I recently put together an STB with the MyHD TV card, which has two video inputs (useful for, e.g., cable and an off-the-air antenna, since most digital cable systems fall far, far short of true HDTV performance) a fairly full-featured remote, SVGA out, a pass through cable/adapter for using a computer and a HDTV or standard monitor, component video inputs (part of the pass through adapter) and Dolby capable dts.
I'm sure I left some features out. There's a lot more info at avsforums.com - the developers (Macro Image Technologies) maintain an active presence there, and are actively upgrading capabilities
It's a bit pricey ($280 at Digital Connections) but with HDTV tuners running $500-900 Best Buy (I don't consider BB any sort of standard for quality, but they're everywhere), it's a bargain for HDTV. Even after shelling out $1500-2500 for a HDTV capable projector or monitor, you usually need a separate tuner to actually receive the HDTV signal.
BTW, don't fall for a fancy HDTV antenna. Few are any better than a $15 Radio Shack 'double bowtie' that sits on your TV (and most are worse). One good thing about HDTV is that you either get a clear signal or none at all (though you may experience drop outs at the fringe)
Before the MyHD, I had a $30 IOmagic PVR card from Outpost.com (actually a rebadged Pinnacle Studio Pro, and a great performer, if you load Pinnacle's drivers instead of IOmagic's) It even came with a $20 rebate, IIRC. For NTSC (standard TV) viewing or recording on a budget, I'd say it's hard to beat $10 after rebate.
I often recompress to a better codec (e.g. DivX 4 fast motion) and get 3 episodes of good VHS quality (or much better, if the source is HDTV) archival on a single CD. If you tape/save your favorite show, you'll definitely prefer CD/HDD over VCR tapes - CDs are cheaper (HDDs are about the same at $1/GB, and 5 eps/GB) and digital media give you better image, near-instant scanning to a desired scene (when you want to answer a question), and many other advantages. (e.g. I make parodies and music videos)
My next NTSC set-top may use the Shuttle SS50 (or a newer incarnation). it has decent audio/video built in (it doesn't use the mediocre capability of its South Bridge chip, as most do, but has discrete audio, video, and LAN chips) as well as 2 PCI slots (a newer version has 1 PCI+1AGP)
"No sure you can get IDE Raid but compared to a good SCSI raid card, IDE raid sucks."
I'm not flaming you, but I don't understand the above sentence. Since you clearly indicate you don't know about them, why do you say they suck? Did you drop a couple of words? (I've done that before, myself)
I'm not saying it doesn't suck compared to SCSI RAID -I honestly don't know- but my ATA RAID controller (vs slapping some software on fully populated mobo controllers) works as well as my SCSI units. I'm not pushing the envelope on either SCSI or IDE, so I'm hardly a technical benchmark, but I think a lot of us are driving Volvos instead of Ferraris and for us, they don't suck.
A circle is 1-D because topology makes a distinction between a 'circle' (a line figure) and the region bounded by it. The region inside the circle isn't part of the circle, any more than the region outside the circle is. Casual English uses the word 'circle' for the line figure and the region it encloses.
Similarly, if we call a basketball a 'sphere', we are discussing its 2-D surface. If we want to talk about a basketball as a 3-D object, including its internal volume, we must call it a 'ball'. There is such a thing as a 3-D sphere (3-sphere), but it is the surface of a four dimensional 'ball', which, I assure you, is like nothing you've ever seen. Many posters seem to have forgotten this today, and are speaking of 3D spheres, when they mean 2-D spheres enclosing 3-D balls.
2-D spheres have more interesting properties than 3-D balls - which you might not suspect if you think of them (as many do) as mere surfaces (i.e a part or property) of 3-D balls.
For example, rape carries a stigma in many societies, and though that stigma is (IMHO) foolish, in the US most (if not all) states have victim privacy laws which prohibit release of victim information without the victim's express (often written) permission. The details vary with jurisdiction and the exact nature of the alleged crime (e.g. adult rape vs. child rape)
Perhaps similar laws would help in computer crimes against corporate victims? One major catch, of course, is that stockholders would not be able to learn of those crimes either (only the relevant employees and management). However, the SEC and Congress have balanced similar interests before, and created workable, if not ideal, delineations.
The limiting quantity in most ISP settings is the the ISP's bandwdth. They get a fat pipe and parcel it out in small chunks of peak bandwidth and time slice i.e. much to the chagrin of many/.'ers they expect your will only be logged in (or actively using bandwidth) a small fraction of the time.
In order to support this business plan, they have to solicit an audience who wants that kind of service (again, many/.'ers choose not to understand this). Home surfers fit this bill, and they are best served by fat downlinks. Asking an ISP (or service plan) that is geared to typical home surfers to provide you with an asymetrical Fat uplink is like scraping the frosting off their cake. They probably have a service that meets your needs, and they probably charge accordingly.
Why is there no reverse ADSL? Because it would only be of interest to those operating servers and consuming sufficient bandwidth that the consumer service plan is unsatisfactory. In short, someone who is using lots of bandwidth for a long period of time.
Why would an ISP go provide a service that would lose it money , and break its business plan? That would be dot-bomb logic (Or Crazy Eddie: "We lose money on each transaction, but we make it up on the volume!")
Re:Badass compression algorithm? Nope! (legible)
on
Share The Pi!
·
· Score: 1
(oops, hit "post" instead of "preview" by accident)
Some have suggested repeatedly applying this algorithm, to generate successively shorter indices, which can "compressed" in turn by the same method, but this will not work for several reasons. Here's just one:
Let's accept the assumption that the index I of an N-bit message will typically be about N/2. N/2 is only a single binary digit shorter than N. Not much of a compression scheme at the best of times!
[To be exact:
Index(50%): k = ln(2)/ln(n/n-1) from the Poisson Eq.: [(n-1)/n]^k = 50%
This factors in the odds of hitting undesired repeats before the desired match.]
However, it is not enough to specify the index, you must also specify how many digits to read beginning at that index. So, instead of an N bit message generating (on average) an N-1 bit index as output, it would generate a pair of numbers (Index, Length), with the Index typically being N-1 bits long, and the Length typically being log(base 2, N) bits long.
Therefore, the output is typically N+log N-1 bits, which is greater than N for all N>2, The "pi index" method does not usually compress the data at all! Quite the opposite!
Re:Badass compression algorithm?
on
Share The Pi!
·
· Score: 1
Some have suggested repeatedly applying this algorithm, to generate successively shorter indices, which can "compressed" in turn by the same method, but this will not work for several reasons. Here's just one:
Let's accept the assumption that the index I of an N-bit message will typically be about N/2. N/2 is only a single binary digit shorter than N. Not much of a compression scheme at the best of times!
To be exact:
I(50%) = ln(2)/ln(n/n-1)
(from the Poisson formula [(n-1)/n]^I = 50%)
factoring a in the odds of hitting undesired repeats before the desired match
[b]However, it is not enough to specify the index, you must also specify how many digits to read beginning at that index.[/b] So, instead of an N bit message generating (on average) an N-1 bit index, it would turn an N bits input into a pair of numbers (Index, Length), with the Index typically being N-1 bits long, and the Length typically being log(base 2, N) bits long.
(N-1) + log(base 2, N) > N for all N>2. Therefore, the "pi index" method does not usually compress the data at all!
Listen, I'm the last one to justify the quality of medicine, as commonly practiced, btu your atatement reveals a glaring lack of familiarity with the art and science of either diagnostics or pharmacology -- or for that matter, the underlying eopistemologic staudies of medical informatics.
I'm not saying that thousands of significant avoidable mistakes are not made each day -- they are and for many of these, the description you gave is entirely accurate. It is inexcusable -- if only because, in my humble opinion, computerized prescription crosschecking should be mandatory, and we should have far better mechanisms for automatic, secure sharing of parient records, with adequate safeguards for privacy.
HOWEVER: Studies indicate that the average patient over the age of 67 is taking eight medications (be aware that, for purposes of drug interactions, many substances other than prescription meds can be highly significant). Also, studies have shown that somewhere between seven or eight meds, the chances of an unintended drug interaction reaches 50%
Further: Look at the literature. Though thousands of drug-drug (or class-class)interactions are known, many more are not verified (or quantified to a degree where they can be adequately weighed in clinical decision making). Worse, only a handful of 3-drug interactions are know, and almost no 4-drug and higher interactions have been documented. Finallly, we have barely begin to scratch the surface of stereo chemical racemic mixes and drug-gene interactions
How could we fully understand unintended drug-gene interactions? (i.e. interaction between a drug and some gene or gene product aside from its intended target) when we have barely mapped the outline of the human genome, and never sequenced a single individual much less the range of variation in the species. (Mapping is like drawing the outlines of the states and the major cities; sequencing is like having a complete roadmap - interpretation of the sequence is a couple of orders of magnitude beyond merely possessing the sequence, a fact we molecular biologists have successfully obscured, in our rush to get others as excited about our work as we are.) Even assuming a rate of growth akin to Moore's Law, we are still many decades from the kind of knowledge you seem to assume we have.
FURTHER, treatment often involves knowingly balancing the risks of a treatment plan against the risks of other candidate plans, and the the risks of the initial condition. It is not always easy to see these risks and effects without deep analysis and careful double blind studies.
For example, it wasn't long ago that many ICUs did not allow the adminitration of ACE inhibitors to certain types of heart failure patients, because the patients would visibly decline, and often die. There were reasons to sus pect ACE drugs might help - but any young physician who tried them soon learned the same cruel lesson. However, long term analyses (which were difficult to get authorized, given the obvious mortality) showed that 1-year survival was actually significantly increased -- those patients who immediately got sicker with ACE inhibitors would likely die in the next 6mos anyway, but after the initial shakeout many would actually do better after several days of ACEI, and overall, more patients were alive (and healthier) at 1, 3, or 5 years with the ACE drug than without.
As a physician and molecular biologist, the problems you cite frustrate me immensely, but it took me years of medical training to fully appreciate how incomplete our knowledge is, and how potentially NP-unsolvable the problem of diagnostics and therapeutics it.
That's not to say that I don't think it can be done much better. It can (and even doctors I admire could often stand some improvement). I've spent a good chunk of my life training for and working on these kinds of solutions in computing and molecular biology, as well as medicine.
I'M SORRY but your blanket indictment, though superficially similar to remarks I have made (backed by appropriate data and studies) in peer-reviewed journals, leads to precisely the wrong conclusion when the time comes to make medical policy, and I felt that I should make some effort to correct it. The factors you cite should be kept in mind, true, but they are not even remotely the entire picture. Worse, it is impossible to assess exacly *how much* of the picture they are, but I think we can safely say they are less than 50%
Every citizen's opinion counts, which makes it important that they hear the "other side". Not only do citizens help mold public policy, but more important, they are my patients, and I believe that the better background they have in their daily lives, the better equipped they will be to make the medical decisions that, in the end, are theirs, not mine, to make.
Sorry this was so long. As Descartes said:" If I had had more time, I'd have written a shorter reply."
I think that there is a simple solution that permits all the "benefits" and avoids most "liabilities" of such mandatory links on all web content (including the numerous technical problems of monitoring and responsibility for (mis)implementation)
Simply put an "other suggested reading" button on the browser that would search all government- mandated sites for the highlighted words/phrases and open a new window with the results.
If a user wants to read more, it's just a button click away. If s/he doesn't, they'd ignore a mandatory link. The code is already available. The servers would be essentially free, since every search engine would soon include a "mandated sites" search option, just to keep eyeballs
I would wait until I had a full 12 months of demonstrable prior art (e.g. a released open source program) before even considering using this system.
Why? Because publishing something does not prevent others from patenting it. If an applicant has up to a year to file the initial application for a patent (e.g. if their own notebooks or old source shows they had the major components of an idea, say, ten months before the publication, their application would be unimpeded). What is a "major component", according to this criterion? That is a little fuzzy. What is clear is that an applicant is guranteed a full year to 'develop' the idea before filing the patent.
I suspect that this database would instead be harvested by patent-hungry companies. At best, ublication of a "new" idea in this database would trigger the 'use it or lose' reflex (pushing the company to patent things they might otherwise let slide). At worst, one might see a flurry of "variant patents", since "prior art" does not enjoy the same rights of enforcement as an actual patent.
The Japanese patent system, for example, if a living demonstration of 'variant patents' run amok. This resonates with Japanese business practice, because it virtually forces quid-pro-quo "cooperation" among companies. If you patent a new type of bulb, Mitsubishi may later patent mounting it in a bayonet base, or using it for growing plants, etc. Net effect: you will have a tough time exploiting your patent, and Mitsubishi has strong leverage to force you to license to them on very favorable terms
Here's an aerial shot to illustrate how arbitrary soveregnity is. Quick - what part of the above photo is an independent nation of long and undisputed standing, and what parts are just an ordinary city?
For that matter - I can think of a dozen nations that are a single rifle shot from end to end. You'd have an interesting time explaining why this structure made of coral is an independent nation (more precisely, a disjointed major fragment of an ocean nation -- just as the Sealand photo in the parent post showed only one platform of Sealand)
National jurisidiction is arbitrary. Who will get mad if the UK invades Sealand? The UK courts, who have already ruled that the UK could not assert territoriality or jurisdiction over these platforms. (The British press would have a nice field day with the 'underdog' angle, too. It's funny how stuff like that weakens or topples administrations in parliamentary systems -- it's kind of like counting chad...)
>here *are* pre-1995 archives out there. For years Deja had a page up announcing that they had aquired an archive back to 1993 and it was going to be up RSN. It never happened
How soon they forget.
I'll admit that their page made it sound like they were acquiring something new, and maybe there's even a good reason for that, but...
DejaNews *was* the Usenet archive from roughly 1993-1995. I remember the outcry they dropped all archives prior to (Aug?) 1995 because it was getting too big (I now suspect the older archives were indexed/databased differently and they didn't feel like bringing them up to the new spec for whatever new commercial plan they had at that time
My recollection of this is clear because Aug '95 was roughly when I went into a moderate shields-up anonymity (for personal reasons) and I was amused to see my hundreds of previous DejaNews hits suddenly diminish to a handful of "whatever happened to..."
In 20+ years of emailing, I have only come across one situation where I was even tempted to send or recieve anything remotely resembling HTML e-mail (namely 'rich text' color to clarify or emphasize variables or passages buried in group code) and even then I prefer MIME attachments.
Further, my primary mail program (Eudora 4.22) is set to "send plain text only" and "don't ask". If I reply to or forward an HTML e-mail, all HTML is stripped out *before* sending (I tested it with an array of HTML spam "from the wild")
This setting should be basic netiquette (for email programs so equipped - and the feature should be *expected* for e-mail programs) Alas, in a world where even the carriage return is considered a nicety, I don't hold out much hope.
If you don't can't send javascript, the recipient can't carelessly mishandle it.
{I welcome any INFORMED, TESTED observations on the deficiencies of this method. One can never be too paranoid)
Don't forget to disable HTML viewing of e-mail (Warning: often the provided check-box alone is not sufficient), and be stingy about what programs are permitted to access the usual HTTP ports. These are virtually painless security procedures. I can't recall ever being particularly inconvenienced by them.
Am I the only one who finds that a properly selected task specific MP3 playlist can add that critical 10%+ performance boost that spells the difference between "adequate" and r-r-rippin'?
I especially value it when I am doing literary stuff ("Mood music" if you will) or need to pick up the pace for a deadline. And then there's the selection of "We kick the world's ass" anthems that I save for emergencies when you have to either go into ultra-arrogance or lapse into despair (Warning: not for use when you may have to deal with actual humans intermittently!)
On this scale, the size of usual connectors is obnoxious.
One of my current projects uses a 44-pin (laptop) IDE (with the ribbon cable peeling nicely out into ser/par/USB, but a connector which looks tiny perched on the end of a 9.5MM IDE drive is a space hog on a 50x60mm (2"x2.4") board. It's cheap, but ultimately limited for some things I'd like to do. It's only marginally better than what they used in the 70's. There seems to be an invisible wall at 1 line per mm (linear). At 0.5 l/mm most connectors use multiple rows. 0.1" (.254mm) spacing must die!
Next time, I may use serial multiplexing (USB), if I can get cheap, tiny USB chips to embed in the cables. It would dramatically simplify layout, and with USB/parallel cables running at $20-40 (street), such chips must exist.
Is there a nonproprietary (mass market) connector with much better than 1 pin/mm^2, or is there some manufacturing/durability limit at smaller spacings (e.g. the 'pocket lint' factor or ribbon cable limits)? I think I recall high end special cables in the 80's that had a grid of pins with less than 1mm spacing, in a round tabbed connector.
Re:To quote Faith No More
on
What is 'IT'?
·
· Score: 1
The only thing i can think of that fits most or all of the description is...
Women
But I'm biased. I like women. A lot.
Then again Clara Bow, once considered Hollywood's 'bad girl' was known as the "IT" girl.
I went looking for a connection to NSI (which would definitely be a worst case scenario), and didn't find one (though their info is sparse on the topic of existing partnerships and affiliations) For those who are curious:
Ron Wiener Co-Founder - Chairman & CEO
Raymond King Co-Founder - Executive Vice President & Director
(Click above names for photo and brief bio. See website link above for info on the other major execs)
Bill Lewis Chief Operating Officer
Len A. Bayles Vice President, Domain Name Industry Relations
Nelson Brady Vice President, Engineering
Jack Williamson, Jr. Chief Financial Officer
Roy Anderson, Jr. Vice President, Branding Industry Relations
More on Ron Weiner: (from a speaker list at a plenary session on aviation in Oregon.
Ron Wiener has twelve years of experience as a founder and CEO of high-tech and aviation-tech start-up companies. Currently he is "Chief Mechanic" at Venture Mechanics, LLC (Portland), a venture catalyst/incubator investing in early-stage technology companies. Previously, he was the founder and CEO of PrintBid.com, which he successfully sold to ImageX.com, Inc. (NASDAQ: IMGX) in December, 1999. PrintBid.com is a B2B e-commerce site for the commercial printing industry and commands an 80% market share in its arena. Combined, ImageX.com now employs 500 people as the industry leader. Prior to founding PrintBid, Wiener was founder and CEO of Distribution Sciences Corp./JetStream Catalog(Hillsboro, OR), a leading software publisher and catalog/internet distributor of over 1,000 products to the aviation industry.
Before moving to Oregon in 1991, Wiener was founder, Chairman and Sr. VP of Marketing of Azure Technology (San Jose, CA), a venture-backed manufacturer of avionics and flight simulation products, and catalog publisher in the aviation industry. Earlier he had served in sales, marketing, product management and product development roles with Central Point Software (Symantec), ICOT (American Airlines' SABRE), Orchid Technology, Eagle Computer and The Computer Factory retail chain.
Wiener was founding president in 1992 of the Oregon Young Entrepreneurs Assoc. (merged with OEF), and author of over 150 published technical, business and aviation articles, and the book Computers and Software in General Aviation (1987). He studied Electrical Engineering/Computer Science at Carnegie-Mellon University, and serves on the Advisory Board of several other Internet companies. He has been an active pilot since 1985 and operates his B36TC Bonanza out of Hillsboro, OR.
Some of them are even less than the $110 offered by the co-op.
I don't think that this is a fair statement. I can see why you made it, but *if* the coop arranges a second shipment at the current price (some of the members are discussing this), e-bay isn't really a competitive price alternative.
When researching E-bay prices, it is critical to search "Completed Items" [not "Active Items"] to see what the units actually sold for, since almost all the bidding takes place in the final hour, and the prices in the search results aren't updated particularly often.
At the time I write this, exactly one out of 70 Webplayers sold in the past several weeks went for under $110. (It had a crummy picture in the ad) Almost all (66/70) sold for $150-250 [one didn't sell and two sold for $120-145], so I wouldn't count on "beating the co-op price" on E-bay any time soon -- based on the observation that winning bids for I-Openers haven't dropped much in 6 months. (the lowball numbers in preceding search are mostly Iopener parts, like SODIMMs)
"Yelling fire in a crowded theater" makes a good slogan. Too good.
I wish people knew more about the case behind it -- Shenck vs. US (1919). Then maybe they'd be ashamed to use it as a rallying cry.
Schenck was only informing the public of their constitutional rights (and no one accused him of not portraying those rights accurately). he was accused of yelling firw in a crowded theatre THAT WAS ACTUALLY ON FIRE. The Justices of the time(many of whom I regard highly) wanted to avoid public tumult at any cost. Shenck spent (IIRC) over a decade in prison for simply pointing out constitutional rights, and he wasn't alone. There were several cases of 'grass roots' leaders being arrested for this. I believe even the Pulitzer Prize winning journalist Upton Sinclair was arrested -- for reading the text of the Constitution to a lawfully asembled crowd. The Vietnam anti-war protests (or Slashdot) could easily be shut down under both the spirit and the letter of Schenk -- if it weren't considered, even within the legal community a dangerous and even bad precedent.
Remember, there are still plenty of places, in and out of the US where peace and order are considered more important than truth or justice. Not in your town? Oh yes - check your local high schools, for example (I have a kid in HS, just for the record). It's a basic human instinct going back to the monkeys
Actually, you've put your finger on an important point -- in the case you cited, the spammer wasn't getting a response, and wasn't likely to for very long (if at all). A trojan may get you SMTP access (or an account password), but once complaints start arriving in the POP box a few hours later, the user will call customer support, change their password, and/or run that virus scan they have been ignoring for months.
So aside from the common sense of the recipient, there is a real question how would-be suckers can actually learn where send their hard-earned money when spammers are so elusive. Web pages are one answer, but In the weeks after that Salon article, I tried a few times, and found that the spammer Web sites were usually either dead or empty.
Much (but by no means all) apparent spam is actually meant to harrass the unwitting "sender", either as a vendetta or a blind script-kiddie prank. In managing several domains, I have seen a lot of spam that would never be commercially viable because the sender could never, even in theory, get a reply.
Of course, if there are only a million idiots out there, and each of them periodically decided to 'give it one more try' with that software they wasted $35 on... that adds up to many billions of spam a year, and that's still just a drop in the bucket.
As far as I have heard, the US Supreme Court merely 'vacated' the Florida Supreme Court ruling, which is not the same as overruling them.
In effect, they set aside the verdict, and told the Florida court to reconsider the case - which may be done with, or without, further hearings. In essence, the Florida court did not give enough information in their ruling for the Supreme Court to fully discern their intent and consideration of various legal ussues (which we've all been hearing about ad nauseum)
The Supreme Court is hesitant to step on the toes of the state supreme courts, and since it was unable to discern the 'quality' and 'correctness' of the legal thinking behind the ruling, they merely sent it back for clarification, in effect.
In the early days, NACA (the predecessor to NASA) was mostly made od can-do types, rather than academics. However, when NASA really got in the public eye (and got glamorous), there was a big push to get their people graduate degrees --any graduate degrees.
A friend of mine's father was big in the space suit design effort, but had little formal education. He told us the 'degree' NASA helped him get was basically a diploma-mill deal with lots of 'life experience' credit. He never put it on his resume, and (perhaps because of being forced to get it) was fond of saying "not bad for a guy who never finished high school" whenever he did something well.
I've seen people do ASCII carbon stick figures of sugars on/. -- but I'm not that ambitious, so do a search. Basically, you can think of a sugar as being a chain of carbons with a H sticking off one side of each carbon and a OH sticking off the other (HOH = water, hence "carbohydrates")
Take the carbon chain and bend it into a ring (you may have one carbon sticking out, and voila! you have a sugar. The different sugars differ *only* in
a) the number of carbons (typically 3-7, but longer would also count)
b) the exact way the end carbons are bound (this is where the stick figure helps)
c) if the OH's are above or below the ring (as biochemists like to say - up-down-up is different then down-up-down or up-up-up)
d) some common sugars (like sucrose = table sugar= alpha-glucosido-beta-D-fructofuranoside) are actually multiple simple sugars tacked together. In the case of table suger, it's a glucose ("blood sugar" aka dextrose) molecule bound to a fructose ("fruit sugar")
That may demystify the 'assymetry' we're talking about.
Secondly, there are most definitely physicochemical non-biological processes that produce assymetric molecules. Your argument goes way back , but it's pretty old and faded today. Indeed, they are widely used in industry, especially the pharmaceutical industry (to make only one 'handedness', which are more effective with fewer side effects) While originally biological byproducts were used to create that handedness, today we can do it from scratch.
How could nature produce such compounds? Here's a simple scenario: many chiral compunds prefer to form crystals with their "own kind" (i.e. a recemic mixture carefullt crystalized, will produce D (right handed) and L (left handed) crystals. So somewhere, some time, primeval goo could slowly have crystalized into D and L crystals in some drying puddle.
Once you have a naturally made D or L crystal, the rest of molecular evolution to 'organic biology' can proceed as we expect. Perhaps there is an advantage to D or L, or maybe it's just that an organism that belongs to the (accidental) majority simply finds more appropriately-oriented "food" or "prey" that it can more easily process. It doesn't really matter which -- though there are some interesting arguments that there are advantages to the overall biosphere to settle on (for example) L-amino acids and D-sugars rather than a random mix of both.
Slashdot Mirror
I recently put together an STB with the MyHD TV card, which has two video inputs (useful for, e.g., cable and an off-the-air antenna, since most digital cable systems fall far, far short of true HDTV performance) a fairly full-featured remote, SVGA out, a pass through cable/adapter for using a computer and a HDTV or standard monitor, component video inputs (part of the pass through adapter) and Dolby capable dts.
I'm sure I left some features out. There's a lot more info at avsforums.com - the developers (Macro Image Technologies) maintain an active presence there, and are actively upgrading capabilities
It's a bit pricey ($280 at Digital Connections) but with HDTV tuners running $500-900 Best Buy (I don't consider BB any sort of standard for quality, but they're everywhere), it's a bargain for HDTV. Even after shelling out $1500-2500 for a HDTV capable projector or monitor, you usually need a separate tuner to actually receive the HDTV signal.
BTW, don't fall for a fancy HDTV antenna. Few are any better than a $15 Radio Shack 'double bowtie' that sits on your TV (and most are worse). One good thing about HDTV is that you either get a clear signal or none at all (though you may experience drop outs at the fringe)
Before the MyHD, I had a $30 IOmagic PVR card from Outpost.com (actually a rebadged Pinnacle Studio Pro, and a great performer, if you load Pinnacle's drivers instead of IOmagic's) It even came with a $20 rebate, IIRC. For NTSC (standard TV) viewing or recording on a budget, I'd say it's hard to beat $10 after rebate.
I often recompress to a better codec (e.g. DivX 4 fast motion) and get 3 episodes of good VHS quality (or much better, if the source is HDTV) archival on a single CD. If you tape/save your favorite show, you'll definitely prefer CD/HDD over VCR tapes - CDs are cheaper (HDDs are about the same at $1/GB, and 5 eps/GB) and digital media give you better image, near-instant scanning to a desired scene (when you want to answer a question), and many other advantages. (e.g. I make parodies and music videos)
My next NTSC set-top may use the Shuttle SS50 (or a newer incarnation). it has decent audio/video built in (it doesn't use the mediocre capability of its South Bridge chip, as most do, but has discrete audio, video, and LAN chips) as well as 2 PCI slots (a newer version has 1 PCI+1AGP)
"No sure you can get IDE Raid but compared to a good SCSI raid card, IDE raid sucks."
I'm not flaming you, but I don't understand the above sentence. Since you clearly indicate you don't know about them, why do you say they suck? Did you drop a couple of words? (I've done that before, myself)
I'm not saying it doesn't suck compared to SCSI RAID -I honestly don't know- but my ATA RAID controller (vs slapping some software on fully populated mobo controllers) works as well as my SCSI units. I'm not pushing the envelope on either SCSI or IDE, so I'm hardly a technical benchmark, but I think a lot of us are driving Volvos instead of Ferraris and for us, they don't suck.
A circle is 1-D because topology makes a distinction between a 'circle' (a line figure) and the region bounded by it. The region inside the circle isn't part of the circle, any more than the region outside the circle is. Casual English uses the word 'circle' for the line figure and the region it encloses.
Similarly, if we call a basketball a 'sphere', we are discussing its 2-D surface. If we want to talk about a basketball as a 3-D object, including its internal volume, we must call it a 'ball'. There is such a thing as a 3-D sphere (3-sphere), but it is the surface of a four dimensional 'ball', which, I assure you, is like nothing you've ever seen. Many posters seem to have forgotten this today, and are speaking of 3D spheres, when they mean 2-D spheres enclosing 3-D balls.
2-D spheres have more interesting properties than 3-D balls - which you might not suspect if you think of them (as many do) as mere surfaces (i.e a part or property) of 3-D balls.
This problem is not unique to computer crime.
For example, rape carries a stigma in many societies, and though that stigma is (IMHO) foolish, in the US most (if not all) states have victim privacy laws which prohibit release of victim information without the victim's express (often written) permission. The details vary with jurisdiction and the exact nature of the alleged crime (e.g. adult rape vs. child rape)
Perhaps similar laws would help in computer crimes against corporate victims? One major catch, of course, is that stockholders would not be able to learn of those crimes either (only the relevant employees and management). However, the SEC and Congress have balanced similar interests before, and created workable, if not ideal, delineations.
The limiting quantity in most ISP settings is the the ISP's bandwdth. They get a fat pipe and parcel it out in small chunks of peak bandwidth and time slice i.e. much to the chagrin of many /.'ers they expect your will only be logged in (or actively using bandwidth) a small fraction of the time.
/.'ers choose not to understand this). Home surfers fit this bill, and they are best served by fat downlinks. Asking an ISP (or service plan) that is geared to typical home surfers to provide you with an asymetrical Fat uplink is like scraping the frosting off their cake. They probably have a service that meets your needs, and they probably charge accordingly.
In order to support this business plan, they have to solicit an audience who wants that kind of service (again, many
Why is there no reverse ADSL? Because it would only be of interest to those operating servers and consuming sufficient bandwidth that the consumer service plan is unsatisfactory. In short, someone who is using lots of bandwidth for a long period of time.
Why would an ISP go provide a service that would lose it money , and break its business plan? That would be dot-bomb logic (Or Crazy Eddie: "We lose money on each transaction, but we make it up on the volume!")
Some have suggested repeatedly applying this algorithm, to generate successively shorter indices, which can "compressed" in turn by the same method, but this will not work for several reasons. Here's just one:
Let's accept the assumption that the index I of an N-bit message will typically be about N/2. N/2 is only a single binary digit shorter than N. Not much of a compression scheme at the best of times!
[To be exact:
Index(50%): k = ln(2)/ln(n/n-1) from the Poisson Eq.: [(n-1)/n]^k = 50%
This factors in the odds of hitting undesired repeats before the desired match.]
However, it is not enough to specify the index, you must also specify how many digits to read beginning at that index. So, instead of an N bit message generating (on average) an N-1 bit index as output, it would generate a pair of numbers (Index, Length), with the Index typically being N-1 bits long, and the Length typically being log(base 2, N) bits long.
Therefore, the output is typically N+log N-1 bits, which is greater than N for all N>2, The "pi index" method does not usually compress the data at all! Quite the opposite!
Some have suggested repeatedly applying this algorithm, to generate successively shorter indices, which can "compressed" in turn by the same method, but this will not work for several reasons. Here's just one: Let's accept the assumption that the index I of an N-bit message will typically be about N/2. N/2 is only a single binary digit shorter than N. Not much of a compression scheme at the best of times! To be exact: I(50%) = ln(2)/ln(n/n-1) (from the Poisson formula [(n-1)/n]^I = 50%) factoring a in the odds of hitting undesired repeats before the desired match [b]However, it is not enough to specify the index, you must also specify how many digits to read beginning at that index.[/b] So, instead of an N bit message generating (on average) an N-1 bit index, it would turn an N bits input into a pair of numbers (Index, Length), with the Index typically being N-1 bits long, and the Length typically being log(base 2, N) bits long. (N-1) + log(base 2, N) > N for all N>2. Therefore, the "pi index" method does not usually compress the data at all!
Listen, I'm the last one to justify the quality of medicine, as commonly practiced, btu your atatement reveals a glaring lack of familiarity with the art and science of either diagnostics or pharmacology -- or for that matter, the underlying eopistemologic staudies of medical informatics.
I'm not saying that thousands of significant avoidable mistakes are not made each day -- they are and for many of these, the description you gave is entirely accurate. It is inexcusable -- if only because, in my humble opinion, computerized prescription crosschecking should be mandatory, and we should have far better mechanisms for automatic, secure sharing of parient records, with adequate safeguards for privacy.
HOWEVER: Studies indicate that the average patient over the age of 67 is taking eight medications (be aware that, for purposes of drug interactions, many substances other than prescription meds can be highly significant). Also, studies have shown that somewhere between seven or eight meds, the chances of an unintended drug interaction reaches 50%
Further: Look at the literature. Though thousands of drug-drug (or class-class)interactions are known, many more are not verified (or quantified to a degree where they can be adequately weighed in clinical decision making). Worse, only a handful of 3-drug interactions are know, and almost no 4-drug and higher interactions have been documented. Finallly, we have barely begin to scratch the surface of stereo chemical racemic mixes and drug-gene interactions
How could we fully understand unintended drug-gene interactions? (i.e. interaction between a drug and some gene or gene product aside from its intended target) when we have barely mapped the outline of the human genome, and never sequenced a single individual much less the range of variation in the species. (Mapping is like drawing the outlines of the states and the major cities; sequencing is like having a complete roadmap - interpretation of the sequence is a couple of orders of magnitude beyond merely possessing the sequence, a fact we molecular biologists have successfully obscured, in our rush to get others as excited about our work as we are.) Even assuming a rate of growth akin to Moore's Law, we are still many decades from the kind of knowledge you seem to assume we have.
FURTHER, treatment often involves knowingly balancing the risks of a treatment plan against the risks of other candidate plans, and the the risks of the initial condition. It is not always easy to see these risks and effects without deep analysis and careful double blind studies.
For example, it wasn't long ago that many ICUs did not allow the adminitration of ACE inhibitors to certain types of heart failure patients, because the patients would visibly decline, and often die. There were reasons to sus pect ACE drugs might help - but any young physician who tried them soon learned the same cruel lesson. However, long term analyses (which were difficult to get authorized, given the obvious mortality) showed that 1-year survival was actually significantly increased -- those patients who immediately got sicker with ACE inhibitors would likely die in the next 6mos anyway, but after the initial shakeout many would actually do better after several days of ACEI, and overall, more patients were alive (and healthier) at 1, 3, or 5 years with the ACE drug than without.
As a physician and molecular biologist, the problems you cite frustrate me immensely, but it took me years of medical training to fully appreciate how incomplete our knowledge is, and how potentially NP-unsolvable the problem of diagnostics and therapeutics it.
That's not to say that I don't think it can be done much better. It can (and even doctors I admire could often stand some improvement). I've spent a good chunk of my life training for and working on these kinds of solutions in computing and molecular biology, as well as medicine.
I'M SORRY but your blanket indictment, though superficially similar to remarks I have made (backed by appropriate data and studies) in peer-reviewed journals, leads to precisely the wrong conclusion when the time comes to make medical policy, and I felt that I should make some effort to correct it. The factors you cite should be kept in mind, true, but they are not even remotely the entire picture. Worse, it is impossible to assess exacly *how much* of the picture they are, but I think we can safely say they are less than 50%
Every citizen's opinion counts, which makes it important that they hear the "other side". Not only do citizens help mold public policy, but more important, they are my patients, and I believe that the better background they have in their daily lives, the better equipped they will be to make the medical decisions that, in the end, are theirs, not mine, to make.
Sorry this was so long. As Descartes said:" If I had had more time, I'd have written a shorter reply."
FreeBSD really rocks,
But only Intel / Alpha box.
The 'Net' one runs on so much more.
The 'Open' one is more secure.
Now I know my BSD's. Won't you come and play with me?
(written in years gone by, when I had a different ID)
I think that there is a simple solution that permits all the "benefits" and avoids most "liabilities" of such mandatory links on all web content (including the numerous technical problems of monitoring and responsibility for (mis)implementation)
Simply put an "other suggested reading" button on the browser that would search all government- mandated sites for the highlighted words/phrases and open a new window with the results.
If a user wants to read more, it's just a button click away. If s/he doesn't, they'd ignore a mandatory link. The code is already available. The servers would be essentially free, since every search engine would soon include a "mandated sites" search option, just to keep eyeballs
I would wait until I had a full 12 months of demonstrable prior art (e.g. a released open source program) before even considering using this system.
Why? Because publishing something does not prevent others from patenting it. If an applicant has up to a year to file the initial application for a patent (e.g. if their own notebooks or old source shows they had the major components of an idea, say, ten months before the publication, their application would be unimpeded). What is a "major component", according to this criterion? That is a little fuzzy. What is clear is that an applicant is guranteed a full year to 'develop' the idea before filing the patent.
I suspect that this database would instead be harvested by patent-hungry companies. At best, ublication of a "new" idea in this database would trigger the 'use it or lose' reflex (pushing the company to patent things they might otherwise let slide). At worst, one might see a flurry of "variant patents", since "prior art" does not enjoy the same rights of enforcement as an actual patent.
The Japanese patent system, for example, if a living demonstration of 'variant patents' run amok. This resonates with Japanese business practice, because it virtually forces quid-pro-quo "cooperation" among companies. If you patent a new type of bulb, Mitsubishi may later patent mounting it in a bayonet base, or using it for growing plants, etc. Net effect: you will have a tough time exploiting your patent, and Mitsubishi has strong leverage to force you to license to them on very favorable terms
Here's an aerial shot to illustrate how arbitrary soveregnity is. Quick - what part of the above photo is an independent nation of long and undisputed standing, and what parts are just an ordinary city?
Okay, let's try another country. The narrow strip of coastal buildings in this photo is a centuries old (far older than Germany, to cite one example) nation of undisputed sovereignity.
For that matter - I can think of a dozen nations that are a single rifle shot from end to end. You'd have an interesting time explaining why this structure made of coral is an independent nation (more precisely, a disjointed major fragment of an ocean nation -- just as the Sealand photo in the parent post showed only one platform of Sealand)
National jurisidiction is arbitrary. Who will get mad if the UK invades Sealand? The UK courts , who have already ruled that the UK could not assert territoriality or jurisdiction over these platforms. (The British press would have a nice field day with the 'underdog' angle, too. It's funny how stuff like that weakens or topples administrations in parliamentary systems -- it's kind of like counting chad...)
>here *are* pre-1995 archives out there. For years Deja had a page up announcing that they had aquired an archive back to 1993 and it was going to be up RSN. It never happened
How soon they forget.
I'll admit that their page made it sound like they were acquiring something new, and maybe there's even a good reason for that, but...
DejaNews *was* the Usenet archive from roughly 1993-1995. I remember the outcry they dropped all archives prior to (Aug?) 1995 because it was getting too big (I now suspect the older archives were indexed/databased differently and they didn't feel like bringing them up to the new spec for whatever new commercial plan they had at that time
My recollection of this is clear because Aug '95 was roughly when I went into a moderate shields-up anonymity (for personal reasons) and I was amused to see my hundreds of previous DejaNews hits suddenly diminish to a handful of "whatever happened to..."
In 20+ years of emailing, I have only come across one situation where I was even tempted to send or recieve anything remotely resembling HTML e-mail (namely 'rich text' color to clarify or emphasize variables or passages buried in group code) and even then I prefer MIME attachments.
Further, my primary mail program (Eudora 4.22) is set to "send plain text only" and "don't ask". If I reply to or forward an HTML e-mail, all HTML is stripped out *before* sending (I tested it with an array of HTML spam "from the wild")
This setting should be basic netiquette (for email programs so equipped - and the feature should be *expected* for e-mail programs) Alas, in a world where even the carriage return is considered a nicety, I don't hold out much hope.
If you don't can't send javascript, the recipient can't carelessly mishandle it.
{I welcome any INFORMED, TESTED observations on the deficiencies of this method. One can never be too paranoid)
Don't forget to disable HTML viewing of e-mail (Warning: often the provided check-box alone is not sufficient), and be stingy about what programs are permitted to access the usual HTTP ports. These are virtually painless security procedures. I can't recall ever being particularly inconvenienced by them.
Am I the only one who finds that a properly selected task specific MP3 playlist can add that critical 10%+ performance boost that spells the difference between "adequate" and r-r-rippin'?
I especially value it when I am doing literary stuff ("Mood music" if you will) or need to pick up the pace for a deadline. And then there's the selection of "We kick the world's ass" anthems that I save for emergencies when you have to either go into ultra-arrogance or lapse into despair (Warning: not for use when you may have to deal with actual humans intermittently!)
On this scale, the size of usual connectors is obnoxious.
One of my current projects uses a 44-pin (laptop) IDE (with the ribbon cable peeling nicely out into ser/par/USB, but a connector which looks tiny perched on the end of a 9.5MM IDE drive is a space hog on a 50x60mm (2"x2.4") board. It's cheap, but ultimately limited for some things I'd like to do. It's only marginally better than what they used in the 70's. There seems to be an invisible wall at 1 line per mm (linear). At 0.5 l/mm most connectors use multiple rows. 0.1" (.254mm) spacing must die!
Next time, I may use serial multiplexing (USB), if I can get cheap, tiny USB chips to embed in the cables. It would dramatically simplify layout, and with USB/parallel cables running at $20-40 (street), such chips must exist.
Is there a nonproprietary (mass market) connector with much better than 1 pin/mm^2, or is there some manufacturing/durability limit at smaller spacings (e.g. the 'pocket lint' factor or ribbon cable limits)? I think I recall high end special cables in the 80's that had a grid of pins with less than 1mm spacing, in a round tabbed connector.
The only thing i can think of that fits most or all of the description is...
Women
But I'm biased. I like women. A lot.
Then again Clara Bow, once considered Hollywood's 'bad girl' was known as the "IT" girl.
Clara Bow:
Servicing the USC football team?
She looks like a girl I knew in high school. I wasn't impressed then either.
I went looking for a connection to NSI (which would definitely be a worst case scenario), and didn't find one (though their info is sparse on the topic of existing partnerships and affiliations) For those who are curious:
Per the SnapNames website:
Ron Wiener Co-Founder - Chairman & CEO
Raymond King Co-Founder - Executive Vice President & Director
(Click above names for photo and brief bio. See website link above for info on the other major execs)
Bill Lewis Chief Operating Officer
Len A. Bayles Vice President, Domain Name Industry Relations
Nelson Brady Vice President, Engineering
Jack Williamson, Jr. Chief Financial Officer
Roy Anderson, Jr. Vice President, Branding Industry Relations
More on Ron Weiner:
(from a speaker list at a plenary session on aviation in Oregon.
Ron Wiener has twelve years of experience as a founder and CEO of high-tech and aviation-tech start-up companies. Currently he is "Chief Mechanic" at Venture Mechanics, LLC (Portland), a venture catalyst/incubator investing in early-stage technology companies. Previously, he was the founder and CEO of PrintBid.com, which he successfully sold to ImageX.com, Inc. (NASDAQ: IMGX) in December, 1999. PrintBid.com is a B2B e-commerce site for the commercial printing industry and commands an 80% market share in its arena. Combined, ImageX.com now employs 500 people as the industry leader. Prior to founding PrintBid, Wiener was founder and CEO of Distribution Sciences Corp./JetStream Catalog(Hillsboro, OR), a leading software publisher and catalog/internet distributor of over 1,000 products to the aviation industry.
Before moving to Oregon in 1991, Wiener was founder, Chairman and Sr. VP of Marketing of Azure Technology (San Jose, CA), a venture-backed manufacturer of avionics and flight simulation products, and catalog publisher in the aviation industry. Earlier he had served in sales, marketing, product management and product development roles with Central Point Software (Symantec), ICOT (American Airlines' SABRE), Orchid Technology, Eagle Computer and The Computer Factory retail chain.
Wiener was founding president in 1992 of the Oregon Young Entrepreneurs Assoc. (merged with OEF), and author of over 150 published technical, business and aviation articles, and the book Computers and Software in General Aviation (1987). He studied Electrical Engineering/Computer Science at Carnegie-Mellon University, and serves on the Advisory Board of several other Internet companies. He has been an active pilot since 1985 and operates his B36TC Bonanza out of Hillsboro, OR.
But at least they'll leave behind a nice building.
Ah yes.
Another case of "Live fast. Die Young. Leave a good-looking corpse"
Some of them are even less than the $110 offered by the co-op.
I don't think that this is a fair statement. I can see why you made it, but *if* the coop arranges a second shipment at the current price (some of the members are discussing this), e-bay isn't really a competitive price alternative.
When researching E-bay prices, it is critical to search "Completed Items" [not "Active Items"] to see what the units actually sold for, since almost all the bidding takes place in the final hour, and the prices in the search results aren't updated particularly often.
At the time I write this, exactly one out of 70 Webplayers sold in the past several weeks went for under $110. (It had a crummy picture in the ad) Almost all (66/70) sold for $150-250 [one didn't sell and two sold for $120-145], so I wouldn't count on "beating the co-op price" on E-bay any time soon -- based on the observation that winning bids for I-Openers haven't dropped much in 6 months. (the lowball numbers in preceding search are mostly Iopener parts, like SODIMMs)
I have absolutely no stake in the Co-op.
"Yelling fire in a crowded theater" makes a good slogan. Too good.
I wish people knew more about the case behind it -- Shenck vs. US (1919). Then maybe they'd be ashamed to use it as a rallying cry.
Schenck was only informing the public of their constitutional rights (and no one accused him of not portraying those rights accurately). he was accused of yelling firw in a crowded theatre THAT WAS ACTUALLY ON FIRE. The Justices of the time(many of whom I regard highly) wanted to avoid public tumult at any cost. Shenck spent (IIRC) over a decade in prison for simply pointing out constitutional rights, and he wasn't alone. There were several cases of 'grass roots' leaders being arrested for this. I believe even the Pulitzer Prize winning journalist Upton Sinclair was arrested -- for reading the text of the Constitution to a lawfully asembled crowd. The Vietnam anti-war protests (or Slashdot) could easily be shut down under both the spirit and the letter of Schenk -- if it weren't considered, even within the legal community a dangerous and even bad precedent.
Here's a readable summary of Shenck and many other classic precedents involving the First Ameendment topics we see on Slashdot -- and for completeness and accuracy, you can check the actual ruling in Schenck, too -- no one is slanting the facts. The truth actually is that disgraceful
Remember, there are still plenty of places, in and out of the US where peace and order are considered more important than truth or justice. Not in your town? Oh yes - check your local high schools, for example (I have a kid in HS, just for the record). It's a basic human instinct going back to the monkeys
Actually, you've put your finger on an important point -- in the case you cited, the spammer wasn't getting a response, and wasn't likely to for very long (if at all). A trojan may get you SMTP access (or an account password), but once complaints start arriving in the POP box a few hours later, the user will call customer support, change their password, and/or run that virus scan they have been ignoring for months.
This is a general problem. Salon did an article reporting their miserable experiences in trying to respond to spam. Basically, they couldn't get through to anyone!
So aside from the common sense of the recipient, there is a real question how would-be suckers can actually learn where send their hard-earned money when spammers are so elusive. Web pages are one answer, but In the weeks after that Salon article, I tried a few times, and found that the spammer Web sites were usually either dead or empty.
Much (but by no means all) apparent spam is actually meant to harrass the unwitting "sender", either as a vendetta or a blind script-kiddie prank. In managing several domains, I have seen a lot of spam that would never be commercially viable because the sender could never, even in theory, get a reply.
Of course, if there are only a million idiots out there, and each of them periodically decided to 'give it one more try' with that software they wasted $35 on... that adds up to many billions of spam a year, and that's still just a drop in the bucket.
As far as I have heard, the US Supreme Court merely 'vacated' the Florida Supreme Court ruling, which is not the same as overruling them.
In effect, they set aside the verdict, and told the Florida court to reconsider the case - which may be done with, or without, further hearings. In essence, the Florida court did not give enough information in their ruling for the Supreme Court to fully discern their intent and consideration of various legal ussues (which we've all been hearing about ad nauseum)
The Supreme Court is hesitant to step on the toes of the state supreme courts, and since it was unable to discern the 'quality' and 'correctness' of the legal thinking behind the ruling, they merely sent it back for clarification, in effect.
Actually, this is highly ironic...
In the early days, NACA (the predecessor to NASA) was mostly made od can-do types, rather than academics. However, when NASA really got in the public eye (and got glamorous), there was a big push to get their people graduate degrees --any graduate degrees.
A friend of mine's father was big in the space suit design effort, but had little formal education. He told us the 'degree' NASA helped him get was basically a diploma-mill deal with lots of 'life experience' credit. He never put it on his resume, and (perhaps because of being forced to get it) was fond of saying "not bad for a guy who never finished high school" whenever he did something well.
Okay, let's simplify this before we debunk it.
/. -- but I'm not that ambitious, so do a search. Basically, you can think of a sugar as being a chain of carbons with a H sticking off one side of each carbon and a OH sticking off the other (HOH = water, hence "carbohydrates")
I've seen people do ASCII carbon stick figures of sugars on
Take the carbon chain and bend it into a ring (you may have one carbon sticking out, and voila! you have a sugar. The different sugars differ *only* in
a) the number of carbons (typically 3-7, but longer would also count)
b) the exact way the end carbons are bound (this is where the stick figure helps)
c) if the OH's are above or below the ring (as biochemists like to say - up-down-up is different then down-up-down or up-up-up)
d) some common sugars (like sucrose = table sugar= alpha-glucosido-beta-D-fructofuranoside) are actually multiple simple sugars tacked together. In the case of table suger, it's a glucose ("blood sugar" aka dextrose) molecule bound to a fructose ("fruit sugar")
That may demystify the 'assymetry' we're talking about.
Secondly, there are most definitely physicochemical non-biological processes that produce assymetric molecules. Your argument goes way back , but it's pretty old and faded today. Indeed, they are widely used in industry, especially the pharmaceutical industry (to make only one 'handedness', which are more effective with fewer side effects) While originally biological byproducts were used to create that handedness, today we can do it from scratch.
How could nature produce such compounds? Here's a simple scenario: many chiral compunds prefer to form crystals with their "own kind" (i.e. a recemic mixture carefullt crystalized, will produce D (right handed) and L (left handed) crystals. So somewhere, some time, primeval goo could slowly have crystalized into D and L crystals in some drying puddle.
Once you have a naturally made D or L crystal, the rest of molecular evolution to 'organic biology' can proceed as we expect. Perhaps there is an advantage to D or L, or maybe it's just that an organism that belongs to the (accidental) majority simply finds more appropriately-oriented "food" or "prey" that it can more easily process. It doesn't really matter which -- though there are some interesting arguments that there are advantages to the overall biosphere to settle on (for example) L-amino acids and D-sugars rather than a random mix of both.