heats the Earth's core.
Once you understand this, then you appreciate the very reason greenhouse gases are called such: they trap this radiation like the walls of a greenhouse.
The size of the polar ice caps is a reflection of the average Earth temperature, not the other way around. Its effect on mean sea level is an indicator of mean temperature, but it certainly isn't the only one (how about the mean temperature itself?)
More interesting to me is the potential effect of increased water vapor in the air, itself a greenhouse gas. Would it have a positive reinforcing effect on a temperature rise, whether it be due to increased CO2, solar radiation, or cow flatuence? On first glance this phenomenon seems to be not well understood.
Another idea I don't see tossed about much is biological equilibria having influence on CO2 levels. It seems quite reasonable that increased CO2 levels and temperatures would increase the biomass (especially in plankton) and thus the drawdown of CO2 out of the atmosphere.
Global climate is a complicated thing. Unfortuantely it is complicated further by global politics....
- dvd_tude
heh...
....
I see you're an Accurate Reloading fan. If so, how could you overlook the Ben and the 577 T.Rex video, widely reposted under another, shall we say non-PC title
They also have a recoil comparison chart here
- dvd_tude
... turns out to be just that. Taking a small, inexpensive motor and attaching it to a bicycle? It's a basically sound idea. That is after all how Soichiro Honda got started in the motorcycle business after the war. Putting the engine in the rim? Hey, another nifty idea, but not original: the Megola did that in 1922.
Okay, sometimes synthesizing old ideas into new ones can yield interesting results. At least it makes you think "hmmm, neat hack". I don't see this as a fine example of this. All that gearing and mechanical complexity of the clutches and gearset strike me as expensive and failure-prone: too many moving parts for what it sets out to do. It would seem simpler is better.
Then there's the question of intrinsic value. Mopeds and motorcycles are cheap for the utility and performance they provide when you compare them to today's overpriced yuppie-toy bicycles. You can find a used motorcycle for $1200 or so, and mopeds even less than that. Mileage? Well, my CBR900RR gets 55mpg, and I don't have to worry about looking good in Spandex. Pollution? Many have pointed out that even 2-stroke motors can be engineered to be quite 'green', such as Aprilia's direct-injected 2-stroke scooter using technology licensed from Orbital Engine Corp.
So, tell me why this is useful?
A small confession: I hate having my time wasted. My personal work ethic and my own desire to have some control over my fate makes me really sensitive to the "GFAR" type of command. I think most bright people feel the same way, unless they're so worn down that they merely accept the status quo and just roll with it (a very dangerous apporach, career-wise.) Intelligence, experience, knowledge and initiative are good things. The question is, how can you use these to your advantage in this type of losing scenario?
I think the number one thing you can do is ask questions. You're not in the military, you're not going to be court-martialled if you start probing the 'why' of a seemingly-stupid assignment.
If you have an actual reasonable person for a boss, the dialog that results from them answering your questions (and vice-versa) not only helps paint a more complete and accurate picture of their problem for you, but also forces them to think more deeply (and creatively) about the problem. What can result at the very least is a better definition of the problem. At best, a better solution can be worked out.
But, the more subtle long-term benefit to the dialog is to build a level of mutual trust with that person. You've shown a genuine interest in helping the business by being interested in that aspect of a technical strategy. Likewise, they understand more fully the technical (and ultimately strategic and financial) aspects of the direction they want to pursure. They become more clueful, and so do you.
What you really have is that you and he/she are both teacher and student at the same time. When this works well, it's magic: you become allies in dodging and reforming ill-informed decisionmaking, becoming reliable resources for each other. Everybody wins: you, the manager, and the business as a whole.
Come to think of it, most important communication in business is by its nature, teaching. But, like any good teacher, prepare to be a good student too: listen, take notes, and don't slack off on the homework.
- dvd_tude
There's actually a good deal more than 3 in the Bay Area. To wit:
;-)
Bay Area:
12 KNTV NBC San Jose Loma Prieta (HD)
19 KBWB WB SF Sutro Tower
24 KGO ABC SF Sutro Tower (HD)
27 KTSF - - SF Mt. San Bruno
29 KPIX CBS SF Sutro Tower (HD)
30 KQED PBS SF Sutro Tower (HD)
39 KCNS - - SF Sutro Tower
45 KBHK UPN SF Sutro Tower
47 KTLN - - Novato Burdell Mtn.
49 KSTS Tele. San Jose Mt. Allison
52 KICU - - San Jose Monument Pk.
56 KTVU FOX Oakland Sutro Tower (SD-16:9)
57 KRON - - SF Sutro Tower
Sacramento:
25 KOVR CBS Sacto Walnut Grove (HD)
35 KCRA NBC Sacto Walnut Grove (HD)
55 KTXL FOX Sacto Walnut Grove (SD-16:9)
61 KXTV ABC Sacto Walnut Grove (HD)
Basically, all the majors and a fair swath of independents are at least on DTV, and all the majors offer at least 1 HD subchannel.
The reason you don't see over-the-air HD in retail stores is the same reason you don't tend to see ANALOG there, either: stores (especially big ones) tend to use house feeds, which better-serves their merchandising. Even if they don't use a house feed for HD, chances are they'll use something off satellite like Discovery HD Theater which looks great (it's all-HD, not upconvert) and doesn't have any swear words
There is also the attendant hassle of setting up a good over-the-air antenna in-store, especially in fringe or heavy-multipath areas (the latter showing up a weakness of 8-VSB). As it is, antenna installation and setup - even for analog - is practically a lost art these days. It's pretty sad when Radio Shack is about the only place that carries this stuff anymore.
- dvd_tude
I actually own a Corbin product (a motorcycle seat) but I wouldn't want to own a Sparrow (all-electric) or a Merlin (gasoline-powered.)
Every Sparrow made (about 250 of them) has had to be recalled numerous times to fix problems. And, the Sparrow has a nasty tendency to tip over (the jury's still out on the Sparrow II and Merlin.)
As far as inherent safety, they are a simple fiberglass shell on a frame which together offer practically no protection to the occupant. Their collision-avoidance ability is suspect at best, and they are so low-slung and small that they easily get lost in blind spots. A bad combination of factors if you ask me.
At least on my bike I have more options in avoiding collisions or injury. I have better visibility because I sit higher (practically eye-to-eye with pickups and minivans) and present a larger visual target.
Even if you ignore the Sparrow's quality problems and its inherent safety issues, the hassle of recharging an all-electric Sparrow, its limited range and its overall inefficiency adds up to a package that basically is a non-starter.
A hybrid is a better investment - at least it's a real car with actual passenger protection. So is a motorcycle, which is arguably 'greener' (consider the overall impact of electricity generation) and and space-efficient than a Sparrow.
- dvd_tude
All-electric cars are a boondoggle since they only serve to translocate the energy problem, and they are not as efficient as hybrids. In fact, all-electrics do worse than nothing: the additional generating capacity needed to serve them as a practical matter has to come from fossil fuel (and some of that the dirtiest kind: coal.) It is a very good thing this path is being shelved.
In contrast, fuel cell technology is not only more efficent than batteries, internal combustion or hybrid, but it offers more flexibility in source feedstock. Yes, you can use electricity to make H2 (at low efficiency) but that's not the most practical or promising path.
First of all, as it is today most hydrogen isn't even made using electricity. The most common process uses a high-temperature steam-and-catalyst reforming process to crack methane or other hydrocarbons into hydrogen and (unfortunately) carbon dioxide. The energy for the process comes from burning the carbon monoxide that is an intermediate reaction product (basically redirecting some of the methane's potential energy) with a subsequent reduction in yield. More recently, there is discussion of using this process on other longer-chain biofuels, such as sink-trap grease (PDF link - sorry.)
Even more promising is a new process that uses algae that requires neither fossil fuels nor generates CO2 emissions. As a photosynthetic process, it could actually tend to remove CO2 from the air instead of produce it.
Moving to hydrogen-based distribution is good if you accept global warming as fact. Moving to renewable resources is even better, at least for the reason that as it decentralizes control of energy resources and the economic power that comes with it. These are goals worth striving for, even if they seem far-fetched now. Just the fact that there is a serious effort and visible progress is of great strategic importance.
- dvd_tude