TiVo's lack of support for standard-definition digital cable is no big deal, but HDTV is another matter.
Digital SDTV is a maximum of 720x480 (or 704x486 (704=22x32) or something else close) which is similar to VGA's 640x480 but the pixels are narrower than they are tall. PAL digital SDTV is 720x540 with almost-square pixels.
This maximum is the same as DVD and also known as CCIR-601. However, digital cable might have lower resolution to save bandwidth. TCTI/AT&T/NuevoComcast uses 352x486 on most channels on the HITS satellite and it's likely the content is softened (low-pass filtered) a bit before real-time-encoding.
Therefore, re-digitizing and re-encoding the standard-def analog stream coming out of your digital cable set-top box is only moderately horrible. Motion artifacts will be a bigger issue than resolution because TiVo encodes in real time so can't go back & choose keyframes more wisely later. It also costs 1/100th of the encoders used by HITS and DirectTV do.
Real-time-encoding of SDTV by a sub-$500 box is a reasonable thing in 2003. HDTV is another matter and the digital cable boxes I know of (Scientific-Atlanta Explorer 2000HD) only have analog video outputs (YCrCb component for HD). Pinnacle Systems makes a system where the HD option alone is $1000. HD on PCs now is where SD was 10 years ago- intra-frame (Motion-JPEG/DV-style) or no compression, using oodles of disk space (even with today's 180GB drives). Uncompressed HD @ 1920x1080x30x12bpp (4:2:0) is 90 megabytes per second. That means burning through a 180GB hard drive in about 1/2 hour.
As the poster suggests, you want to get the MPEG-2 stream & just slap it on a disk instead of trying to recompress. For Over-the-Air HDTV broadcasts, this should be no problem. For cable systems that keep OTA in 8VSB...
(something boxes were required to do a few years ago even if the provider doesn't support it- they have to pass through 8VSB with enough bandwidth/low enough noise that a receiver can still demodulate&decode it)
an 8VSB-in-only HDTV PVR would work. Many systems are demodulating 8VSB and re-modulating at QAM64. If they also apply their conditional access (CA), it gets really sticky.
The fact that there aren't digital-cable-ready TVs like there were(are) cable-ready TVs is something the industry, their Cable Labs group and the FCC have been working on for years. The biggest obstacle is Scientific-Atlanta and General Instrument (now Motorola)'s incompatible systems in the US. It's possible to run both on a single network under an agreement called Harmony, but they still see CA as the crown jewels.
POD (point-of-deployment CA, rented from the cable company) was supposed to solve that by putting all the proprietary stuff in a PCMCIA-like card & making the boxes or TV's or VCRs or PVRs use standard interfaces.
Google terms: PowerKEY (SA's system), DigiCipher (Motorola's), Conditional Access.
NASCAR race cars still are based on road cars, just not road cars that are still made.
Remember that the NASCAR you watch on network TV (Winston Cup and Busch Lights (Lites?)) is only the top of the ladder. My local dirt track is NASCAR-sanctioned.
The cars that race there are old American V8 beaters upgraded with fancy carbs, heads, oil coolers & suspension pieces.
It's a pretty cheap hobby to get into and that's part of the appeal of the whole ladder. While many may not seriously pursue moving all the way to Winston Cup, they must feel some kinship with Rusty & Jeff & Tony.
As the pro side of Import Drags develops, it should benefit (the sport, sponsors and upgrade manufacturers) from the same kind of hero worship & emulation that superbikes & stock cars do.
In Europe & Australia, Touring Cars are just "modern" stock cars and might also prosper from this principle. Mazda's gently (they're cash strapped) pushing production-car (road) racing in the US, but it has a long way to go.
On the gripping hand, the inability for a driver to overrev (and subsequently destroy) an engine lowers costs for less well funded teams like Minardi so they while they might be dominated, at least they can stay in the show.
The 20GB 5.25" 1/2 height drive was a disaster. I worked in a clone shop in those days & half the drives I pulled from stock failed, with a good number failing before the customer even took the machine home (burn-in was usually ~1 day).
However, the 20GB 3.5" drive was faster (40ms vs 65ms average seek) and I never had one fail. It was more expensive and one of the first drives in that form factor.
I had few failures with the 40GB 5.25" 1/2 heights . The 30GB RLLs failed completely less often than the 20GB MFMs, but had horrible data corruption problems.
My favorite drives were the 40GB full-height drives. They used some kind of linear actuator instead of a swing-arm for the heads & shook the whole bench when running seek tests. Start-up currents are instructive (12V supply): 4A for the ST4051 full-height 5.25", 2.4A for the ST225 1/2 height 5.25" and 2A for the ST125 1/2 height 3.5".
Anyone know if the 20GB MFM coincided with a new factory opening in Singapore?
Some point-and-shoot film cameras in some situations can take photos that are technically excellent (resolution, contrast, color accuracy), but point-and-shoot digital cameras are limited to small photosites in their CCDs (noise or poor sensitivity) and 2-5 megapixels (poor resolution).
The trick is to get a good lens. It's hard to build a small, cheap, light and fast zoom lens, but a fixed focal length lens like the f/2.8 on the US$100 Olympus Stylus Epic plus good light and good film (Kodak Gold 100 is OK) and a subject wearing a striped shirt (or a UPC tatoo near the eyes- even better) so the autofocus really works has given me really nice photographs.
If you go to ASA 400, you're going to get grain, but you'll do better than the noise of a cheap CCD with that much gain.
Digital point-and-shoots are likely to get better, but to take advantage of the developments, you have to buy a new camera. To take advantage of chemical photo developments (we've come a long way in ~100 years & over 50 years of 35mm), just get a new roll of film.
However, if this happens, their profit model will immediately face a new menace: "virtual copyright infringment". Some hackers will create/modify a software MPEG player so that it applies a simple EDL (Edit Directive List) to the video as it plays.
1) An EDL doesn't have to be spread on a p2p network or through email. Under present US copyright law, the EDL is probably not a derivative work.
That's why I advocated distributing things like The Phantom Editor's cut of Star Wars Episode I as EDLs instead of divx over p2p.
However...
2) Your assumption that when a source of high quality digitized video appears, a tool to play the video using an EDL will magically appear also is wishful thinking (I wish too, but recgonize it only as a wish).
DVDs already exists & Netflix claims to have 12,000 of them- up to 9GB each, less than 48 hours from my DVD-ROM drive. That's 1/2 as fast as my DSL on a good day & in some ways more convenient.
There are plenty of open source DVD players that could be modified to play CMX EDLs of every enthusiasts's alternate edit of TPM, Memento and the Matrix.
On the gripping hand, eliminating commercials seems to be more compelling than eliminating Jar-Jar. I see that EDLs for Babylon 5 for manually editing out commercials already exist.
Remember the movie Deep Impact? Came out a short time before Y2K and made me seriously consider getting a dirt bike to escape the rioting hordes.
Was talking to some co-workers (all engineers) and we came up with a small diesel-powered dead-simple dirt bike that'll run on fry grease, heating oil (for the NE USA), or kerosene as the ultimate post-apocalyptic vehicle. Should have power-take-off to run well pumps, generators (electrical, not electronic stuff that survives), huge torque & not a lot of power (no clear roads, so no 100MPH cruising anyway) and light enough for an average adult to right if it falls over (hard to do with a diesel).
If the center of gravity is very low, it'll be easier to right even if it is kind of heavy. It needs the torque & traction to run off road, but we aren't jumping dunes, so we don't need the suspension travel of a supercross bike, making the COG requirement easier.
---
Back on topic, Linksys ought to offer a trade-up program for the 1st generation WPC11 cards to get a 3.0 card for 1/2 price or something. That would go a long way towards people believing 802.11b actually works. The antennas on the early cards are TERRIBLE and though the WAP11 is ony 30mW (don't do the boost hack to 100mW- just pushes up sidelobes), and the card is not too powerful or sensitive, other cards using the same chipset have MUCH better ranges. Must be the antenna, eh?
Someone else mentioned serial ports for connecting to LCDs like these. If you want a graphical LCD with a fast update rate parallel may be the way to go.
Note that I'm talking about small LCDs and VFDs which might be monochrome & only have a resolution of 240x64 or 240x128. While the case modders tend to use them for displaying temperatures, mp3 song titles and spectrum bargraphs, think about the embedded possibilities. Not only are these LCDs smaller, but they draw less power than VGA and better color TFTs.
You could boot from a CompactFlash card and have a really teeny hot tub controller that uses fingerd to report current temperature & power consumption, using the parallel port to talk to the sensor/control board.
Consider a box with a 200GB 3.5" hard drive and a "universal" 12V power supply that'll run from a car, a big battery or a lump-in-the-line (power consumption a bit high for a wall-wart). Not a 1.7" or 2.5" drive, but enough room to hold every CD you own maybe even in FLAC instead of OGG as well as every digiphoto you've ever taken.
With no CD-ROM or AGP slot, it would be considerably smaller than a Shuttle XPC or even a Cubid. The new VIA EPIA M will have USB 2.0, FireWire and 100Mb/s ethernet for getting stuff in & out of your "All Box."
For controlling playback on the go add a few buttons and a rotary encoder scanned by the parallel port and a serial LCD (or a PIC to handle user input & a parallel LCD).
You can build smaller systems for ip to real-world, but this thing runs Linux and BSD, a real 32 bit processor with an MMU and plenty of memory. None of that mucking about with 8-bit assembly or (horrors) BASIC Stamps. Write your nuclear-bomb-tipped earth-boring machine control code in Java if you like!
Hardware for the embedded market is usually more expensive for equivalent capability as PC stuff. You can get an EPIA board for a lot less than an ARM development kit. [I know about power consumption, cost in volume, etc. etc., but for fun small projects, off-the-shelf wins.]
By the way, this guy has solve the problem of how to get 802.11b in a teeny box. The mini-ITX boards don't have PCMCIA and a PCI Wi-Fi card is going to take up a lot of room. Solution: gut a USB Wi-Fi box. The Orinoco one has a teensy USB to PCMCIA interface & a regular Wi-Fi card. Dunno 'bout Linux drivers for it, tho.
All that said, I would like to see a "legacy free" mini-ITX board for building regular-old personal computers.
It looks like Dragon's "modern" products (Naturally Speaking) are still published (& maintained?) by ScanSoft. Some vendors still show the "obsolete" discrete-recognition Dictate available (Dragon Dictate Power Edition 3.0 $189). If you Google for the part number 01-022-24-01 you'll find other vendors.
L&H also took down Kurzweil (the K. that did voice recognition, not the K. music systems co.).
Kurzweil made a discrete recognition system (Voice) that I used for programming for a while when whateverthefuck is wrong with my hands was kicking up.
I ran it on a Pentium 120MHz laptop w/ 1GB hard drive & it worked pretty well.
I still have the CD's, but it'll only run on Win95, not Win98SE.
Discrete is better for programming because continuous relies on the underlying syntax of the natural language (e.g. English) to hint it at reaonable words for ambiguous sounds. In programming, I say things like "left paren eye spacebar equalsign spacebar..."
Dragon's big advantage over Kurzweil was that it could be used completely hands free (including mouse) so they got the disabled market.
The discrete recognizers had to be trained (Voice called it "enrolling") so they'd know how you pronounce the phonemes, but they had huge vocabularies- you didn't have to speak every word, just enough to represent every sound. The version of Voice I had let me add words, so I could add gss_SetPixelMap & other symbols. There was a way to import a whole list of words, so I munged the output of nm & spent a couple of days training it to recognize every function, type, and variable in the PowerTV operating system.
Kurzweil Voice & Dragon Dictate had extensive "correction" features because they made a lot of mistakes. This started as very irritating, but you'd get used to it & you could talk ahead a few words, then correct the 3rd word back, then continue where you left off. When correcting, it would present a list of close-sounding words to pick from or you could spell out the word & add it if it was new.
There was even a "you keep screwing up" helper where you could give it two words that it confused a lot & re-train them.
One problem with speaking discretely (pausing between each word) is that the back of your throat takes more wear & you can start to go hoarse, causing the recognizer to get worse. Yer supposed to avoid carbonated beverages @ lunch.
Instead of running these old apps on a PDA, why not run 'em on a Crusoe subnotebook or something like the OQO?
...but what the heck does 'mere aggregation' mean in an embedded system???
The GPL & FAQ are written assuming PCs & distros on tape or CD-ROM. On a CD containing Nvidia driver binaries and the Linux kernel, you can pull the differently-licensed pieces apart.
What about an ROM containing some GPL code in binary form and proprietary closed-source code also in binary form, but the two parts are separately linked & one dynamically loads & links the other?
The aggregation doesn't seem so "mere" anymore & is less easily reversed by end-users.
What if the GPL binary is in a separate solid-state media (NAND Flash/Smartmedia soldered to the board) and loaded by the closed code running from ROM (NOR Flash)?
What if the GPL is on removable solid-state media?...if the interfaces are public ones?
If you don't like the idea of the binaries sharing a single ROM, what about TiVo? There the binaries share a hard drive and the device is not a PC, so the aggregation is not easily reversible. TiVo provided hackers a backdoor, but was this just a bone thrown so we wouldn't look too hard at GPL compliance or does it really make the product compliant?
[I'm not talking about TiVo's mods to the kernel, which they did release source for, but all the other code (drivers and application) which are closed source, but distributed with GPL'd code.]
nothing was said about broadcast flags, does this mean there wont be any? Or that it's still under debate? or did the FCC actually say "screw you" to the MPAA?
Actually, the broadcast flag was one of the other items on the agenda:
BUREAU: MEDIA TITLE: Digital Broadcast Copy Protection SUMMARY: The Commission will consider a Notice of Proposed Rulemaking concerning digital broadcast copy protection.
It's item #1, but the agenda was followed in exactly reverse order for some reason. A RealVideo recording of the stream is available, but will probably only be up for 6 months, so archive it now.
Any member of the public can comment on proposed rules. It's possible the stream of mail they get is quite a bit smaller than Congresspeople, so this push (by Hollings) of this issue to the FCC may actually be to cyber-activists advantage. Instructions on making a comment are on the FCC website.
is it possible for HBO et al. to broadcast Macrovision copy protection on their signal so that one cannot record such broadcasts?
On digital cable and direct-broadcast satellite, yes.
I don't know whether Macrovision can be sent over analog cable & whether it would be possible to unscramble the "premium channel" CATV scrambling while leaving (or adding in the box) Macrovision.
On digital boxes, however, extra bits in the MPEG sream or simply stuff in the program guide information can be used to turn on the Macrovision-generation stuff in the Digital ENCoder (DENC) which encodes YUV digital video into NTSC or PAL.
That is if the DENC has such a feature. I know that some chips (which are usually well-integrated: MPEG & AC-3 decode + graphics + DENC & in Broadcom's case, a complete "settop on a chip") are offered in two versions: Macrovision & not because some box builders might not want to pay for the license.
So, yes, in some cable/sat. systems (where the operator bought the right box that had the right chip), you could prevent a TiVo from being able to "back up" a show to VCR. Macrovision in cable boxes was primarily considered for Pay-per-view, though, not for premium channels. I don't know of any USA systems doing it, but I get the impression from a UK TiVo review that BskyB does.
One thought in the US was to have two pay-per-view prices-- pay more to disable macrovision. I believe that was an idea the box makers & system operators cooked up & don't know if the MPAA would go for it, but it suggests one way out of this that pisses off fewer people.
Maybe the policymakers need to distinguish between boxes like TiVo that make "ephemeral" copies and VCRs/DVD-recorders.
Let's suppose someone comes up with a way to create signed binaries that it will run without a mod chip (which might be pulled from the market under the DMCA).
Are these not-created-by-Microsoft authentication tags (for xbox linux, not a "backup" of a game) a violation of Biden's bill?
The "property" with a counterfeit authentication isn't Microsoft's, it's FSF's/Linus's/et al. so maybe the bill doesn't apply. I get hung up on the definition of "illicit authentication feature" in section 3 6 C "appears to be genuine, but is not." Is it A and (B or C) or A or B or C??
The casual user will not be able to do this. One guy in a thousand will (like almost everyone who reads/.)
Michael Eisner:
"DVD has been hacked, I believe. But, but, 10%, 5% of the people hack it, they have it on their T-shirts, they have the code, you know, they have to be a genius to figure out how do it. NORMAL people just put it in and say 'I'll pay the money.'"
Eisner also decried Apple's "Rip. Mix. Burn" campaign at Hollings' hearing, but said he's "addicted" to his iPod at the Disney earnings conference call a month later.
Apple should really prepare themselves for the sheep migration by selling a Mac for under $350, w/out monitor of course.
The box should have VGA out on a HD15 connector and PS/2 mouse and keyboard jacks so you don't even have to buy a new USB keyboard.
Your old PC getting slow & you were planning to upgrade? But a new PC requires Windows SS that wants to know your shoe size before booting, so you get a pcMac & just plop it where the PC used to go. Comes with Virtual PC & the last version of Windows that only sucked mostly instead of completely (2000).
1) USB has an isochronous mode (since at least 1.1).
2) You probably don't want to use isochronous for what is essentially SCSI emulation. isoc can be fast because bad data (usually) isn't retransmitted and error checking might be done on larger chunks o' data (e.g. up to 1023 bytes vs. 64 for USB 1.1 iso vs. bulk transfers).
Think UDP vs. TCP.
isoc.'s goal isn't speed but guaranteed throughput anyway. For USB, isoc. has a lower theoretical maximum throughput.
3) Your point of looking at how the raw bandwidth is actually used is a good one, but your details are misleading.
Here's a better example: USB 1.1 bulk transfers (used for Mass Storage Class among other things) are a maximum of 64 bytes and you can send a max. of 19 of them in one frame. There are 1000 frames per second. 19 * 64 * 8 / 1000 = 9.7 megabits/s vs. a raw rate of 12 mbits/s.
USB 2.0 bulk transfers can be up to 512 bytes per packet and 13 fit in a microframe for 426.0 mbits/s vs. 480 raw (there are 8 microframes per frame).
Mass Storage class requires a 31 byte command and 13 byte answer for every transaction. The largest I can remember seeing is maybe 32 kbytes at a time, so we're only losing 1/10th of 1% to protocol overhead at that level.
I only program USB controllers (though I love firewire for DV), so I'll leave it to someone more qualified to answer the efficiency questions for it.....
------
Incidentally, USB 1.1 isoc. does 2 512 byte packets per frame for 8.2 mbits/s. USB 2.0 can do 2 3072 byte packets per microframe for 393 mbits/s. In both cases it's actually less than for bulk transfers.
This won't be like region-locked DVD players that become region-free by cutting a jumper. Those can't be sold under the agreement. Waiting for the next Apex to make one without signing the agreement? Watch Congress put the force of law behind this agreement. After all, it isn't being rammed down our throat like the CPDTPA was, no, no, industries agreed (sarcasm) in private action to do this.
CSS wasn't actually that easy to crack. Robust products must be built to resist reverse-engineering so look out for tricks like the Xbox uses (you aren't running Linux on one, so don't tell me that bunnie's crack means much). Code obfuscation, secure busses, yep, it'll all be in your way.
Building your own HD receiver might end up easier. You'll have to get a tuner/QAM decoder that gives up the decoded signal in the clear, but at least open source MPEG-2 decoders are available.
Also, buy your (over-the-air) HD receivers now! They will never be encumbered by this bullshit. Satellite & cable boxes might be recalled, but they can't take away your Samsung SIR-T150 (~$600). Don't buy a combo box (they can update your firmware). Get an OTA-only one. Unencumbered HD tuner cards for PCs are another tech that will soon be out of production & $$$$ on ebay.
Hollings and others want industries to agree so they can back up the agreement with force of law.
Hollings said this in the Commerce Committe hearing on March 14th. He mentioned Macrovision as something the industry came up with that Congress then required in all VCRs.
He claims to not want to have to write the technical standards that would become law but to only be using that threat to drive industries to agree.
You might or might not believe him, but Orrin Hatch who is much calmer on this issue and seems a lot more relucant to write a tech. standard seems to believe in backing a "private" agreement.
The problem is that this isn't an agreement based on mutual compromise. It's everyone caving to the MPAA.
How about letting the heads of corporations (and their bootlickers in Congress) portray themselves.
I can't figure out if they are evil or stupid. It's hard to be manipulative & devious if you were shortchanged at the neuron dispenser. I'm drifting towards the opinion that it's all an act to ingratiate know-nothings who are suspicious of technology.
History of Publishing Website covers press control and censorship under Elizabeth I (1558-1603) and censorship used for military ends during the English Civil War (1642-1649).
it sounds to me that your district has already chosen a preferred representative in Congress. I'm sorry that he's not who you would have chosen.
I could be misinterpreting your implication, but it sounds like you're suggesting that the unliked candidate won fair & square & the poster has a case of sour grapes.
The candidate won in a particular system with rules that are not neutral nor the only possibility.
More parties won't help in most cases because they simply fracture the vote. We do already have a multi-party system to some extent because at least alternative parties are legal. Nader was banned from even watching the debates so there's a ways to go, but there are other alternatives:
At-large candidates. Increase the number of citizens per Representative to free up a few seats. Make those seats national so the candidates can campaign for the black vote, the geek vote or the pave-the-earth vote.
Running a national campaign for a seat that represents 1/435th of Congress might seem unwieldy and impractical, but many Congressional campaigns already receive national attention. Also narrow-cast media outlets such as issue-centric magazines, websites and discussion forums let a candidate get the word out much cheaper.
1) The right of monopoly control over distribution (a.k.a. copyright) is in the (main body of the) Constitution
2) The First Amendment seems to contradict this since copied speech is still speech.
3) The resolution was to employ the fair use doctrine. It was originally not laid out so plainly with the four tests in chapter 12 of the US Code, but was just a doctrine (that might pre-date the Constitution for all I know) that courts used to decide infringement cases.
4) Eventually, the four tests were codified in law, but they are still vague & given to courts' interpretation.
Fair use as defined in the USC is not a right, but a defense. However, fair use derived from the first amendment is a right.
Epson printers also have an anti-refill chip and they use separate head & ink. This is even in their sub-$100 printers.
It really is a (give us all your) money issue, not a QA one, though different inks & papers really do have an affect on not just quality in general but color balance and response curve. Print Image Matching (PIM) (profiling by design rather than instance) can only work if the ink and paper are known in advance.
Epson is also the only company (of hp, Lexmark and Canon) to use piezography instead of heat to drive the ink out of the head onto the paper.
Not having replaceable heads can lead to clogging problems. One way Epson tries to solve this is by running an ink-wasting "cleaning" cycle if you don't use the printer for a while. Might as well print....
If an Epson clogs, it's off to the service center. Canon offers the best compromise: heads not replaced with ink, but new heads are user replaceable. They also have Epson-esqe print quality for photos (hp & lexmark aren't in the same league) and separate ink tanks.
Congress doesn't just pass laws. They also threaten to pass laws.
That's what happened in the PMRC/porn rock mess: with a threat of a requirement from Congress, the record labels adopted "voluntary" labels. That one was a bizarre public/private issue: If Congress had really gone ahead & mandated labels, there was a good chance it would be ruled unconstitutional, but not only did they not introduce legislation, they had the Washington Wives in the PMRC hold the cudgel over the music industry's head.
In the Commerce Committe hearing on the CBDTPA, Hollings & Stevens made repeated reference to the fact that they were trying to resolve deadlock between the content pimps union vs. the consumer electronics & computer manufacturers.
They're trying to stay two steps back from mandating copy controls. If everyone makes nice tomorrow & agrees on the Broadcast Flag, the Analog Hole, a industry-wide DRM solution and what to do about P2P, Congress can go home. If they don't get to it soon, Congress will pass the bill, giving 'em another year to get around to it. After that, Congress picks the tech.
Hollings made reference to Macrovision: the movie studios said they'd distribute movies if they had Macrovision on 'em, the VCR mfg.'s said they'd put it on the VCRs, then Congress ratified the monopoly with legislation.
The problem is that some of the issues the various industries must agree on are unsolvable. Eisner wants some way to protect movies on a screen in a theater from videotaping, divxing and p2ping (he showed a clip of Black Hawk Down downloaded from Morpheus while the film was still in theaters). He was nearly foaming at the mouth he was so worked up about these crazy hackers (he made reference to the DeCSS t-shirt!) and their evil networks (maybe Disney should buy the Internet & do for it what they did for Times Square) copying unprotected content in the "ether."
The rep from Intel, Vadasz, seemed bewildered with Eisner's requests. I think Eisner knows there's no way to protect against this kind of stuff without turning computers into VCRs (after bitching about Rip.Mix.Burn. billboards encouraging piracy, he mentioned he's "addicted" to his iPod in the Disney earnings conference call). Valenti wants "two nerds in a garage in San Diego" to figure it out if big tech industry can't.
Slashdot Mirror
TiVo's lack of support for standard-definition digital cable is no big deal, but HDTV is another matter.
Digital SDTV is a maximum of 720x480 (or 704x486 (704=22x32) or something else close) which is similar to VGA's 640x480 but the pixels are narrower than they are tall. PAL digital SDTV is 720x540 with almost-square pixels.
This maximum is the same as DVD and also known as CCIR-601. However, digital cable might have lower resolution to save bandwidth. TCTI/AT&T/NuevoComcast uses 352x486 on most channels on the HITS satellite and it's likely the content is softened (low-pass filtered) a bit before real-time-encoding.
Therefore, re-digitizing and re-encoding the standard-def analog stream coming out of your digital cable set-top box is only moderately horrible. Motion artifacts will be a bigger issue than resolution because TiVo encodes in real time so can't go back & choose keyframes more wisely later. It also costs 1/100th of the encoders used by HITS and DirectTV do.
Real-time-encoding of SDTV by a sub-$500 box is a reasonable thing in 2003. HDTV is another matter and the digital cable boxes I know of (Scientific-Atlanta Explorer 2000HD) only have analog video outputs (YCrCb component for HD). Pinnacle Systems makes a system where the HD option alone is $1000. HD on PCs now is where SD was 10 years ago- intra-frame (Motion-JPEG/DV-style) or no compression, using oodles of disk space (even with today's 180GB drives). Uncompressed HD @ 1920x1080x30x12bpp (4:2:0) is 90 megabytes per second. That means burning through a 180GB hard drive in about 1/2 hour.
As the poster suggests, you want to get the MPEG-2 stream & just slap it on a disk instead of trying to recompress. For Over-the-Air HDTV broadcasts, this should be no problem. For cable systems that keep OTA in 8VSB...
(something boxes were required to do a few years ago even if the provider doesn't support it- they have to pass through 8VSB with enough bandwidth/low enough noise that a receiver can still demodulate&decode it)
an 8VSB-in-only HDTV PVR would work. Many systems are demodulating 8VSB and re-modulating at QAM64. If they also apply their conditional access (CA), it gets really sticky.
The fact that there aren't digital-cable-ready TVs like there were(are) cable-ready TVs is something the industry, their Cable Labs group and the FCC have been working on for years. The biggest obstacle is Scientific-Atlanta and General Instrument (now Motorola)'s incompatible systems in the US. It's possible to run both on a single network under an agreement called Harmony, but they still see CA as the crown jewels.
POD (point-of-deployment CA, rented from the cable company) was supposed to solve that by putting all the proprietary stuff in a PCMCIA-like card & making the boxes or TV's or VCRs or PVRs use standard interfaces.
Google terms: PowerKEY (SA's system), DigiCipher (Motorola's), Conditional Access.
Other sources: Multichannel News and Communications Engineering and Design (CED) Magazine
NASCAR race cars still are based on road cars, just not road cars that are still made.
Remember that the NASCAR you watch on network TV (Winston Cup and Busch Lights (Lites?)) is only the top of the ladder. My local dirt track is NASCAR-sanctioned.
The cars that race there are old American V8 beaters upgraded with fancy carbs, heads, oil coolers & suspension pieces.
It's a pretty cheap hobby to get into and that's part of the appeal of the whole ladder. While many may not seriously pursue moving all the way to Winston Cup, they must feel some kinship with Rusty & Jeff & Tony.
As the pro side of Import Drags develops, it should benefit (the sport, sponsors and upgrade manufacturers) from the same kind of hero worship & emulation that superbikes & stock cars do.
In Europe & Australia, Touring Cars are just "modern" stock cars and might also prosper from this principle. Mazda's gently (they're cash strapped) pushing production-car (road) racing in the US, but it has a long way to go.
-M
On the gripping hand, the inability for a driver to overrev (and subsequently destroy) an engine lowers costs for less well funded teams like Minardi so they while they might be dominated, at least they can stay in the show.
Two-way telemetry was legalized in the 2002 season. They used it to fix a stuck valve in Coulthard's car @ Monaco.
Seagate made good drives back then also.
The 20GB 5.25" 1/2 height drive was a disaster. I worked in a clone shop in those days & half the drives I pulled from stock failed, with a good number failing before the customer even took the machine home (burn-in was usually ~1 day).
However, the 20GB 3.5" drive was faster (40ms vs 65ms average seek) and I never had one fail. It was more expensive and one of the first drives in that form factor.
I had few failures with the 40GB 5.25" 1/2 heights . The 30GB RLLs failed completely less often than the 20GB MFMs, but had horrible data corruption problems.
My favorite drives were the 40GB full-height drives. They used some kind of linear actuator instead of a swing-arm for the heads & shook the whole bench when running seek tests. Start-up currents are instructive (12V supply): 4A for the ST4051 full-height 5.25", 2.4A for the ST225 1/2 height 5.25" and 2A for the ST125 1/2 height 3.5".
Anyone know if the 20GB MFM coincided with a new factory opening in Singapore?
This post on the Classic Computing mailing list has more.
Also search for segate st225 quality at Google...
Some point-and-shoot film cameras in some situations can take photos that are technically excellent (resolution, contrast, color accuracy), but point-and-shoot digital cameras are limited to small photosites in their CCDs (noise or poor sensitivity) and 2-5 megapixels (poor resolution).
The trick is to get a good lens. It's hard to build a small, cheap, light and fast zoom lens, but a fixed focal length lens like the f/2.8 on the US$100 Olympus Stylus Epic plus good light and good film (Kodak Gold 100 is OK) and a subject wearing a striped shirt (or a UPC tatoo near the eyes- even better) so the autofocus really works has given me really nice photographs.
If you go to ASA 400, you're going to get grain, but you'll do better than the noise of a cheap CCD with that much gain.
Digital point-and-shoots are likely to get better, but to take advantage of the developments, you have to buy a new camera. To take advantage of chemical photo developments (we've come a long way in ~100 years & over 50 years of 35mm), just get a new roll of film.
1) An EDL doesn't have to be spread on a p2p network or through email. Under present US copyright law, the EDL is probably not a derivative work.
That's why I advocated distributing things like The Phantom Editor's cut of Star Wars Episode I as EDLs instead of divx over p2p.
However...
2) Your assumption that when a source of high quality digitized video appears, a tool to play the video using an EDL will magically appear also is wishful thinking (I wish too, but recgonize it only as a wish).
DVDs already exists & Netflix claims to have 12,000 of them- up to 9GB each, less than 48 hours from my DVD-ROM drive. That's 1/2 as fast as my DSL on a good day & in some ways more convenient.
There are plenty of open source DVD players that could be modified to play CMX EDLs of every enthusiasts's alternate edit of TPM, Memento and the Matrix.
On the gripping hand, eliminating commercials seems to be more compelling than eliminating Jar-Jar. I see that EDLs for Babylon 5 for manually editing out commercials already exist.
P.S. I'm pretty sure EDL is Edit Decision List.
Nope. Roads will be blocked- cars useless.
Remember the movie Deep Impact? Came out a short time before Y2K and made me seriously consider getting a dirt bike to escape the rioting hordes.
Was talking to some co-workers (all engineers) and we came up with a small diesel-powered dead-simple dirt bike that'll run on fry grease, heating oil (for the NE USA), or kerosene as the ultimate post-apocalyptic vehicle. Should have power-take-off to run well pumps, generators (electrical, not electronic stuff that survives), huge torque & not a lot of power (no clear roads, so no 100MPH cruising anyway) and light enough for an average adult to right if it falls over (hard to do with a diesel).
If the center of gravity is very low, it'll be easier to right even if it is kind of heavy. It needs the torque & traction to run off road, but we aren't jumping dunes, so we don't need the suspension travel of a supercross bike, making the COG requirement easier.
---
Back on topic, Linksys ought to offer a trade-up program for the 1st generation WPC11 cards to get a 3.0 card for 1/2 price or something. That would go a long way towards people believing 802.11b actually works. The antennas on the early cards are TERRIBLE and though the WAP11 is ony 30mW (don't do the boost hack to 100mW- just pushes up sidelobes), and the card is not too powerful or sensitive, other cards using the same chipset have MUCH better ranges. Must be the antenna, eh?
Someone else mentioned serial ports for connecting to LCDs like these. If you want a graphical LCD with a fast update rate parallel may be the way to go.
Note that I'm talking about small LCDs and VFDs which might be monochrome & only have a resolution of 240x64 or 240x128. While the case modders tend to use them for displaying temperatures, mp3 song titles and spectrum bargraphs, think about the embedded possibilities. Not only are these LCDs smaller, but they draw less power than VGA and better color TFTs.
You could boot from a CompactFlash card and have a really teeny hot tub controller that uses fingerd to report current temperature & power consumption, using the parallel port to talk to the sensor/control board.
Consider a box with a 200GB 3.5" hard drive and a "universal" 12V power supply that'll run from a car, a big battery or a lump-in-the-line (power consumption a bit high for a wall-wart). Not a 1.7" or 2.5" drive, but enough room to hold every CD you own maybe even in FLAC instead of OGG as well as every digiphoto you've ever taken.
With no CD-ROM or AGP slot, it would be considerably smaller than a Shuttle XPC or even a Cubid. The new VIA EPIA M will have USB 2.0, FireWire and 100Mb/s ethernet for getting stuff in & out of your "All Box."
For controlling playback on the go add a few buttons and a rotary encoder scanned by the parallel port and a serial LCD (or a PIC to handle user input & a parallel LCD).
You can build smaller systems for ip to real-world, but this thing runs Linux and BSD, a real 32 bit processor with an MMU and plenty of memory. None of that mucking about with 8-bit assembly or (horrors) BASIC Stamps. Write your nuclear-bomb-tipped earth-boring machine control code in Java if you like!
Hardware for the embedded market is usually more expensive for equivalent capability as PC stuff. You can get an EPIA board for a lot less than an ARM development kit. [I know about power consumption, cost in volume, etc. etc., but for fun small projects, off-the-shelf wins.]
By the way, this guy has solve the problem of how to get 802.11b in a teeny box. The mini-ITX boards don't have PCMCIA and a PCI Wi-Fi card is going to take up a lot of room. Solution: gut a USB Wi-Fi box. The Orinoco one has a teensy USB to PCMCIA interface & a regular Wi-Fi card. Dunno 'bout Linux drivers for it, tho.
All that said, I would like to see a "legacy free" mini-ITX board for building regular-old personal computers.
It looks like Dragon's "modern" products (Naturally Speaking) are still published (& maintained?) by ScanSoft. Some vendors still show the "obsolete" discrete-recognition Dictate available (Dragon Dictate Power Edition 3.0 $189). If you Google for the part number 01-022-24-01 you'll find other vendors.
L&H also took down Kurzweil (the K. that did voice recognition, not the K. music systems co.).
Kurzweil made a discrete recognition system (Voice) that I used for programming for a while when whateverthefuck is wrong with my hands was kicking up.
I ran it on a Pentium 120MHz laptop w/ 1GB hard drive & it worked pretty well.
I still have the CD's, but it'll only run on Win95, not Win98SE.
Discrete is better for programming because continuous relies on the underlying syntax of the natural language (e.g. English) to hint it at reaonable words for ambiguous sounds. In programming, I say things like "left paren eye spacebar equalsign spacebar..."
Dragon's big advantage over Kurzweil was that it could be used completely hands free (including mouse) so they got the disabled market.
The discrete recognizers had to be trained (Voice called it "enrolling") so they'd know how you pronounce the phonemes, but they had huge vocabularies- you didn't have to speak every word, just enough to represent every sound. The version of Voice I had let me add words, so I could add gss_SetPixelMap & other symbols. There was a way to import a whole list of words, so I munged the output of nm & spent a couple of days training it to recognize every function, type, and variable in the PowerTV operating system.
Kurzweil Voice & Dragon Dictate had extensive "correction" features because they made a lot of mistakes. This started as very irritating, but you'd get used to it & you could talk ahead a few words, then correct the 3rd word back, then continue where you left off. When correcting, it would present a list of close-sounding words to pick from or you could spell out the word & add it if it was new.
There was even a "you keep screwing up" helper where you could give it two words that it confused a lot & re-train them.
One problem with speaking discretely (pausing between each word) is that the back of your throat takes more wear & you can start to go hoarse, causing the recognizer to get worse. Yer supposed to avoid carbonated beverages @ lunch.
Instead of running these old apps on a PDA, why not run 'em on a Crusoe subnotebook or something like the OQO?
...but what the heck does 'mere aggregation' mean in an embedded system???
...if the interfaces are public ones?
The GPL & FAQ are written assuming PCs & distros on tape or CD-ROM. On a CD containing Nvidia driver binaries and the Linux kernel, you can pull the differently-licensed pieces apart.
What about an ROM containing some GPL code in binary form and proprietary closed-source code also in binary form, but the two parts are separately linked & one dynamically loads & links the other?
The aggregation doesn't seem so "mere" anymore & is less easily reversed by end-users.
What if the GPL binary is in a separate solid-state media (NAND Flash/Smartmedia soldered to the board) and loaded by the closed code running from ROM (NOR Flash)?
What if the GPL is on removable solid-state media?
If you don't like the idea of the binaries sharing a single ROM, what about TiVo? There the binaries share a hard drive and the device is not a PC, so the aggregation is not easily reversible. TiVo provided hackers a backdoor, but was this just a bone thrown so we wouldn't look too hard at GPL compliance or does it really make the product compliant?
[I'm not talking about TiVo's mods to the kernel, which they did release source for, but all the other code (drivers and application) which are closed source, but distributed with GPL'd code.]
Inquiring minds want to know!
-M
nothing was said about broadcast flags, does this mean there wont be any? Or that it's still under debate? or did the FCC actually say "screw you" to the MPAA?
Actually, the broadcast flag was one of the other items on the agenda:
BUREAU: MEDIA
TITLE: Digital Broadcast Copy Protection
SUMMARY: The Commission will consider a Notice of Proposed Rulemaking concerning digital broadcast copy protection.
It's item #1, but the agenda was followed in exactly reverse order for some reason. A RealVideo recording of the stream is available, but will probably only be up for 6 months, so archive it now.
Any member of the public can comment on proposed rules. It's possible the stream of mail they get is quite a bit smaller than Congresspeople, so this push (by Hollings) of this issue to the FCC may actually be to cyber-activists advantage. Instructions on making a comment are on the FCC website.
is it possible for HBO et al. to broadcast Macrovision copy protection on their signal so that one cannot record such broadcasts?
On digital cable and direct-broadcast satellite, yes.
I don't know whether Macrovision can be sent over analog cable & whether it would be possible to unscramble the "premium channel" CATV scrambling while leaving (or adding in the box) Macrovision.
On digital boxes, however, extra bits in the MPEG sream or simply stuff in the program guide information can be used to turn on the Macrovision-generation stuff in the Digital ENCoder (DENC) which encodes YUV digital video into NTSC or PAL.
That is if the DENC has such a feature. I know that some chips (which are usually well-integrated: MPEG & AC-3 decode + graphics + DENC & in Broadcom's case, a complete "settop on a chip") are offered in two versions: Macrovision & not because some box builders might not want to pay for the license.
So, yes, in some cable/sat. systems (where the operator bought the right box that had the right chip), you could prevent a TiVo from being able to "back up" a show to VCR. Macrovision in cable boxes was primarily considered for Pay-per-view, though, not for premium channels. I don't know of any USA systems doing it, but I get the impression from a UK TiVo review that BskyB does.
One thought in the US was to have two pay-per-view prices-- pay more to disable macrovision. I believe that was an idea the box makers & system operators cooked up & don't know if the MPAA would go for it, but it suggests one way out of this that pisses off fewer people.
Maybe the policymakers need to distinguish between boxes like TiVo that make "ephemeral" copies and VCRs/DVD-recorders.
Xbox is already a sort of "trusted" computer.
Let's suppose someone comes up with a way to create signed binaries that it will run without a mod chip (which might be pulled from the market under the DMCA).
Are these not-created-by-Microsoft authentication tags (for xbox linux, not a "backup" of a game) a violation of Biden's bill?
The "property" with a counterfeit authentication isn't Microsoft's, it's FSF's/Linus's/et al. so maybe the bill doesn't apply. I get hung up on the definition of "illicit authentication feature" in section 3 6 C "appears to be genuine, but is not." Is it A and (B or C) or A or B or C??
The casual user will not be able to do this. One guy in a thousand will (like almost everyone who reads /.)
Michael Eisner:
"DVD has been hacked, I believe. But, but, 10%, 5% of the people hack it, they have it on their T-shirts, they have the code, you know, they have to be a genius to figure out how do it. NORMAL people just put it in and say 'I'll pay the money.'"
Eisner also decried Apple's "Rip. Mix. Burn" campaign at Hollings' hearing, but said he's "addicted" to his iPod at the Disney earnings conference call a month later.
Disney says slashdotters are abnormal.
Apple should really prepare themselves for the sheep migration by selling a Mac for under $350, w/out monitor of course.
The box should have VGA out on a HD15 connector and PS/2 mouse and keyboard jacks so you don't even have to buy a new USB keyboard.
Your old PC getting slow & you were planning to upgrade? But a new PC requires Windows SS that wants to know your shoe size before booting, so you get a pcMac & just plop it where the PC used to go. Comes with Virtual PC & the last version of Windows that only sucked mostly instead of completely (2000).
1) USB has an isochronous mode (since at least 1.1).
2) You probably don't want to use isochronous for what is essentially SCSI emulation. isoc can be fast because bad data (usually) isn't retransmitted and error checking might be done on larger chunks o' data (e.g. up to 1023 bytes vs. 64 for USB 1.1 iso vs. bulk transfers).
Think UDP vs. TCP.
isoc.'s goal isn't speed but guaranteed throughput anyway. For USB, isoc. has a lower theoretical maximum throughput.
3) Your point of looking at how the raw bandwidth is actually used is a good one, but your details are misleading.
Here's a better example: USB 1.1 bulk transfers (used for Mass Storage Class among other things) are a maximum of 64 bytes and you can send a max. of 19 of them in one frame. There are 1000 frames per second. 19 * 64 * 8 / 1000 = 9.7 megabits/s vs. a raw rate of 12 mbits/s.
USB 2.0 bulk transfers can be up to 512 bytes per packet and 13 fit in a microframe for 426.0 mbits/s vs. 480 raw (there are 8 microframes per frame).
Mass Storage class requires a 31 byte command and 13 byte answer for every transaction. The largest I can remember seeing is maybe 32 kbytes at a time, so we're only losing 1/10th of 1% to protocol overhead at that level.
I only program USB controllers (though I love firewire for DV), so I'll leave it to someone more qualified to answer the efficiency questions for it.....
------
Incidentally, USB 1.1 isoc. does 2 512 byte packets per frame for 8.2 mbits/s. USB 2.0 can do 2 3072 byte packets per microframe for 393 mbits/s. In both cases it's actually less than for bulk transfers.
You've read about them in the agreement I hope?
This won't be like region-locked DVD players that become region-free by cutting a jumper. Those can't be sold under the agreement. Waiting for the next Apex to make one without signing the agreement? Watch Congress put the force of law behind this agreement. After all, it isn't being rammed down our throat like the CPDTPA was, no, no, industries agreed (sarcasm) in private action to do this.
CSS wasn't actually that easy to crack. Robust products must be built to resist reverse-engineering so look out for tricks like the Xbox uses (you aren't running Linux on one, so don't tell me that bunnie's crack means much). Code obfuscation, secure busses, yep, it'll all be in your way.
Building your own HD receiver might end up easier. You'll have to get a tuner/QAM decoder that gives up the decoded signal in the clear, but at least open source MPEG-2 decoders are available.
Also, buy your (over-the-air) HD receivers now! They will never be encumbered by this bullshit. Satellite & cable boxes might be recalled, but they can't take away your Samsung SIR-T150 (~$600). Don't buy a combo box (they can update your firmware). Get an OTA-only one. Unencumbered HD tuner cards for PCs are another tech that will soon be out of production & $$$$ on ebay.
-M
Hollings and others want industries to agree so they can back up the agreement with force of law.
Hollings said this in the Commerce Committe hearing on March 14th. He mentioned Macrovision as something the industry came up with that Congress then required in all VCRs.
He claims to not want to have to write the technical standards that would become law but to only be using that threat to drive industries to agree.
You might or might not believe him, but Orrin Hatch who is much calmer on this issue and seems a lot more relucant to write a tech. standard seems to believe in backing a "private" agreement.
The problem is that this isn't an agreement based on mutual compromise. It's everyone caving to the MPAA.
-M
How about letting the heads of corporations (and their bootlickers in Congress) portray themselves.
I can't figure out if they are evil or stupid. It's hard to be manipulative & devious if you were shortchanged at the neuron dispenser. I'm drifting towards the opinion that it's all an act to ingratiate know-nothings who are suspicious of technology.
History of Publishing Website covers press control and censorship under Elizabeth I (1558-1603) and censorship used for military ends during the English Civil War (1642-1649).
it sounds to me that your district has already chosen a preferred representative in Congress. I'm sorry that he's not who you would have chosen.
I could be misinterpreting your implication, but it sounds like you're suggesting that the unliked candidate won fair & square & the poster has a case of sour grapes.
The candidate won in a particular system with rules that are not neutral nor the only possibility.
More parties won't help in most cases because they simply fracture the vote. We do already have a multi-party system to some extent because at least alternative parties are legal. Nader was banned from even watching the debates so there's a ways to go, but there are other alternatives:
At-large candidates. Increase the number of citizens per Representative to free up a few seats. Make those seats national so the candidates can campaign for the black vote, the geek vote or the pave-the-earth vote.
Running a national campaign for a seat that represents 1/435th of Congress might seem unwieldy and impractical, but many Congressional campaigns already receive national attention. Also narrow-cast media outlets such as issue-centric magazines, websites and discussion forums let a candidate get the word out much cheaper.
-M
Congress gave "Fair Use" via legislation
No, I think the way it worked was this:
1) The right of monopoly control over distribution (a.k.a. copyright) is in the (main body of the) Constitution
2) The First Amendment seems to contradict this since copied speech is still speech.
3) The resolution was to employ the fair use doctrine. It was originally not laid out so plainly with the four tests in chapter 12 of the US Code, but was just a doctrine (that might pre-date the Constitution for all I know) that courts used to decide infringement cases.
4) Eventually, the four tests were codified in law, but they are still vague & given to courts' interpretation.
Fair use as defined in the USC is not a right, but a defense. However, fair use derived from the first amendment is a right.
Epson printers also have an anti-refill chip and they use separate head & ink. This is even in their sub-$100 printers.
It really is a (give us all your) money issue, not a QA one, though different inks & papers really do have an affect on not just quality in general but color balance and response curve. Print Image Matching (PIM) (profiling by design rather than instance) can only work if the ink and paper are known in advance.
Epson is also the only company (of hp, Lexmark and Canon) to use piezography instead of heat to drive the ink out of the head onto the paper.
Not having replaceable heads can lead to clogging problems. One way Epson tries to solve this is by running an ink-wasting "cleaning" cycle if you don't use the printer for a while. Might as well print....
If an Epson clogs, it's off to the service center. Canon offers the best compromise: heads not replaced with ink, but new heads are user replaceable. They also have Epson-esqe print quality for photos (hp & lexmark aren't in the same league) and separate ink tanks.
Congress doesn't just pass laws. They also threaten to pass laws.
That's what happened in the PMRC/porn rock mess: with a threat of a requirement from Congress, the record labels adopted "voluntary" labels. That one was a bizarre public/private issue: If Congress had really gone ahead & mandated labels, there was a good chance it would be ruled unconstitutional, but not only did they not introduce legislation, they had the Washington Wives in the PMRC hold the cudgel over the music industry's head.
In the Commerce Committe hearing on the CBDTPA, Hollings & Stevens made repeated reference to the fact that they were trying to resolve deadlock between the content pimps union vs. the consumer electronics & computer manufacturers.
They're trying to stay two steps back from mandating copy controls. If everyone makes nice tomorrow & agrees on the Broadcast Flag, the Analog Hole, a industry-wide DRM solution and what to do about P2P, Congress can go home. If they don't get to it soon, Congress will pass the bill, giving 'em another year to get around to it. After that, Congress picks the tech.
Hollings made reference to Macrovision: the movie studios said they'd distribute movies if they had Macrovision on 'em, the VCR mfg.'s said they'd put it on the VCRs, then Congress ratified the monopoly with legislation.
The problem is that some of the issues the various industries must agree on are unsolvable. Eisner wants some way to protect movies on a screen in a theater from videotaping, divxing and p2ping (he showed a clip of Black Hawk Down downloaded from Morpheus while the film was still in theaters). He was nearly foaming at the mouth he was so worked up about these crazy hackers (he made reference to the DeCSS t-shirt!) and their evil networks (maybe Disney should buy the Internet & do for it what they did for Times Square) copying unprotected content in the "ether."
The rep from Intel, Vadasz, seemed bewildered with Eisner's requests. I think Eisner knows there's no way to protect against this kind of stuff without turning computers into VCRs (after bitching about Rip.Mix.Burn. billboards encouraging piracy, he mentioned he's "addicted" to his iPod in the Disney earnings conference call). Valenti wants "two nerds in a garage in San Diego" to figure it out if big tech industry can't.
Hollings, Eisner and Valenti sing the copyright blues.mp3