Where I work, we've got a few barcodes taped to the counter. Thermal-printed, been there for years: the paper is turning brown, and the black is a somewhat-vague purple.
We scan these fairly frequently on a daily basis, without problem.
The USPS seems to be happy with uneven dot-matrix printed barcodes; look at the lower-right corner of the stuff that drops through your mailslot sometime. And this is for so-fast-it's-blurry mail-sort systems, on particularly-lumpy material.
I've noticed 2-dimensional UPS barcodes (the funky ~1" square you see on some shipping labels, with a circular target in thing in the middle) printed dot-matrix, too.
And I've seen no indication that either system is in any way flawed.
So. We've established that the scanners aren't very particular; let's talk about printers.
Laser printing isn't so hot. Bend it a feww times, and the toner begins flaking off.
Lexmark, and probably others, offer what they claim to be waterproof ink. This is probably at least as durable as laserprint in a typical wallet.
The Alps MD-1000 I have here prints using wax ribbons. It tends not to flake, it tends not to fade, and it's definately waterproof. Oh, and it was cheap.
Most laser printers top out at 1200dpi. 2400dpi inkjets are now commonplace.
UPC barcodes have only two line widths - features which, given the scalability of barcodes, are probably quite easily implemented with a 24-bit printer at reasonable size.
Coca-Cola uses very large, sprayed dot-matrix barcodes on their 24-can cases of 6-packs. They're very rough, and I imagine they work justfine.
Now that we've got printing out of the way, let's talk about the barcodes I carry in my wallet:
I've got an Ohio driver's license, dye-sub printed plus holographic lamination, made 2.5 years ago. The barcode is quite plain and obviously usable, as sharp as I remember it being when it was issued.
I've got a Blockbuster membership card. 24-pin dot-matrix printed, issued at least 5 years ago, and laminated: The barcoode is quite plain, and obviously usable.
I've got a Sam's Club membership, issued a few years ago, printing style unknown (but probably thermal). The barcode is wearing off, but is still quite usable.
Obviously, you don't want to take a crucial water-soluable barcode out in the rain and use it. However, I feel that you need to look around a bit more: There's a plethora of low-res, functional barcodes attached to items in the world around you which you are obviously oblivious to, many of which are expected to be exposed to the elements.
And remember: Anything can be laminated, usually at a shop within walking distance. Why might one expect to be able to print barcoded keys at home, while conventional machined brass keys require a trip downtown? One shouldn't, at this point: Let's take it one step at a time, starting with email delivery.
Oh. And JPEG, as a format, is fine. It can encode sharp lines with ease, as long as the encoder is aware of the requirements and/or the quality settings are set sanely (which is not a problem with standard libjpeg) -- efficiency, in this instance, is rather not relevent. PNG, as a purist ideal, would be somewhat better. But even monochromatic BMP (or XBM or PBM...or PCX for old-school PC users)-format barcodes would be quite sufficient for the task at hand. Not to mention GIF, which will be readable by everything for a really.long.time. You could probably even distribute barcodes as HTML tables with colored backgrounds without problems.
Thus, I find all of your presented points to be misleading, inaccurate FUD.
I used to have a large rooftop antenna before I moved into this tin box/apartment, and had my own RadioTiVo.
It wasn't at all difficult, though I did spend way too much time optimizing the commandline for LAME and setting levels correctly. I just put a YMF724-based sound card into my headless, does-everything FreeBSD box, plugged it into a 1980s-vintage standalone Kenwood digital tuner (find something similar at a pawnshop or Ebay), and made some cron jobs to run things. The 724 was nice because its ADC stage generally sounded very good, and it had a loopback mode that it could be massaged into which would let you hear immediately if you had clipped the input.
The box, a K6-2 350, isn't quite fast enough to do VBR MP3 encoding in realtime, and I was dead-set on VBR. So, I had it record the entire program as standard 44.1KHz 16-bit PCM, and then run a nice'd encode process on the file after the radio program had finished.
Sometimes, usually on the weekends, this meant that 2 or 3 processes of LAME were running at a time trying to catch up. Not that FreeBSD ever broke a sweat...
It ran extremely reliably, and with an NTP-synced clock, the start- and stop-times were consistantly dead on.
Every few months, I'd burn a CD or two of Car Talk for archiving and nuke whatever was left over.
Of course, there was no way to change stations. I considered briefly the notion of building a machine from mindstorms that would push the radio's preset buttons, but then I realized that nothing but NPR had any programming which I actually wanted to listen to.:) Using an FM tuner card was always out of the question for reasons of noise and interference.
Hint: Use lame's lowpass filter to cut everything above 15KHz. There's nothing there but noise with commercial FM broadcasts, which are already band-limited to 15KHz anyway per FCC rules. That said, resist the temptation to use a 32KHz sampling rate and stick with 44.1. It's what the Nyquist filters and samplerate converters in consumer gear are optimized to work with, and makes burning audio CDs easier. These translate to better sound, overall.
It was (mostly) my own work, quite some time ago when I was a cub scout.
I say "mostly" because I had some help from my dad: the design was the result of much groupthink. Most of the work, however, I did myself, including the shaping. Dad did the initial rip down the length of the car with a radial arm saw, did most of the handling of molten lead, and other potentially-hazardous things. I also had a friend of mine help paint the flames.
I consider it a perfect example of what a father-son project is supposed to be -- we both learned a lot, and had fun.
Immitation is the most sincere form of inadequacy
on
Pinewood Derby Tips?
·
· Score: 2
How odd that your dissertation looks just like the comment that I posted myself on these pages, an hour before you.
Strange, isn't it? I mean, just look: Compare this to this. The similarity is striking, isn't it?
Tell me, if you would, which of the following explanations is most applicable to this deceitful prose: A database error, some sort of maligned human oops, or just good old-fashioned plagiarism?
If it were a database error, the entire comment would have been duplicated verbatim, but it was not: the tags I used to italicize "nth degree" are absent, for example.
If it were some sort of human oops, I imagine that the such a hastily-construed cut-and-paste job would have included my sig, not to mention chunks of headers, leading and trailing stories, etc. In other words, by the time you'd reached a sufficient level of intoxication that your brain would permit you to accidentally duplicate a comment of such length, other mistakes would be apparent.
Therefore, I deduce plagiarism to be the cause.
Which leads me to be curious: How much code do you intend to steal for inclusion in your upcoming PDK product?
It's not too hard. As others have pointed out, center of gravity means nothing, except that a lower center would tend to have less cross-sectional area (which is good). All you have to worry about is friction and weight.
Friction is the hard part, and you've got two kinds to deal with: Friction with the air, and friction between the wheels and their axles.
For wind resistance, just start slicing the car down as much as you feel you can without sacrificing strength, or capacity for added weight. Round off both ends of the block into something that looks reasonably aerodynamic - I used a bench-mounted belt sander. You want it low, and flat, so as to displace as little air as possible as it moves down the track. Using similar curves to a high-speed train will yield better results than mimicking an F14 or Lamborghini. That the resultant form is rather boring is not an accident.
Friction against paint is probably not a big deal. My finish was a hand-brushed yellow-and-black flame paint job, and was nowhere near smooth. But it certainly wouldn't hurt to make it as mirror-like as possible with careful painting and sanding, along with a coat of well-buffed wax.
One trick I used which I've never seen repeated: A small circle of electrical tape, placed over the outside of each wheel, to further reduce drag. They were cut with a ball peen hammer and a coin of appropriate diameter, and fit precisely. It should be flat or slightly convex, but not concave, when applied to the wheel. This will increase rotational inertia (==bad) somewhat, but it seems like a beneficial thing to do.
To reduce friction between wheel and axle, I first removed the burr on the supplied axles using progressively fine sandpaper - there were two axial burrs, along with one across the back of the head. I also removed burrs on the inside of the wheels, and slightly rounded the flat part near the middle, inside of the wheels where it can touch the body, to reduce contact area.
Use graphite on the wheels and bearings, and use it as much as a polish as you do a lubricant. Work it into the microscopic texture of the axles, the inside of the wheels, and the bearing surfaces between the axle and the body by whatever means you can find, and then add some more and roll the car around a bit.
I used feeler stock (thinner-than-hair bits of flat metal) to get each wheel to within a thousandth-or-so of an inch of being on the same plane, such that they'd all be in contact with the track surface. I don't know if this helped, but it seemed like a good idea.
At least, try to keep your axles as straight and square as possible with eachother and the body. You do not want any pressure on the points where the wheel contacts the body, or the head of the axle.
And adjust the axles so that there is as little play in them as possible. The wheel should not move appreciably side-to-side, nor should it forcably rub on any point. Slop here translates to lost energy. It may take several iterations of gentle tapping on the axle and even-more-gentle pulling on the wheel to get this right.
Weighting is a science. You want the car to be as heavy as it possibly can be. Remember that you won't be immediately disqualified for having a very slightly heavy car, and that you'll get at least a couple of chances to bring the weight down.
Drill holes in the bottom of the car, and pour lead into them. I used an antique balance I had at home and added weight until I measured it being just -over- specification, and so the lead was protruding slightly below the bottom of the car. This allowed me to use the official scales to tune the car at the event, removing a bit of lead with a file and re-weighing several times before nailing it precisely, while in the process filing the lead flat and reducing drag.
Place a strip or two of electrical tape over the lead, to further reduce drag and add the nth degree of tuning: If you file a bit of lead off and the car still measures.001g over, use a sharp knife and remove a sliver of tape to finish the tuning. You'd be amazed how much a bit of vinyl electrical tape can weigh...
If your track levels out toward the bottom, put the weight as far toward the back of the car as you can. You'll get a few more inches of "thrust" by doing this, vs. your front-weighted opponent.
The devil's in the details. When it's all done, handle the car as if it were full of nitroglycerine - don't even look at it funny, or all of the precarious work you've put into it will begin to undo itself. Never let others handle it, and absolutely never leave it unattended at the event. It's a finely-tuned, very delicate instrument, and is deserving of respect.
Tablets of this sort have been around for a long time -- the most popular made by Wacom, and the cheapest model being the $99 Graphire. I've seen these for sale even at Wal-Mart..
Not a bad kit, really - you get a pen and a scrollwheel mouse that you can use interchangably. Neither the pen nor the mouse require any electricity at all, instead relying on magnets and some trick inductive stuff inside the tablet. It's pressure-sensitive, too.
Wireless is good so that you can watch porn straight from your neighbor's PC from the comfort of your couch, without all the hassle of maxing both cable connections, stringing Cat5, or even notifying him...
Just buy a Plextor. I've been using their 8x PR-820 drive for almost four years to the day.
A few thousand burns later, I've got no trouble to report except with a bad batch of Verbatim media toward the end of 1999.
And, mind you, this is for all manner of material -- from PSX archiving to music production to bulk duplication, usually on the cheapest media I can find. I've never burned at less than maximum speed.
YMMV, HTH. But given this experience I'm not likely to ever buy anything other than Plextor in the future -- that is, if this drive ever dies so that I can justify replacing it.
I used to live in a 100-year-old house. It had modern, 200-amp electricity and indoor plumbing. Those must've been pretty serious installation projects compared to the little stuff I did with it.
It's had RG-58 installed for 10base-2, replaced by Cat5 for 10/100. The ISDN demarc was on the back of the house; the ISDN TA was at the front, upstairs, fed by a new run of Cat5. It's had cable TV, Primestar, and DirecTV. Along with a rooftop TV antenna and an FM antenna up in the attic. Telephones in every room without a toilet, with up to four live lines coming in at a given time.
It's just not that hard to do with old wooden houses. They've all got complete crawlspaces or basements, and are topped with capacious attics - you can get anywhere on either floor from one of these places. The walls frequently contain completely disused chimneys, which make fine vertical conduit. If you're worried about water and chimney cruft harming your precious Cat5, just buy some that's Teflon insulated (read: "plenum").
And since they've had plumbing and electricity added, all of these internal spaces are very likely to be easily accessible (unlike a modern buttoned-down house).
If you're lucky, the structure might even be balloon-framed, with studs made of -long- lumber stretching from the roof to the foundation without interruption.
And since the walls are either uninsulated or blown (unless it's been on fire in the past), you can shove wires down these stud cavities with a rod anywhere you want, without accumulating a snowball of fiberglass on the end of the rod.
If it has forced-air heat, it's usually really, really easy to get wires from Point A to point B by way of ductwork.
There's always a plethora of open passages left over at different times from the installation of indoor plumbing, centralized heating, knob-and-tube wiring, and then jacketed wiring.
You've got it easy, running your skinny little Cat5 around, compared to the efforts of those who've lived there before to install modern utilities.
Where have we heard what before? Are you talking about The Bubble?
Of course companies seek to profit.
As history dictates, the health of a company is not related to the long-term support of its products. Support gets killed in the interest of profitability as much as it does for the lack of it. It doesn't take much of a beancounter to realize these few things:
Products which are no longer produced generate zero revenue, but cost money to support. The longer you support old products, the longer your users will postpone buying your new products.
Thus, the more poorly you support your old products, the more likely people are to be in the market for something new. If all companies operated like this (they do), it'd be a goldmine (it is).
Since we're thusly assured to have zero support from the vendor in the not-distance future, I submit that the only way to ensure any future usefulness from the products you buy is to make an effort to get the same thing everyone else seems to have. Or at least something built from the same, or compatibile, component parts.
Think NE2000 (and clones), 3c509, DEC Tulip (and clones), HP LaserJet (and clones). Think Oxford 911 firewire/ATAPI bridges, NCR/Symbios/LSI/whoevertheyarenow 53c87x SCSI chipsets, and VESA BIOS.
If you've got the NE2000 or Tulip of 802.11a, you won't have any problems keeping your gear working with current software -- Microsoft and a dedicated team of rodent-like OSS hackers will keep it alive for as long as it remains useful, plus a few years.
I'm afraid I don't quite follow what you're saying.
If I have an unrestricted 10Mbps between myself and my friends (and their friends, and their freinds' friends, and...), you better believe that MP3 will floweth freely amongst the group, without any one person needing to spend a dime.
Of course, a number of these people will also subscribe to some form of consumer broadband. Not to mention those who rip their own CDs. There's just as many avenues for new material to enter the mix as there is for cross-pollenation of, say, Gnutella and Kazaa.
And speaking of broadband, I pay ~$50/mo for 2000/384kbps RoadRunner. I can't fathom sharing a paltry 1536kbps amongst 10 of my greediest peers (I like burning ISOs in realtime as they download), nor can I imagine that sharing of such services would be tolerated for very long.
Only one thing is needed for this to succeed: People willing to spend money in a small geographic area in order to talk amongst themselves. It's not gonna be easy, though.
Back in the day, it'd probably have been fairly easy to get people to drop a few hundred bucks on hardware in exchange for free local-area bandwidth.
You'd need only post a message to a local BBS suggesting the idea, debate it for a week or two, round up a couple of capable people willing to help with installations. After that, proceed with the leeching.
First, the sysops would connect to eachother, and by way of their ubiquitous, incessant bragging would encourage others to tie into the network.
If 802.11/a/b had been available back then, at today's prices, such networks would have been plentiful. But it's much too late for that.
The local scene is all-but-nonexistant these days. LUGs and 2600 notwithstanding, there's no way to communicate with a semi-clued cross section of the local populace anymore.
This social problem combined with the fact that there is no percieved advantage to talking to a computer in the same town vs. one is Sydney, Australia makes the prospects of participation look awfully dim.
So. Since there's no local forum by which to arrange such activities, and there's no compelling reason for people to join such a network instead of paying for DSL, good salesmanship is the only way to produce such a beast today.
People will first think you're crazy when you tell them that you want them to buy a few hundred dollars with of cabling, Yagis, and gear, only to talk for free with a half-dozen strange geeks around town. And then they'll question your motives, wondering what your take in it is.
It's a little early for the masses to get ready to participate in something of this scale and organization. In order for people to want to join such a network, a few things need to happen first:
1) Transfer caps, and/or the return of metered dialup access. People, as a rule, are afraid of buying new things, especially when what they have appears to work justfine. It doesn't matter if this seems irrational or not. People who drive Fords generally continue to drive Fords until their Crown Vic explodes one morning on the way to church.
2) Remember that killer app everyone was looking for during the dot.com boom? It remains elusive to this day. If it is ever discovered, and has sufficient local interest for people to justify dropping cash on hardware in order to participate, things might have a chance. And AFAICT, except in new markets, the broadband thing is getting pretty stale. The freenet needs a killer app.
P2P is close, but the potential for such unrefined protocols to spill over from the freenet onto paid internet links makes for some hefty financial baggage for someone to tote. You could charge for internet access, but what would be the point? It's supposed to be free, remember?
3) 802.11 gear needs to come down in price. With Wal-Mart selling new PCs for less than the cost of my first CD player, wireless gear must look quite expensive to Joe Average in comparison to their "free" cable modem.
4) Additionally, we need more spectrum in a band capable of traversing more than a few miles of open terrain, without hanging antennas several hundred feet in the air. Swapping files and playing Unreal with Joe Across Town would be fun, but is rather limited in scope. People in Toledo ought to be able to be able to at least send email to folks in Detroit without unreasonable delay. There will be a handful of savvy volunteers willing to maintain short-haul inter-city links, but only if there's popular demand and a low-cost technical means of putting them together -- after all, any way to avoid paying Ma Bell for T1 circuits is a Good Thing in this context.
5) I touched on this before, but we'll need better protocols, or at least good shims for existing ones. If such a network ever happens, it will be one where there are relatively vast amounts of relatively local bandwidth for free, with things slowing down considerably as distance inreases (along with increased costs that someone will have to cover).
A diverse network of transparent squid proxies will help markedly with http, but what about ftp? Cursedly indiscriminate P2P? Even IRC can consume substantial bandwidth with a few thousand consumers. All of these things can be cached/proxied/otherwise-massaged, rate-limited, or simply operated on a strictly-local basis. But, someone along the way will have to write, implement, and enforce these policies at the border. The investment in time required for such needed, hatred-inspiring rulemaking will not be small. (Read: operating gateways is going to be expensive, time-consuming, and quite thankless.)
6) And for the sanity-in-infrastructure requirements, it'll have to run IPV6 more-or-less exclusively -- nobody wants to be known as 10.4.249.57, and real IPV4 address space costs money. Besides, in a free network, there's no reason for every man, woman, child's palm pilot, bicycle, and toaster oven to not have unique addresses. Is the state of IPV6 under any incarnation of Windows sufficient to support this endeavor?
That all said, you could count me in if I didn't live in a downstairs apartment with windows facing outside of town.
Is the idea too early, or too late? You tell me. Personally, I don't see many of these problems being solved very soon in the current climate, but you be the judge.
I used to repair fire-damaged structures for a living, so I've seen this more than a few times.
The easiest, and often cheapest, thing to do is throw everything away and cash an insurance check for new stuff.
For the stuff you don't want to throw away (because it is sentimental or otherwise irreplaceable), try Formula 409, Simple Green, or another general-purpose detergent. Use a brush, a cloth, or whatever seems appropriate.
Electronics can have a tendancy to not survive such cleaning, however. There's (at least) a thousand types of glue used to hold the bits that comprise them together, and no telling how they'll react to chemical treatment.
Water, of course, is generally harmless, but has limited application on smoke residue.
A friend of mine's house burned a few years ago, and the fire department vented the ceiling of his computer room...which is to say that vast amounts of smoke and steam went flying past his gear. I found a number of CDs in that room with their jewel cases melted off of them.
Most of it is still working justfine today without any cleaning, though the CDs did require some special care to come back to life.
You'll also do well to hire a company who specializes in such cleaning projects, if you want to try salvaging stuff. I used to contract with Serv-Pro (they've got offices all over - check the phone book), and they were often able to restore things to new.
Plus, they had a fleet of athletic 18-24 girls to do the work, which always brightened up my workday.
Once you're ready to start painting and carpeting, make sure you coat everything with a good primer. I usually rented a fairly serious airless paint sprayer, and used Killz, or Pro-Block from Sherwin Williams. Killz does a somewhat better job, being shellac, but the alcohol base will kill you dead if you're not extremely careful with your respirator. Plus, there's probably flash hazards with spraying alcohol everywhere... Pro-Block works very nearly as well, is oil-based (thus possible to coexist with) and is a great deal less expensive.
Cover the walls, ceiling, any exposed studs, the floorboards, and everything else you want to never smell of smoke again.
A few months ago, I installed Slack 3.9 on a 12-meg, 386SL/25 laptop. 3.9 was the last to includes a 2.0 kernel and libc5, so I figured it'd be fastest.
It ran fairly well. No accelerated drivers for X, or anything else other than Windows 3.11, so it lived in text mode.
It ran surprisingly well. Of course, the 386SL is an odd beast, with on-chip memory and cache controllers -- this was doubtless the source of some improvement in speed over what I expected.
Some time later, something-or-other trounced the hard drive rather completely, so I installed Slackware 8.1.
It runs surprisingly well. Some things even seemed to be a touch faster than they were with 3.9.
Thus, I'd like to submit that -all- versions of Slackware are suitable for old hardware.
According to this chart, channel 6's FM audio resides at 87.75MHz.
College/non-profit radio station FM frequency allocations start at 88.1MHz.
Meteor scatter manifests as the ability to recieve distant, over-the-horizon broadcasts which are normally inaccessible, not as the ability to recieve local stations at a different frequency than perhaps you should.
What you are experiencing is just one FM broadcaster stomping on another's frequency, and the tuner being unable to sort them out. You've heard this before: When driving in the car trying listen to 103.7, but getting bleed from the adjacent stations at 103.5 and 103.9.
A Rotel tuner that I have will tune down to 87.5MHz. I used to get a kick out of listening to channel 6's audio with the stereo.
Airflow? He's mounting a single 1U box. And even if he wanted more, it is usually exceedingly difficult to find road cases in excess of 12U. Which just isn't very much equipment.
Remember, the intended purpose of these racks is not just for the shipping of equipment, but the use of it. Sensitive analog (and increasingly, digital) electronics upon which multi-million dollar shows (and hundreds of thousands of potentially-angry people) are relying on.
This type of stuff generally makes a lot more heat than a couple of PCs. And it's sitting in direct sunlight on a 105 degree day...that is, unless it's getting rained on.
Somehow, airflow seems to be OK for these folks. And they take it at least as seriously as any of you computer kids.
Is there any particular reason to be using copper?
I mean, sure: it's a fine conductor. But for soaking up RF in the context of a home computer, it would seem that steel would work fine. And rather than being ~$3/ft^2, steel is usually of negligible cost.
The best place to look for this sort of stuff is not at a business selling fabric to housewives or mail-order distributors, but your neighborhood hardware store. Not Lowe's or Home Depot, but the tired-looking place downtown that sells shotgun shells out of the same glass display case that props up their singular cash register.
In this store, you will find a man who looks as old as the building. He will be helpful. You'll either be directed toward the precise item that you're looking for, be able to order it for little cost, or be given information about another tired-looking local business that specializes in such things as wire mesh.
Myself, I've seen copper window screen sold in these places, by the foot. No idea if the holes are too big for the frequencies in question.
Failing that, there's other options. AFAIK, the requirements for a Faraday cage do not stipulate that the material must be hole-y.
From the same hardware merchant (or any of the fabric stores that you're so fond of), you'll be able to buy a can of spray-on adhesive. From your local food service vendor, you'll be able to buy a wide roll of aluminum foil.
You'll need one (1) aerosol can of adhesive, and one (1) roll of foil.
Apply adhesive according to directions on label. Apply foil according to common sense, bearing in mind that the adhesive will not let go of the foil, ever. If grounding connections must be made between movable portions of the case (side panels, for instance), make sure that the foil is applied in such a way that they're able to touch eachother with at least slight pressure.
Or, if you want to be a pedant, join the sections together with copper foil tape.
Done. Time saved: days. Money saved: At least $65.
[OT: What, pray tell, is Slashdot's average IQ these days? I deal with stupid, helpless people for a living, but I'm often amazed by what I discover on these pages.]
I think that your 20% figure is rather low, and greatly overestimates the level of knowledge posessed by today's typical internet user[1].
I therefore submit the following conjecture:
70% of bandwidth-equipped internet users are too stupid to do anything at all about sharing. 1/3 of these people are behind Linksys routers that they know nothing about, or are otherwise irrevocably firewalled.
This leaves about 46% of the unwashed masses sharing by default. That's about 32% of the whole Internet populace can do nothing but share, because they don't know any better.
And then there's the remaining 30%-big category of people who might be able to control whether or not they share with any surety.
A third of the people in this group are spiteful leeches and will never share. Another third favor equality, and will make an effort to give back what they take. And the remaining third are intelligent enough to realize that the more people share, the lesser the individual bandwidth penalty for sharing becomes. This latter group will spew forth as much data as needs dictate and their connection permits, thus cancelling the antisocial efforts of the spiteful portion.
Thus, on average, 62% of all internet users share.
This means that to sustain a stream, this 62% sharing majority of users will need to relay the program to an average of 1.6 people each. Some will do a bit more, some a bit less.
My RR cable modem wouldn't even notice if I started streaming at 50kbps (which sounds remarkably good with ogg) to 4 or 5 other people, let alone the paltry 1 or 2 listeners that such a system as this requires to stay healthy.
That all said, it'd sure be nice if multicast IP became a widespread reality. It'd put all of these issues to bed with much haste...
1: See "Jargon File," under section titled "The September that never ended."
I'd like to add that I'm quite happily running Slackware 8.1 with a 2.4 kernel on a 25MHZ 386SLC laptop. It won't win awards, but it does pull a (quite remarkable, really) 300KB/sec from its PCMCIA NE2K.
It also suffers from a laggy 2-gig Hitachi drive, which is probably not dissimilar from that which you have but is, I assure you, a good deal faster than the 150-meg JVC it replaced.
And, as you might surmise, I'm reasonably pleased with it. I don't ask much, it doesn't do much. It is rock-solid stable, though...
I only hope it might put your disheartening comments toward the forlonged 486 genre into a more proper perspective.
But that's just a toy, such as it is. On to more important things:
Karma: Big Swinging Dick (mostly affected by moderation done to your comments)
Now, wrap it into a (pre-existing?) HOWTO. Get it published on defcon1.org. Whatevevr it takes.
But don't leave it as a random, 1-sentence Score:3 posting on Slashdot, where it will do little good for future masses encountering the same, doubtless growing, problem.
Yep. Might as well use it for other things. That's why people buy SUVs instead of more practical vehicles.
From the sound of what you're now saying, you're more interested in building a general-purpose home server than an inexpensive, reliable replacement for your troubled Linksys box.
Your priorities seem to be misaligned toward serving this end. If speed is enough of a factor that you care about the CPU at all (honestly, a 486 can handle justabout any consumer routing scenario without breaking a sweat), you're not going to be happy with that slothly laptop drive.
You're probably also not going to be happy with a small, unexpandable system, as long as you've got other uses in mind. Better off going at least to MicroATX so that you get a few real slots, and aren't so tempted to put everything into a hideously small case.
That all said, I still feel that a single-purpose, solid-state mini-itx system would be the ideal replacement for the single-purpose, solid-state Linksys box: it follows KISS procedure. That a 400MHz CPU may be underutilized by merely routing your broadband connection does not mean that it must be burdened by serving other unrelated tasks.
You are after stability, right? You're sick of tinkering with the Linksys thing, trying to make it work correctly. So, build a replacement that works correctly, and resist the urge to tinker with it. Leave it the hell alone. Unless some security issue forces a kernel upgrade or somesuch, if it's not broke, don't fix it. Being solid state, "breaking" is rather unlikely to occur unless it is disturbed.
In other words, just put it in the corner of your basement and forget about it.
Good routers, in this context, are transparent and invisible, as all infrastructure should be. Treat it like you thought you'd be treating your Linksys router when you bought it, before discovering that it doesn't work well for you; it shouldn't require much more care and feeding than your water heater.
And if, for some reason, you still can't get over the urge to make maximal use of what you've built, simply run seti@home, cat/dev/random >/dev/null, or some other CPU-eating task. Rest well knowing that your inexpensive processor is always working as hard as possibly it can.
It's just a firewall. It doesn't need mass storage, or at least nothing more than few megs. It just needs to be reliable.
So. Just beg your friend for the throwaway 8- or 16-meg compactflash card that came with his camera, and plug it into one of these.
Less power (can we say "fanless PSU"?), more speed, and superb reliability. With proper research, the adapter should be in the same price range as the 2.5" IDE adapter kit that you'd need for a laptop drive...
Save the hard drive for things that can benefit from the space.
The DjVu folks seem to have created a very fine format for the storage of scanned documents of random layout. It uses space efficiently, and produces good-looking (though imperfect - think JPEG) results. It's the only way I know of to preserve the look of a paper document without throwing away vast amounts of storage.
But, as far as I can tell, that's where the fun stops. AFAIK, it doesn't handle vector graphics, and has nothing to offer over PDF for strictly digital documents. In the digital world, PDF produces perfect results, automatically.
OTOH, PDF is not geared toward scanned documents. I've seen a lot of examples of, say, scanned datasheets in PDF; all of them were bad.
I thus submit that each respective format has its own well-defined niche, and fills it admirably.
Ghostview certainly could support DjVu, if someone wrote the appropriate code for ghostscript. That it supports GIF, JPEG, PNG and PDF would tend to indicate that it's well within the scope of the software's design parameters.
Slashdot Mirror
Eh?
Barcoding isn't a very demanding exercise.
Where I work, we've got a few barcodes taped to the counter. Thermal-printed, been there for years: the paper is turning brown, and the black is a somewhat-vague purple.
We scan these fairly frequently on a daily basis, without problem.
The USPS seems to be happy with uneven dot-matrix printed barcodes; look at the lower-right corner of the stuff that drops through your mailslot sometime. And this is for so-fast-it's-blurry mail-sort systems, on particularly-lumpy material.
I've noticed 2-dimensional UPS barcodes (the funky ~1" square you see on some shipping labels, with a circular target in thing in the middle) printed dot-matrix, too.
And I've seen no indication that either system is in any way flawed.
So. We've established that the scanners aren't very particular; let's talk about printers.
Laser printing isn't so hot. Bend it a feww times, and the toner begins flaking off.
Lexmark, and probably others, offer what they claim to be waterproof ink. This is probably at least as durable as laserprint in a typical wallet.
The Alps MD-1000 I have here prints using wax ribbons. It tends not to flake, it tends not to fade, and it's definately waterproof. Oh, and it was cheap.
Most laser printers top out at 1200dpi. 2400dpi inkjets are now commonplace.
UPC barcodes have only two line widths - features which, given the scalability of barcodes, are probably quite easily implemented with a 24-bit printer at reasonable size.
Coca-Cola uses very large, sprayed dot-matrix barcodes on their 24-can cases of 6-packs. They're very rough, and I imagine they work justfine.
Now that we've got printing out of the way, let's talk about the barcodes I carry in my wallet:
I've got an Ohio driver's license, dye-sub printed plus holographic lamination, made 2.5 years ago. The barcode is quite plain and obviously usable, as sharp as I remember it being when it was issued.
I've got a Blockbuster membership card. 24-pin dot-matrix printed, issued at least 5 years ago, and laminated: The barcoode is quite plain, and obviously usable.
I've got a Sam's Club membership, issued a few years ago, printing style unknown (but probably thermal). The barcode is wearing off, but is still quite usable.
Obviously, you don't want to take a crucial water-soluable barcode out in the rain and use it. However, I feel that you need to look around a bit more: There's a plethora of low-res, functional barcodes attached to items in the world around you which you are obviously oblivious to, many of which are expected to be exposed to the elements.
And remember: Anything can be laminated, usually at a shop within walking distance. Why might one expect to be able to print barcoded keys at home, while conventional machined brass keys require a trip downtown? One shouldn't, at this point: Let's take it one step at a time, starting with email delivery.
Oh. And JPEG, as a format, is fine. It can encode sharp lines with ease, as long as the encoder is aware of the requirements and/or the quality settings are set sanely (which is not a problem with standard libjpeg) -- efficiency, in this instance, is rather not relevent. PNG, as a purist ideal, would be somewhat better. But even monochromatic BMP (or XBM or PBM...or PCX for old-school PC users)-format barcodes would be quite sufficient for the task at hand. Not to mention GIF, which will be readable by everything for a really.long.time. You could probably even distribute barcodes as HTML tables with colored backgrounds without problems.
Thus, I find all of your presented points to be misleading, inaccurate FUD.
Think now, post later. K?
I used to have a large rooftop antenna before I moved into this tin box/apartment, and had my own RadioTiVo.
:) Using an FM tuner card was always out of the question for reasons of noise and interference.
It wasn't at all difficult, though I did spend way too much time optimizing the commandline for LAME and setting levels correctly. I just put a YMF724-based sound card into my headless, does-everything FreeBSD box, plugged it into a 1980s-vintage standalone Kenwood digital tuner (find something similar at a pawnshop or Ebay), and made some cron jobs to run things. The 724 was nice because its ADC stage generally sounded very good, and it had a loopback mode that it could be massaged into which would let you hear immediately if you had clipped the input.
The box, a K6-2 350, isn't quite fast enough to do VBR MP3 encoding in realtime, and I was dead-set on VBR. So, I had it record the entire program as standard 44.1KHz 16-bit PCM, and then run a nice'd encode process on the file after the radio program had finished.
Sometimes, usually on the weekends, this meant that 2 or 3 processes of LAME were running at a time trying to catch up. Not that FreeBSD ever broke a sweat...
It ran extremely reliably, and with an NTP-synced clock, the start- and stop-times were consistantly dead on.
Every few months, I'd burn a CD or two of Car Talk for archiving and nuke whatever was left over.
Of course, there was no way to change stations. I considered briefly the notion of building a machine from mindstorms that would push the radio's preset buttons, but then I realized that nothing but NPR had any programming which I actually wanted to listen to.
Hint: Use lame's lowpass filter to cut everything above 15KHz. There's nothing there but noise with commercial FM broadcasts, which are already band-limited to 15KHz anyway per FCC rules. That said, resist the temptation to use a 32KHz sampling rate and stick with 44.1. It's what the Nyquist filters and samplerate converters in consumer gear are optimized to work with, and makes burning audio CDs easier. These translate to better sound, overall.
Good luck.
It was (mostly) my own work, quite some time ago when I was a cub scout.
I say "mostly" because I had some help from my dad: the design was the result of much groupthink. Most of the work, however, I did myself, including the shaping. Dad did the initial rip down the length of the car with a radial arm saw, did most of the handling of molten lead, and other potentially-hazardous things. I also had a friend of mine help paint the flames.
I consider it a perfect example of what a father-son project is supposed to be -- we both learned a lot, and had fun.
How odd that your dissertation looks just like the comment that I posted myself on these pages, an hour before you.
Strange, isn't it? I mean, just look: Compare this to this. The similarity is striking, isn't it?
Tell me, if you would, which of the following explanations is most applicable to this deceitful prose: A database error, some sort of maligned human oops, or just good old-fashioned plagiarism?
If it were a database error, the entire comment would have been duplicated verbatim, but it was not: the tags I used to italicize "nth degree" are absent, for example.
If it were some sort of human oops, I imagine that the such a hastily-construed cut-and-paste job would have included my sig, not to mention chunks of headers, leading and trailing stories, etc. In other words, by the time you'd reached a sufficient level of intoxication that your brain would permit you to accidentally duplicate a comment of such length, other mistakes would be apparent.
Therefore, I deduce plagiarism to be the cause.
Which leads me to be curious: How much code do you intend to steal for inclusion in your upcoming PDK product?
It's not too hard. As others have pointed out, center of gravity means nothing, except that a lower center would tend to have less cross-sectional area (which is good). All you have to worry about is friction and weight.
.001g over, use a sharp knife and remove a sliver of tape to finish the tuning. You'd be amazed how much a bit of vinyl electrical tape can weigh...
Friction is the hard part, and you've got two kinds to deal with: Friction with the air, and friction between the wheels and their axles.
For wind resistance, just start slicing the car down as much as you feel you can without sacrificing strength, or capacity for added weight. Round off both ends of the block into something that looks reasonably aerodynamic - I used a bench-mounted belt sander. You want it low, and flat, so as to displace as little air as possible as it moves down the track. Using similar curves to a high-speed train will yield better results than mimicking an F14 or Lamborghini. That the resultant form is rather boring is not an accident.
Friction against paint is probably not a big deal. My finish was a hand-brushed yellow-and-black flame paint job, and was nowhere near smooth. But it certainly wouldn't hurt to make it as mirror-like as possible with careful painting and sanding, along with a coat of well-buffed wax.
One trick I used which I've never seen repeated: A small circle of electrical tape, placed over the outside of each wheel, to further reduce drag. They were cut with a ball peen hammer and a coin of appropriate diameter, and fit precisely. It should be flat or slightly convex, but not concave, when applied to the wheel. This will increase rotational inertia (==bad) somewhat, but it seems like a beneficial thing to do.
To reduce friction between wheel and axle, I first removed the burr on the supplied axles using progressively fine sandpaper - there were two axial burrs, along with one across the back of the head. I also removed burrs on the inside of the wheels, and slightly rounded the flat part near the middle, inside of the wheels where it can touch the body, to reduce contact area.
Use graphite on the wheels and bearings, and use it as much as a polish as you do a lubricant. Work it into the microscopic texture of the axles, the inside of the wheels, and the bearing surfaces between the axle and the body by whatever means you can find, and then add some more and roll the car around a bit.
I used feeler stock (thinner-than-hair bits of flat metal) to get each wheel to within a thousandth-or-so of an inch of being on the same plane, such that they'd all be in contact with the track surface. I don't know if this helped, but it seemed like a good idea.
At least, try to keep your axles as straight and square as possible with eachother and the body. You do not want any pressure on the points where the wheel contacts the body, or the head of the axle.
And adjust the axles so that there is as little play in them as possible. The wheel should not move appreciably side-to-side, nor should it forcably rub on any point. Slop here translates to lost energy. It may take several iterations of gentle tapping on the axle and even-more-gentle pulling on the wheel to get this right.
Weighting is a science. You want the car to be as heavy as it possibly can be. Remember that you won't be immediately disqualified for having a very slightly heavy car, and that you'll get at least a couple of chances to bring the weight down.
Drill holes in the bottom of the car, and pour lead into them. I used an antique balance I had at home and added weight until I measured it being just -over- specification, and so the lead was protruding slightly below the bottom of the car. This allowed me to use the official scales to tune the car at the event, removing a bit of lead with a file and re-weighing several times before nailing it precisely, while in the process filing the lead flat and reducing drag.
Place a strip or two of electrical tape over the lead, to further reduce drag and add the nth degree of tuning: If you file a bit of lead off and the car still measures
If your track levels out toward the bottom, put the weight as far toward the back of the car as you can. You'll get a few more inches of "thrust" by doing this, vs. your front-weighted opponent.
The devil's in the details. When it's all done, handle the car as if it were full of nitroglycerine - don't even look at it funny, or all of the precarious work you've put into it will begin to undo itself. Never let others handle it, and absolutely never leave it unattended at the event. It's a finely-tuned, very delicate instrument, and is deserving of respect.
Tablets of this sort have been around for a long time -- the most popular made by Wacom, and the cheapest model being the $99 Graphire. I've seen these for sale even at Wal-Mart..
Not a bad kit, really - you get a pen and a scrollwheel mouse that you can use interchangably. Neither the pen nor the mouse require any electricity at all, instead relying on magnets and some trick inductive stuff inside the tablet. It's pressure-sensitive, too.
Simple:
Wireless is good so that you can watch porn straight from your neighbor's PC from the comfort of your couch, without all the hassle of maxing both cable connections, stringing Cat5, or even notifying him...
Just buy a Plextor. I've been using their 8x PR-820 drive for almost four years to the day.
A few thousand burns later, I've got no trouble to report except with a bad batch of Verbatim media toward the end of 1999.
And, mind you, this is for all manner of material -- from PSX archiving to music production to bulk duplication, usually on the cheapest media I can find. I've never burned at less than maximum speed.
YMMV, HTH. But given this experience I'm not likely to ever buy anything other than Plextor in the future -- that is, if this drive ever dies so that I can justify replacing it.
You're not creative enough.
I used to live in a 100-year-old house. It had modern, 200-amp electricity and indoor plumbing. Those must've been pretty serious installation projects compared to the little stuff I did with it.
It's had RG-58 installed for 10base-2, replaced by Cat5 for 10/100. The ISDN demarc was on the back of the house; the ISDN TA was at the front, upstairs, fed by a new run of Cat5. It's had cable TV, Primestar, and DirecTV. Along with a rooftop TV antenna and an FM antenna up in the attic. Telephones in every room without a toilet, with up to four live lines coming in at a given time.
It's just not that hard to do with old wooden houses. They've all got complete crawlspaces or basements, and are topped with capacious attics - you can get anywhere on either floor from one of these places. The walls frequently contain completely disused chimneys, which make fine vertical conduit. If you're worried about water and chimney cruft harming your precious Cat5, just buy some that's Teflon insulated (read: "plenum").
And since they've had plumbing and electricity added, all of these internal spaces are very likely to be easily accessible (unlike a modern buttoned-down house).
If you're lucky, the structure might even be balloon-framed, with studs made of -long- lumber stretching from the roof to the foundation without interruption.
And since the walls are either uninsulated or blown (unless it's been on fire in the past), you can shove wires down these stud cavities with a rod anywhere you want, without accumulating a snowball of fiberglass on the end of the rod.
If it has forced-air heat, it's usually really, really easy to get wires from Point A to point B by way of ductwork.
There's always a plethora of open passages left over at different times from the installation of indoor plumbing, centralized heating, knob-and-tube wiring, and then jacketed wiring.
You've got it easy, running your skinny little Cat5 around, compared to the efforts of those who've lived there before to install modern utilities.
Where have we heard what before? Are you talking about The Bubble?
Of course companies seek to profit.
As history dictates, the health of a company is not related to the long-term support of its products. Support gets killed in the interest of profitability as much as it does for the lack of it. It doesn't take much of a beancounter to realize these few things:
Products which are no longer produced generate zero revenue, but cost money to support. The longer you support old products, the longer your users will postpone buying your new products.
Thus, the more poorly you support your old products, the more likely people are to be in the market for something new. If all companies operated like this (they do), it'd be a goldmine (it is).
Since we're thusly assured to have zero support from the vendor in the not-distance future, I submit that the only way to ensure any future usefulness from the products you buy is to make an effort to get the same thing everyone else seems to have. Or at least something built from the same, or compatibile, component parts.
Think NE2000 (and clones), 3c509, DEC Tulip (and clones), HP LaserJet (and clones). Think Oxford 911 firewire/ATAPI bridges, NCR/Symbios/LSI/whoevertheyarenow 53c87x SCSI chipsets, and VESA BIOS.
If you've got the NE2000 or Tulip of 802.11a, you won't have any problems keeping your gear working with current software -- Microsoft and a dedicated team of rodent-like OSS hackers will keep it alive for as long as it remains useful, plus a few years.
I'm afraid I don't quite follow what you're saying.
If I have an unrestricted 10Mbps between myself and my friends (and their friends, and their freinds' friends, and...), you better believe that MP3 will floweth freely amongst the group, without any one person needing to spend a dime.
Of course, a number of these people will also subscribe to some form of consumer broadband. Not to mention those who rip their own CDs. There's just as many avenues for new material to enter the mix as there is for cross-pollenation of, say, Gnutella and Kazaa.
And speaking of broadband, I pay ~$50/mo for 2000/384kbps RoadRunner. I can't fathom sharing a paltry 1536kbps amongst 10 of my greediest peers (I like burning ISOs in realtime as they download), nor can I imagine that sharing of such services would be tolerated for very long.
Only one thing is needed for this to succeed: People willing to spend money in a small geographic area in order to talk amongst themselves. It's not gonna be easy, though.
Back in the day, it'd probably have been fairly easy to get people to drop a few hundred bucks on hardware in exchange for free local-area bandwidth.
You'd need only post a message to a local BBS suggesting the idea, debate it for a week or two, round up a couple of capable people willing to help with installations. After that, proceed with the leeching.
First, the sysops would connect to eachother, and by way of their ubiquitous, incessant bragging would encourage others to tie into the network.
If 802.11/a/b had been available back then, at today's prices, such networks would have been plentiful. But it's much too late for that.
The local scene is all-but-nonexistant these days. LUGs and 2600 notwithstanding, there's no way to communicate with a semi-clued cross section of the local populace anymore.
This social problem combined with the fact that there is no percieved advantage to talking to a computer in the same town vs. one is Sydney, Australia makes the prospects of participation look awfully dim.
So. Since there's no local forum by which to arrange such activities, and there's no compelling reason for people to join such a network instead of paying for DSL, good salesmanship is the only way to produce such a beast today.
People will first think you're crazy when you tell them that you want them to buy a few hundred dollars with of cabling, Yagis, and gear, only to talk for free with a half-dozen strange geeks around town. And then they'll question your motives, wondering what your take in it is.
It's a little early for the masses to get ready to participate in something of this scale and organization. In order for people to want to join such a network, a few things need to happen first:
1) Transfer caps, and/or the return of metered dialup access. People, as a rule, are afraid of buying new things, especially when what they have appears to work justfine. It doesn't matter if this seems irrational or not. People who drive Fords generally continue to drive Fords until their Crown Vic explodes one morning on the way to church.
2) Remember that killer app everyone was looking for during the dot.com boom? It remains elusive to this day. If it is ever discovered, and has sufficient local interest for people to justify dropping cash on hardware in order to participate, things might have a chance. And AFAICT, except in new markets, the broadband thing is getting pretty stale. The freenet needs a killer app.
P2P is close, but the potential for such unrefined protocols to spill over from the freenet onto paid internet links makes for some hefty financial baggage for someone to tote. You could charge for internet access, but what would be the point? It's supposed to be free, remember?
3) 802.11 gear needs to come down in price. With Wal-Mart selling new PCs for less than the cost of my first CD player, wireless gear must look quite expensive to Joe Average in comparison to their "free" cable modem.
4) Additionally, we need more spectrum in a band capable of traversing more than a few miles of open terrain, without hanging antennas several hundred feet in the air. Swapping files and playing Unreal with Joe Across Town would be fun, but is rather limited in scope. People in Toledo ought to be able to be able to at least send email to folks in Detroit without unreasonable delay. There will be a handful of savvy volunteers willing to maintain short-haul inter-city links, but only if there's popular demand and a low-cost technical means of putting them together -- after all, any way to avoid paying Ma Bell for T1 circuits is a Good Thing in this context.
5) I touched on this before, but we'll need better protocols, or at least good shims for existing ones. If such a network ever happens, it will be one where there are relatively vast amounts of relatively local bandwidth for free, with things slowing down considerably as distance inreases (along with increased costs that someone will have to cover).
A diverse network of transparent squid proxies will help markedly with http, but what about ftp? Cursedly indiscriminate P2P? Even IRC can consume substantial bandwidth with a few thousand consumers. All of these things can be cached/proxied/otherwise-massaged, rate-limited, or simply operated on a strictly-local basis. But, someone along the way will have to write, implement, and enforce these policies at the border. The investment in time required for such needed, hatred-inspiring rulemaking will not be small. (Read: operating gateways is going to be expensive, time-consuming, and quite thankless.)
6) And for the sanity-in-infrastructure requirements, it'll have to run IPV6 more-or-less exclusively -- nobody wants to be known as 10.4.249.57, and real IPV4 address space costs money. Besides, in a free network, there's no reason for every man, woman, child's palm pilot, bicycle, and toaster oven to not have unique addresses. Is the state of IPV6 under any incarnation of Windows sufficient to support this endeavor?
That all said, you could count me in if I didn't live in a downstairs apartment with windows facing outside of town.
Is the idea too early, or too late? You tell me. Personally, I don't see many of these problems being solved very soon in the current climate, but you be the judge.
If you had a proper backup of that data, you'd have little trouble migrating.
tar is about as universal as imaginable.
I used to repair fire-damaged structures for a living, so I've seen this more than a few times.
The easiest, and often cheapest, thing to do is throw everything away and cash an insurance check for new stuff.
For the stuff you don't want to throw away (because it is sentimental or otherwise irreplaceable), try Formula 409, Simple Green, or another general-purpose detergent. Use a brush, a cloth, or whatever seems appropriate.
Electronics can have a tendancy to not survive such cleaning, however. There's (at least) a thousand types of glue used to hold the bits that comprise them together, and no telling how they'll react to chemical treatment.
Water, of course, is generally harmless, but has limited application on smoke residue.
A friend of mine's house burned a few years ago, and the fire department vented the ceiling of his computer room...which is to say that vast amounts of smoke and steam went flying past his gear. I found a number of CDs in that room with their jewel cases melted off of them.
Most of it is still working justfine today without any cleaning, though the CDs did require some special care to come back to life.
You'll also do well to hire a company who specializes in such cleaning projects, if you want to try salvaging stuff. I used to contract with Serv-Pro (they've got offices all over - check the phone book), and they were often able to restore things to new.
Plus, they had a fleet of athletic 18-24 girls to do the work, which always brightened up my workday.
Once you're ready to start painting and carpeting, make sure you coat everything with a good primer. I usually rented a fairly serious airless paint sprayer, and used Killz, or Pro-Block from Sherwin Williams. Killz does a somewhat better job, being shellac, but the alcohol base will kill you dead if you're not extremely careful with your respirator. Plus, there's probably flash hazards with spraying alcohol everywhere... Pro-Block works very nearly as well, is oil-based (thus possible to coexist with) and is a great deal less expensive.
Cover the walls, ceiling, any exposed studs, the floorboards, and everything else you want to never smell of smoke again.
Yes; we should be acting like spoilt little children, to as much of an extent as our pocketbooks will permit.
.
It is, after all, the American dream
A few months ago, I installed Slack 3.9 on a 12-meg, 386SL/25 laptop. 3.9 was the last to includes a 2.0 kernel and libc5, so I figured it'd be fastest.
It ran fairly well. No accelerated drivers for X, or anything else other than Windows 3.11, so it lived in text mode.
It ran surprisingly well. Of course, the 386SL is an odd beast, with on-chip memory and cache controllers -- this was doubtless the source of some improvement in speed over what I expected.
Some time later, something-or-other trounced the hard drive rather completely, so I installed Slackware 8.1.
It runs surprisingly well. Some things even seemed to be a touch faster than they were with 3.9.
Thus, I'd like to submit that -all- versions of Slackware are suitable for old hardware.
Naah.
According to this chart, channel 6's FM audio resides at 87.75MHz.
College/non-profit radio station FM frequency allocations start at 88.1MHz.
Meteor scatter manifests as the ability to recieve distant, over-the-horizon broadcasts which are normally inaccessible, not as the ability to recieve local stations at a different frequency than perhaps you should.
What you are experiencing is just one FM broadcaster stomping on another's frequency, and the tuner being unable to sort them out. You've heard this before: When driving in the car trying listen to 103.7, but getting bleed from the adjacent stations at 103.5 and 103.9.
A Rotel tuner that I have will tune down to 87.5MHz. I used to get a kick out of listening to channel 6's audio with the stereo.
Airflow? He's mounting a single 1U box. And even if he wanted more, it is usually exceedingly difficult to find road cases in excess of 12U. Which just isn't very much equipment.
Remember, the intended purpose of these racks is not just for the shipping of equipment, but the use of it. Sensitive analog (and increasingly, digital) electronics upon which multi-million dollar shows (and hundreds of thousands of potentially-angry people) are relying on.
This type of stuff generally makes a lot more heat than a couple of PCs. And it's sitting in direct sunlight on a 105 degree day...that is, unless it's getting rained on.
Somehow, airflow seems to be OK for these folks. And they take it at least as seriously as any of you computer kids.
Is there any particular reason to be using copper?
I mean, sure: it's a fine conductor. But for soaking up RF in the context of a home computer, it would seem that steel would work fine. And rather than being ~$3/ft^2, steel is usually of negligible cost.
The best place to look for this sort of stuff is not at a business selling fabric to housewives or mail-order distributors, but your neighborhood hardware store. Not Lowe's or Home Depot, but the tired-looking place downtown that sells shotgun shells out of the same glass display case that props up their singular cash register.
In this store, you will find a man who looks as old as the building. He will be helpful. You'll either be directed toward the precise item that you're looking for, be able to order it for little cost, or be given information about another tired-looking local business that specializes in such things as wire mesh.
Myself, I've seen copper window screen sold in these places, by the foot. No idea if the holes are too big for the frequencies in question.
Failing that, there's other options. AFAIK, the requirements for a Faraday cage do not stipulate that the material must be hole-y.
From the same hardware merchant (or any of the fabric stores that you're so fond of), you'll be able to buy a can of spray-on adhesive. From your local food service vendor, you'll be able to buy a wide roll of aluminum foil.
You'll need one (1) aerosol can of adhesive, and one (1) roll of foil.
Apply adhesive according to directions on label. Apply foil according to common sense, bearing in mind that the adhesive will not let go of the foil, ever. If grounding connections must be made between movable portions of the case (side panels, for instance), make sure that the foil is applied in such a way that they're able to touch eachother with at least slight pressure.
Or, if you want to be a pedant, join the sections together with copper foil tape.
Done. Time saved: days. Money saved: At least $65.
[OT: What, pray tell, is Slashdot's average IQ these days? I deal with stupid, helpless people for a living, but I'm often amazed by what I discover on these pages.]
I think that your 20% figure is rather low, and greatly overestimates the level of knowledge posessed by today's typical internet user[1].
I therefore submit the following conjecture:
70% of bandwidth-equipped internet users are too stupid to do anything at all about sharing. 1/3 of these people are behind Linksys routers that they know nothing about, or are otherwise irrevocably firewalled.
This leaves about 46% of the unwashed masses sharing by default. That's about 32% of the whole Internet populace can do nothing but share, because they don't know any better.
And then there's the remaining 30%-big category of people who might be able to control whether or not they share with any surety.
A third of the people in this group are spiteful leeches and will never share. Another third favor equality, and will make an effort to give back what they take. And the remaining third are intelligent enough to realize that the more people share, the lesser the individual bandwidth penalty for sharing becomes. This latter group will spew forth as much data as needs dictate and their connection permits, thus cancelling the antisocial efforts of the spiteful portion.
Thus, on average, 62% of all internet users share.
This means that to sustain a stream, this 62% sharing majority of users will need to relay the program to an average of 1.6 people each. Some will do a bit more, some a bit less.
My RR cable modem wouldn't even notice if I started streaming at 50kbps (which sounds remarkably good with ogg) to 4 or 5 other people, let alone the paltry 1 or 2 listeners that such a system as this requires to stay healthy.
That all said, it'd sure be nice if multicast IP became a widespread reality. It'd put all of these issues to bed with much haste...
1: See "Jargon File," under section titled "The September that never ended."
Good points, all of them.
;)
I'd like to add that I'm quite happily running Slackware 8.1 with a 2.4 kernel on a 25MHZ 386SLC laptop. It won't win awards, but it does pull a (quite remarkable, really) 300KB/sec from its PCMCIA NE2K.
It also suffers from a laggy 2-gig Hitachi drive, which is probably not dissimilar from that which you have but is, I assure you, a good deal faster than the 150-meg JVC it replaced.
And, as you might surmise, I'm reasonably pleased with it. I don't ask much, it doesn't do much. It is rock-solid stable, though...
I only hope it might put your disheartening comments toward the forlonged 486 genre into a more proper perspective.
But that's just a toy, such as it is. On to more important things:
Karma: Big Swinging Dick (mostly affected by moderation done to your comments)
Yes, but who got +5 in this thread?
Good information.
Now, wrap it into a (pre-existing?) HOWTO. Get it published on defcon1.org. Whatevevr it takes.
But don't leave it as a random, 1-sentence Score:3 posting on Slashdot, where it will do little good for future masses encountering the same, doubtless growing, problem.
Thanks.
Yep. Might as well use it for other things. That's why people buy SUVs instead of more practical vehicles.
/dev/random > /dev/null, or some other CPU-eating task. Rest well knowing that your inexpensive processor is always working as hard as possibly it can.
From the sound of what you're now saying, you're more interested in building a general-purpose home server than an inexpensive, reliable replacement for your troubled Linksys box.
Your priorities seem to be misaligned toward serving this end. If speed is enough of a factor that you care about the CPU at all (honestly, a 486 can handle justabout any consumer routing scenario without breaking a sweat), you're not going to be happy with that slothly laptop drive.
You're probably also not going to be happy with a small, unexpandable system, as long as you've got other uses in mind. Better off going at least to MicroATX so that you get a few real slots, and aren't so tempted to put everything into a hideously small case.
That all said, I still feel that a single-purpose, solid-state mini-itx system would be the ideal replacement for the single-purpose, solid-state Linksys box: it follows KISS procedure. That a 400MHz CPU may be underutilized by merely routing your broadband connection does not mean that it must be burdened by serving other unrelated tasks.
You are after stability, right? You're sick of tinkering with the Linksys thing, trying to make it work correctly. So, build a replacement that works correctly, and resist the urge to tinker with it. Leave it the hell alone. Unless some security issue forces a kernel upgrade or somesuch, if it's not broke, don't fix it. Being solid state, "breaking" is rather unlikely to occur unless it is disturbed.
In other words, just put it in the corner of your basement and forget about it.
Good routers, in this context, are transparent and invisible, as all infrastructure should be. Treat it like you thought you'd be treating your Linksys router when you bought it, before discovering that it doesn't work well for you; it shouldn't require much more care and feeding than your water heater.
And if, for some reason, you still can't get over the urge to make maximal use of what you've built, simply run seti@home, cat
Good luck, though, whatever you do.
Why bother with a laptop disk?
It's just a firewall. It doesn't need mass storage, or at least nothing more than few megs. It just needs to be reliable.
So. Just beg your friend for the throwaway 8- or 16-meg compactflash card that came with his camera, and plug it into one of these.
Less power (can we say "fanless PSU"?), more speed, and superb reliability. With proper research, the adapter should be in the same price range as the 2.5" IDE adapter kit that you'd need for a laptop drive...
Save the hard drive for things that can benefit from the space.
The DjVu folks seem to have created a very fine format for the storage of scanned documents of random layout. It uses space efficiently, and produces good-looking (though imperfect - think JPEG) results. It's the only way I know of to preserve the look of a paper document without throwing away vast amounts of storage.
But, as far as I can tell, that's where the fun stops. AFAIK, it doesn't handle vector graphics, and has nothing to offer over PDF for strictly digital documents. In the digital world, PDF produces perfect results, automatically.
OTOH, PDF is not geared toward scanned documents. I've seen a lot of examples of, say, scanned datasheets in PDF; all of them were bad.
I thus submit that each respective format has its own well-defined niche, and fills it admirably.
Ghostview certainly could support DjVu, if someone wrote the appropriate code for ghostscript. That it supports GIF, JPEG, PNG and PDF would tend to indicate that it's well within the scope of the software's design parameters.