Actually, the basic technology here is not at all new. These are what is know as "McKibben Muscles", first developed for artificial arms in the 1950s. (The link is to a good overview of them from my robotics bookmarks list.)
They are cheap, strong, compact, and easy to build, but they require a lot of energy in the form of not-too-easily available compressed air or other gas. For this reason, they're not used too frequently for prosthetics since carrying enough gas becomes a problem.
These have been used to build a number of camplex life-mimicking robots, including many of the better legged walking robots. Check out the this page for an idea of how you can use them in place of real muscles to achieve very lifelike results.
If you've got a good source of compressed gas, these are an excellent design choice, and probably the only thing we have that can fairly approximate natural muscle at a reasonable cost. ("Muscle wires" like Nitinol require gobs of power and need way too much cooling time to contract to be useful in most applications.)
I'm not a DB expert by any means, but here's what I do know about Intersystems' Cache - it can be faster than anything else in certain real-world applications:
I used to be Market Segment Manager for healthcare at Sun Microsystems. This is back when Sun was a "who?" in healthcare IT (outside of medical imaging, where we were dominant, but buried in OEM gear behind the name of a Seimens, GE, etc.)
When the Ultras came out, we wanted to aggressively go after some of the spaces where we would be a good fit. One of those was electronic medical records and managed care systems, both of which tend to be large, sparsely populated databases.
We worked closely with a clinical system vendor (which has an excellent product and has moved from a relatively unknown newcomer to one of the most favored products in the industry) and helped them face off not only Sun gear vs. IBM, HP, and DEC, but also to find the best DB to work with the hardware environment.
According to thier own tests, this user reported the following: Oracle and DB2 were essentially neck and neck, but Oracle offered better network support, was cheaper, and was considerably easier to do business with than IBM. They were stunned to find that for thier application, Cache was THIRTY times faster than Oracle out-of-the-box. It was still nearly 10X faster even after doing "unnatural things to Oracle that you wouldn't want to live with."
Similar results were reported a few months later by another ISV that built the first large-scale managed care enrollment system that was proven capable of scaling to millions of members.
I'm also somewhat familiar with what was for a long time the showcase site for Intersystems: Boston Children's Hospital and the Harvard Medical Center. This system has grown along with the hospital server-by-server. In 1995-ish, the clinical system there was based on around 150 PC servers scattered throughout the facility. Intersystems Distributed Cache Protocol keeps the data on the move so it tends to reside neares its point of most frequent use automatically. I don't know of anything other DB that provides this sort of modular scalability and caching performance that's so important in the real world., although there may be something else similar.
I'm impressed because people I know that really know this stuff are impressed, again, I make no claim to being a database guru myself.
I want views!!! Precious views!! I need to secure certain data from the prying eyes of account managers and sales folk.
Sounds to me like you're really trying to hammer that socket-head capscrew into place, rather than finding an appropriate Allen wrench.
It's amazing to me how many people in the open source community continually try to force mySQL into doing jobs it's clearly not up to. This is especially puzzling, since there *are* good, modern, high-performance, high-function databases out there, things like Postgresql, Interbase,and the Sleepycat Berkeley DB (not SQL, but really fast and solid) among many others.
For quick, dirty, and fast, without the hassles of dealing with SQL at all, there are options like NoSQL and Starbase, both of which are based on the Stream/Modifier concept of the original/rdb that allows fast, flexible database operations and allowing the use of the full power of the Unix environment. ( It's easy, leverages a zillion other existing tools, and everything you learn is directly applicable to anything else you do in a Unix environment.)
It seems most people want SQL just because it somehow legitimizes their project as "serious", when in reality, it often just adds undesirable complexity and support cost. SQL is often unnecessary and actually undesirable, if you allow yourself to consider the possible benefits of NOT using a SQL database.
Also, keep in mind that *any* database (and especially ones that are already very fast,like BerkeleyDB or Starbase) is lighting fast in RAM, and memory is now cheap enough to make putting the whole DB in memory a very do-able thing: A gigabyte or several of database is BIG, and easily supportable on very affordable server hardware.
And of course, there is a reason those commercial database companies exist: They often offer capabilities that open source users may never have. I am looking right now at a new application which will be *much* faster and cheaper based on Intersystems' unique Cache database than it eould be on any of the more common big names. It's important to choose the right tool for the job. (In this case, I need lots of transactions in a very large sparsely populated database, Cache is excellent there, an order of magnitude better than Oracle or DB2, which is why Cache is the leading DB choice for electronic medical records systems, and has been for years.)
MySQL is not the answer to every database problem. Or even most of them, truth be told. Personally, I find it to be a minimally competent, fussy, and underpowered tool. It's good for some things, but to be honest, I've always thought its appeal had more to do with being the first real open source SQL DB than with being good.
As always, this boils down to choosing the right tool for the job. Not exactly rocket science, but something way too many people don't bother to do.
As the architect of a high-performance multi-terabyte storage subsystem based on IDE disk array technology, I beg to differ about Maxtor's quality. Just this past fall, we spent a long time evaluating which drives were the best. The general consensus (and our own conclusion) is Maxtor's large drives run cooler and last longer (failure rates start to become an important item when you have several hundred such drives in an array.)
IBM drives were better at one time, but that's not been the case for a while, and I doubt Hitachi can change that anytime soon.
Anyway, Maxtor's good enough for large (256 node and up) Linux and Windows clusters doing seismic processing work worth hundreds of millions, if that's any consolation.
Actually, almost all audio gear is only 17" wide. In fact, if you measure some actual rackmount gear, you'll find 17" is the most common width. Most audio gear is not designed to be directly mounted in a rack, but there's a very good reason that extra 2 inches is there: So that it can fit in a 19" rack.
See, this way, there's room to bolt an adapter onto the sides of the box with enough room for rack handles and slides between the rails. (And don't forget, you don't have 19" between the rails, that's the space between the holes in the rails, IIRC...)
Of course, telco equipment uses 23" rails, just to be different. Go figure.
Re:Now begins the hardest part...
on
Ogg Vorbis 1.0
·
· Score: 2
Of course, any computer program or operating system that is well-designed will not need or require anachronisms like filename extensions to figure out how to use a file. Turning the filename into a database with the last three or four characters determining file type is a BAD idea! The fact that it's common practice doesn't make it any better.
These are a throwback to the bad old DEC-isms (8.3, extension identifies file type) that made it into DOS and have continued to pollute even Unix-like operating systems since then.
Old Unix users remember when the "file" command and/etc/magic were the system-wide database that allowed the OS or any application to correctly ID a file format without an extension. They're still there, just not used so much anymore.
And of course, the Mac keeps the type/creator in the resource fork so again, file names only need contian information that's *useful* to the user, as it should be.
HELP STAMP OUT FILE EXTENSIONS!
(Oddly (and sadly), it looks like Microsoft gets this better than anyone - one of the stated goals for the upcoming "Longhorn" OS is that file type/name/location transparency. This is an *incredibly* powerful idea, and it looks like they're taking it several steps further than Apple's resource fork. If Longhorn delivers on half of what it's supposed to, it's the death of every other user-centric OS out there, and MS's desktop/laptop/workstation/tablet dominance will be complete. Longhorn scares me, but it's an exciting concept - if I could buy it today, I'd do so in a heartbeat, because it solves real problems that real users face every day; problems that are not even being thought about by anyone else outside of Apple..)
* Real-world example of the DHTML problems. Go to www.gatorade.com [gatorade.com] , and click on one of the product listings. There is some DHTML which causes a scroolbar to appear in the middle of the screen so that one can scrool down to see all of the products listed. This scrollbar does not work in Mozilla; it does work in IE.
[rant]What a horrible site - way too much bloat - I'm getting *really* tired of waiting too long for code bloat (Java, J/Java/ECMAscript, or the worst offender, the dreaded Flash) to load even over a very fast inet link.[/rant]
Even so, it may not work in Mozilla, but it works perfectly in Netscape 7.0 beta 1, on W98SE, which is what I'm running at the moment. Either this is a byug which was fixed by Netscape and not the Mozilla crowd, or it's platform-related (you don't say what your're running it on.) I know it may not be popular to say so here, but I've found Netscape 7.0b1 to be *considerably* more stable than Mozilla, which wants to create 4 TB bookmark files if given the chance...
The interesting thing, from a scientific perspective, is the sheer lack of study - and the reticence from the scientific community to question the party line.
Wow. Sounds just like evolution. What a coincidence. (Seriously, this isn't a troll (although I fear it will be moderated as one), but rather a sober observation that science is not often interested in investigating things that don't fit with the current body of popular opinion. Regardless of one's opinions on diets or evolution, there is clearly much more real science needing to be done before anyone should run around claiming an exclusive on the facts. In general that hardest thing for scientists to admit is that we simply don't know, even when that's the honest answer...)
FWIW, the Lucent, Cisco, and Buffalo 802.11 gear is excellent, and all are built by the same Japanese company. Here in the US, they use the Buffalo brand - take them apart, and you'll see that Buffalo, Cisco, and Lucent 802.11 gear is indeed all identical, except for a few questionable "value-added" features (which are pretty much non-standard, by definition) and, of course, the price. (I think they may not build the high-end Cisco units, but they're priced out of reason, anyhow.)
Buffalo has also been known to offer good group discounts to "wireless clubs" like the one here in Austin. BTW, Buffalo also has a wide range of antennas and antenna adapters/cables.
NTSC video is probably the safest bet. Make high-quality Beta or VHS copies of what you want to keep, and store them properly for maximum longevity. (Check with tape manufacturers about magnetic tape archival procedures for maximum life.) Because of the volume of such tapes from the 20th and 21st ceturies, this video will probably be usable for a very long time, decades or more.
Remember that with an analog format, degradation will be very slow, and after an initial loss in a few years, it will remain quite stable and usable for a very long time. In general, digital formats are likely to be obsolete and problematic long before you face the same problem with analog formats.
If you insist on digital, DV might be an acceptable option, but it is far from certain that DV will ever achieve anything like the ubiquity that plain old NTSC analog offers. Burning real DVDs may well work as well, since it looks like DVD's becoming a pretty standard format, and players are simple and cheap.
Of course, if you care about preserving the information, be sure to preserve a player (or two) for it, too. (It might even be a good idea to include a monitor or TV, as analog display devices could be hard to find down the road.)
You don't want to be in the awkward situation of a time capsule opened a few decades ago that contained a puzzing item: a spool of fine wire. I turned out to be from a wire recorder, although it took those opening the capsule some time to a) realize what it was, and b) find a working wire recorder on which to play it back. (Wire recorders were popular as business dictation devices in the 1930s and '40s, before Scotch 110 tape allowed magnetic tape to be a real option, creating the magnetic audio recording industry that eventually led Ampex to develop the first magnetic video tape recorder.) Interestingly, magnetic recording wire has one attribute that makes it well-suited for use in time capsules and such: since the magnetic patterns are larger and stored on a more magnetically reslient medium, the signal can be preserved for a very long time - much longer than the usable life of tape. Sorry, I don't know of anyone making video wire recorders, though...:-)
It's amazing how ours may be the first generation in a century to have such a perishable history - our video, our snapshots, and such are all in very vulnerable digital formats, unlike the family photos, slides, and movies that are holding up pretty well after several decades, and do not require any Media-industry-approved equipment to view them. We somehow think burning these onto CDs will preserve them - it may, but not with anywhere near the reliability and robustness of real photographs...
It's really pretty hilarious to hear the/. crowd talking badly about people talking in buzzwords! Have a look around - most of the people here seem incapable of having a conversation with ordinary people. (Most of who are more intelligent, although less technically inclined that the self-righteous crowd that hangs out here.)
Believe it or not, there *is* a real world out there, and people live and work in it. It's that world that produces the surplus that gets wasted on salaries for ungrateful Stallman slaves.
It's crap like this that really makes me wonder if supporting Linux is a good idea. (after many years of supporting Linux, I am seriously pondering whether it is wise to continue to do so.)
If Ransome Love had any sense, he'd recognize this crowd of backstabbers for what it is and go pursure his plan based on BSD, where at least he has a chance at being successful. Linux is doomed to play along the edges of the market so long as the "community" insists on bludgeoning to death anyone that tries to bring it sucessfully into the real world.
If the charge is based on "connections", then what happens when Internet radion uses multicast like it should in the first place? Do the radio rules apply then? (Seems logical, since there's no way of knowing how many people are listening to what is now a true "broadcast".
Of course, this won't happen unless there's significant consumer demand for Mbone connections - enough that "broadband" (I hate the now-common misuse of that term...) ISPs start to offer it as a differentiator for their service.
Technology can make a difference here in the next few years: A near-ubiquitous IPv6 Mbone could change the game substantially. This won't happen as long as the Telcos are controlling the infrastructure, since many of them will face bankruptcy in the next few years even without these additional infrastructure investments. It will be interesting to see what happens, but expect the "content owners" to start making more plays to own the delivery infrastructure viz. Time Warner.
For a small company offsite backups are important too. Look at the USCG data that was lost inthe WTC with an offsite backup apparently in the other tower. If you are paranoid send of the hdd to your mother in arkansas.
I don't know about the USCG, but Cantor Fitzgerald's backup system did indeed copy to another set of servers in the other tower. I know the guy that worked for them to set this up (he's my account rep with one of our storage vendors.) He's a very sharp guy, and it seemed completely reasonable at the time, although he feels horrible now. He says they actually considered whether this was safe, but figured the only risk they ran was that of a monstrous tidal wave, and even that was iffy, so it seemed like a very safe bet.
Sadly, backups would have done little good for Cantor Fitzgerald, since the tragedy of thier loss was people - there was virtually no one left to use the backups even if they had been available...
A striking somewhat-related fact: More Americans are killed in abortion clinics every day than were killed in the Twin Towers Attack. Check it out for yourself - it's true.
A lot of people here are way too quick to dismiss the incredible usefulness of a device like this. Those of us that have owned and used the Tandy Model 100, 102, or the later, sleeker (but buggier) WP-2 know better. So do those millions of us for whom our Palm device is *by far* the most indispensible computer we own. (You can have my PC and 17" LCD monitor long before I turn loose of my Kyocera Smartphone or some other Palm device.)
While clearly not capable of doing everything a real laptop does, these things are in a usefulness class that's *far* beyond any laptop (and I think I speak with some authority, as a former program manager with responsibility for both Latitude and Inspiron at Dell.) Windows CE/PocketPC/HPC has tried to play here, but still doesn't "get it" - there's tremendous value in instant-on, simple applications that just allow the work to get done - fancy GUIs are actually a detriment in such form factors. Battery life measured in weeks or months of ordinary use is incredibly liberating. My dream is to one day own a computer with real desktop capabilities, that can operate through an entire day of hard use (conferences or client meetings), and that does not require luggage for all its life support equipment (rechargers, floppy & CD drives, etc.) Although some have come close (Some ThinkPads, and the new iBook), I've never seen such a thing actually make it to market. (Keep in mind that one of the largest target market for portable devices is the healthcare field, but they can really only consider devices that can make it through an entire shift without recharging - that pretty much eliminates all laptops, leaving you back with a Palm or CE choice. Sadly, too many are being decived by color screens and a "familiar" (not really) Windows interface.
This is a very interesting and useful gadget, but lacks one crucial capability to be a real, viable replacement for a laptop on a business trip: a graphical web browser. Sadly, the last time I checked, there were no native graphical web browsers for Palm devices (WAP is useless, text-based browsers are nearly so, and the TopGun derivatives require impractical transcoding proxies.)
Even so, this is a great tool for those that write a lot, either for deliverables, or simply notes to document the work they do. I can see lots of these in use by consultants, lawyers, and businessmen of all kinds. The main strategic error I see is one of physical form factor - my slim WP-2 is far easier to throw into my breifcase "just in case" than the Dana, with that wierd curved design that doubles its thickness. The second biggest erro is the lack of a real network connection (wired or unwired), but that's a Palm problem and not properly something that's the Dana's fault. (It is inexcusable that Palm hasn't fixed this in the years they've had to do so. It is the single biggest reason the CE devices are starting to kick their butt in vertical applications, but they seem oblivious, as usual.)
This sort of thing could be really interesting with the next-gen BeOS-derived environment. I can't wait. Buying one of these would be easier (and about the same cost) as hacking my WP-2 to run a uSimm/uCLinux card internally, although that would have the advantage of a real Ethernet interface...
Like Columbus' discovery of the new world, it's not news because it's first, it's news because it signals the beginning of commercialization of the concept.
The appearance of such a BIOS product on the market, especially when driven by an industry leader like Phoenix, may well change what's possible on bare metal. That's a very good thing, and a welcome return to computers that boot into a useful and usable state, like all old pre-PC computers (Apples, Commodores, TRS-80s, etc.) used to do. (Actually, the original IBM PCs (and early clones) had a socket for an optional BASIC ROM, since IBM realized most existing PC users expected the computer to be useful for something (even programming) at power-on. I've still got an old Xerox 1500 PC-clone with it's extra sockets filled with a BASIC ROM and the much-desired 8087 math coprocessor, which really sped up VisiCalc...)
In a few years, this may well be viewed as one of the most important announcements in the history of the PC. It literally has the potential to change everything. This is a big deal. Good job, Phoenix.
And the license cost for all of that from QNX would be far more than Phoenix can get awwy with charging even for such a rich BIOS.
Economics matter, and QNX has never really grasped that, which is why despite the elegance and power of their architecture and code, they are still a mere curiosity and a non-starter in the market. (I know, I used to work in set-top box architecture - QNX offered great capabilities, but we never even seriously considered them, because they made it quite clear that they expected license fees that would have scuttled the economic viability of the entire product. Sad.)
To quote a famous industry leader, "You have no security anyway. Get over it."
Seriously, this is probably far more secure than the way things are done today. I would love to have this now, if for no other reason than because it would make getting new drivers a breeze when doing the semi-annual Windows re-install.
Besides, outside a small percentage of geeks, any security that requires any user attention at all simply isn't going to get done. If you're worried about security, perhaps you shouldn't be using Linux at all (unless it's the NSA's SELinux), but should look at something like OpenBSD instead.
There's no reason to assume that Phoenix' FirstBIOS is in any way related to LANL's Linux BIOS project you point out. There have been several such Linux-based BIOs efforts, the first of which was, I believe, San Mehat's Linux BIOS for the Corel Netwinder. (I'm not sure whare San is these days - he was at VA Research, but now that they're out of the Linux biz, I suppose he's moved on somewhere else - perhaps he went to work for Phoenix?)
Re:Do they expect to sell these?
on
IMSAI Series Two
·
· Score: 2
his box would be more comparable to some overpriced kit car that tries to look like a 1932 Ford, rather than an actual 1932 Ford.
A better comparison might be to a Cobra kit car as compared to an original Cobra. There definitely aren't enough of those to go around, and the ones that are left are ridiculously expensive. And to top it off, the better modern Cobras (say, Factory Five's, for example) are actually *better* than the originals. All the good of the original, with some of the stupidity filed off. (If you've ever ridden in a real big-block Cobra, you'll remember the funny smell of your shoes being roasted by their close proximity to the exhaust, shielded (or is that just re-radiated?) only by a bit of flimsy sheet metal...)
No this isn't for everyone. It's cool, though, and I wish them well. In fact, I wish I'd thought of building them myself.
We were robbed when they took away our front-panel switches and LEDs! Keyboard? We don't need no stinkin' keyboard...
Actually, "Software Automatic Mouth". I remember because I always thought it was a bit awkward...
I posted an article here only a few days ago lamenting the creativity and innovation we saw in the hacker community in those days. Maybe I'm just getting old, but back then, hacks were real, and really did something, they weren't just poser "hacks" consisting of case mods, neon lighting, and replacing the front panel LEDs with blue ones. Oooooh, that's impressive - NOT!
"It's about freakin' time! Now when are we going to replace these markedly inferior CD-ROMs and DVD-ROMs with the gloriously mellifluous LP-ROM?"
Actually, it's already been done. In my garage, I have a copy of Interface Age Magazine, one of the early PC mags (back when PC meant IMSAI 8080 or Altair or SWTPC 6800 boxes with S-100 buses, or if you were really flush with cash, the incredibly slick boxes from The Digital Group or CompuColor!)
This issue includes one of those LPs pressed on thin vinyl that were sometimes bound into magazines in the '70s. The interesting thing about this one (which they billed as the first "floppy ROM") is that instead of musical audio, it contained a BASIC interpreter (I think, I really don't remember what he program was, since I had no computer myself, I just read and dreamed) encoded on the disk in "Standard" Kansas City format - the same way data was commonly encoded on cassette tapes at the time. (Those of you that have never loaded a program from cassette have no concept of "slow" - I want no more griping about kernel build times on your 2 GHz P4s and Athlons!)
Therefore, you could just hook your turntable and stereo (line level out) up to the cassette interface on your computer, and voila!, you had a BASIC interpreter, dumped directly into your computer from a magazine delivered via regular mail.
This is so cool that it's one of the few magazines I kept from that era - I wish I had hung on to more, as the hacking was really far more adventurous back then than it is now... It's amazing how we're just now getting back to doing the things they were doing back then, things like voice recognition and such, which hasn't really improved all that much (perhaps 10-20x) despite CPU hardware that's 100,000 times faster.
Makes you wonder where we'd be if the vibrant CP/M hacker community that was dominant prior to the IBM PC had been able to somehow survive and continue to grow.
Well, while I was a fish in the dorms I changed a sign from "All Visitors Must Be Escorted" to "All Escorts Must Be Visited"... went unnoticed by most everyone all year long... gave me a chuckle every time I entered... guess I'm an artist too...
I love those signs with the removable letters and the clear, bubble-type covers. We had several of those at the Sun office in Houston, and as I saw them daily, it one day struck me that anagrams were in order. Stealing letters from other signs was not allowed - you had to play the hand you were dealt.
Two in particular I remember - One of them had said "OPEN DOOR SLOWLY" (it really was easy to cold-cock somebody with it otherwise) - that became "OPEN SLOW OLD ROY", not terribly clever, but not bad, and preserved the meaning acceptably.
The other was a junk room which at one time had actually been what it was labelled, "TAPE LIBRARY". I just coud not resist turning that one into "TRIPE RAY LAB". Both signs stayed in their modified form for the last year I was there, and presumably remained until Sun moved to another building a couple of years later....
I still get a chuckle out of the mental picture of a tripe ray, though...
Many, many people have been killed (read any good history book) just for trying to organize. Remember that the next time you say "we don't want a union".
Sorry, I have to disagree here: Unions are inherently violent, corrupt, and the most egregious oppressors of the workers that have ever existed. As a young engineer at an aerospace company in California, I was given the job of watching for Union attacks from the top of the roof of the main assembly building. (This after the company acquiesced to all significant demands of the Union - it was later revealed that they struck "because we had the money in the strike fund".) After the mob of Union thugs turned over three Police cars and set fire to them outside the company gates, just "to make a point", they STARTED SHOOTING AT US later in the afternoon. No one was injured, but they could have been, and Union bulletin boards encouraged trying to take us out. (It gets darn cold on the roof in Riverside at night, BTW!)
Let there be no mistake about it: Unions are BAD!!! Nothing they've brought to American wokers is worth the continual price we pay in corruption, murder, and mayhem. Thank God I now live in a right-to-work state, where no one can be forced to join a union and have their money forcibly confiscated to pay for criminal activities or lobbying for causes they disagree with. Fortunately, the American people are beginning to see Unions for what they really are, and Union membership is the lowest point in decades.
Last year, I picked up a Xerox WorkCentre M940 for less than $100 at OfficeMax - not bad for a decent scanner, copier, and printer, especially one with indiviual ink taks for each color so you don't have to waste ink just because one color wears out.
I was a bit shocked when I went in to buy an extra set of cartridges at the same store: I was looking at almost $150 for a set of replacement cartridges (to be fair, that included the higher capacity black cartridge.)
Anyway, I balked and went online to find a better deal. I ordered from AcuJetUSA, since it looked like they had a high quality, professinal cartridge. Whn I finally tried them, I found that the AcuJet ink (these were replaceemnt cartridges, not some cheesy refill kit) was of a VASTLY DIFFERENT COMPOSITION from the Xerox originals. After replacing only the black cartridge, it's now impossible to print anything with multiple colors - the black ink bleeds almost 1/16" into any adjacent colored area, something that never happened with the Xerox ink. In addition, the AcuJet ink is "wetter", leaving the paper puckered and wavy from moisture even when printing an all-black page.
To AcuJet's credit, even though the normal 30-day return period was up, they've agreed to take all the cartridges back and return my money. (I'll send them back in a day or two, and we'll see.)
I will be buying geniune Xerox cartridges from now on, though. The hassle of returning these alone isn't worth the money I would have saved. Amazon seems to have about the best deal among the reliable vendors. (I don't want a refilled cartridge passed off as new...)
Slashdot Mirror
Actually, the basic technology here is not at all new. These are what is know as "McKibben Muscles", first developed for artificial arms in the 1950s. (The link is to a good overview of them from my robotics bookmarks list.)
They are cheap, strong, compact, and easy to build, but they require a lot of energy in the form of not-too-easily available compressed air or other gas. For this reason, they're not used too frequently for prosthetics since carrying enough gas becomes a problem.
These have been used to build a number of camplex life-mimicking robots, including many of the better legged walking robots. Check out the this page for an idea of how you can use them in place of real muscles to achieve very lifelike results.
If you've got a good source of compressed gas, these are an excellent design choice, and probably the only thing we have that can fairly approximate natural muscle at a reasonable cost. ("Muscle wires" like Nitinol require gobs of power and need way too much cooling time to contract to be useful in most applications.)
I'm not a DB expert by any means, but here's what I do know about Intersystems' Cache - it can be faster than anything else in certain real-world applications:
I used to be Market Segment Manager for healthcare at Sun Microsystems. This is back when Sun was a "who?" in healthcare IT (outside of medical imaging, where we were dominant, but buried in OEM gear behind the name of a Seimens, GE, etc.)
When the Ultras came out, we wanted to aggressively go after some of the spaces where we would be a good fit. One of those was electronic medical records and managed care systems, both of which tend to be large, sparsely populated databases.
We worked closely with a clinical system vendor (which has an excellent product and has moved from a relatively unknown newcomer to one of the most favored products in the industry) and helped them face off not only Sun gear vs. IBM, HP, and DEC, but also to find the best DB to work with the hardware environment.
According to thier own tests, this user reported the following: Oracle and DB2 were essentially neck and neck, but Oracle offered better network support, was cheaper, and was considerably easier to do business with than IBM. They were stunned to find that for thier application, Cache was THIRTY times faster than Oracle out-of-the-box. It was still nearly 10X faster even after doing "unnatural things to Oracle that you wouldn't want to live with."
Similar results were reported a few months later by another ISV that built the first large-scale managed care enrollment system that was proven capable of scaling to millions of members.
I'm also somewhat familiar with what was for a long time the showcase site for Intersystems: Boston Children's Hospital and the Harvard Medical Center. This system has grown along with the hospital server-by-server. In 1995-ish, the clinical system there was based on around 150 PC servers scattered throughout the facility. Intersystems Distributed Cache Protocol keeps the data on the move so it tends to reside neares its point of most frequent use automatically. I don't know of anything other DB that provides this sort of modular scalability and caching performance that's so important in the real world., although there may be something else similar.
I'm impressed because people I know that really know this stuff are impressed, again, I make no claim to being a database guru myself.
I want views!!! Precious views!! I need to secure certain data from the prying eyes of account managers and sales folk.
/rdb that allows fast, flexible database operations and allowing the use of the full power of the Unix environment. ( It's easy, leverages a zillion other existing tools, and everything you learn is directly applicable to anything else you do in a Unix environment.)
Sounds to me like you're really trying to hammer that socket-head capscrew into place, rather than finding an appropriate Allen wrench.
It's amazing to me how many people in the open source community continually try to force mySQL into doing jobs it's clearly not up to. This is especially puzzling, since there *are* good, modern, high-performance, high-function databases out there, things like Postgresql, Interbase,and the Sleepycat Berkeley DB (not SQL, but really fast and solid) among many others.
For quick, dirty, and fast, without the hassles of dealing with SQL at all, there are options like NoSQL and Starbase, both of which are based on the Stream/Modifier concept of the original
It seems most people want SQL just because it somehow legitimizes their project as "serious", when in reality, it often just adds undesirable complexity and support cost. SQL is often unnecessary and actually undesirable, if you allow yourself to consider the possible benefits of NOT using a SQL database.
Also, keep in mind that *any* database (and especially ones that are already very fast,like BerkeleyDB or Starbase) is lighting fast in RAM, and memory is now cheap enough to make putting the whole DB in memory a very do-able thing: A gigabyte or several of database is BIG, and easily supportable on very affordable server hardware.
And of course, there is a reason those commercial database companies exist: They often offer capabilities that open source users may never have. I am looking right now at a new application which will be *much* faster and cheaper based on Intersystems' unique Cache database than it eould be on any of the more common big names. It's important to choose the right tool for the job. (In this case, I need lots of transactions in a very large sparsely populated database, Cache is excellent there, an order of magnitude better than Oracle or DB2, which is why Cache is the leading DB choice for electronic medical records systems, and has been for years.)
MySQL is not the answer to every database problem. Or even most of them, truth be told. Personally, I find it to be a minimally competent, fussy, and underpowered tool. It's good for some things, but to be honest, I've always thought its appeal had more to do with being the first real open source SQL DB than with being good.
As always, this boils down to choosing the right tool for the job. Not exactly rocket science, but something way too many people don't bother to do.
As the architect of a high-performance multi-terabyte storage subsystem based on IDE disk array technology, I beg to differ about Maxtor's quality. Just this past fall, we spent a long time evaluating which drives were the best. The general consensus (and our own conclusion) is Maxtor's large drives run cooler and last longer (failure rates start to become an important item when you have several hundred such drives in an array.)
IBM drives were better at one time, but that's not been the case for a while, and I doubt Hitachi can change that anytime soon.
Anyway, Maxtor's good enough for large (256 node and up) Linux and Windows clusters doing seismic processing work worth hundreds of millions, if that's any consolation.
Actually, almost all audio gear is only 17" wide. In fact, if you measure some actual rackmount gear, you'll find 17" is the most common width. Most audio gear is not designed to be directly mounted in a rack, but there's a very good reason that extra 2 inches is there: So that it can fit in a 19" rack.
See, this way, there's room to bolt an adapter onto the sides of the box with enough room for rack handles and slides between the rails. (And don't forget, you don't have 19" between the rails, that's the space between the holes in the rails, IIRC...)
Of course, telco equipment uses 23" rails, just to be different. Go figure.
Of course, any computer program or operating system that is well-designed will not need or require anachronisms like filename extensions to figure out how to use a file. Turning the filename into a database with the last three or four characters determining file type is a BAD idea! The fact that it's common practice doesn't make it any better.
/etc/magic were the system-wide database that allowed the OS or any application to correctly ID a file format without an extension. They're still there, just not used so much anymore.
These are a throwback to the bad old DEC-isms (8.3, extension identifies file type) that made it into DOS and have continued to pollute even Unix-like operating systems since then.
Old Unix users remember when the "file" command and
And of course, the Mac keeps the type/creator in the resource fork so again, file names only need contian information that's *useful* to the user, as it should be.
HELP STAMP OUT FILE EXTENSIONS!
(Oddly (and sadly), it looks like Microsoft gets this better than anyone - one of the stated goals for the upcoming "Longhorn" OS is that file type/name/location transparency. This is an *incredibly* powerful idea, and it looks like they're taking it several steps further than Apple's resource fork. If Longhorn delivers on half of what it's supposed to, it's the death of every other user-centric OS out there, and MS's desktop/laptop/workstation/tablet dominance will be complete. Longhorn scares me, but it's an exciting concept - if I could buy it today, I'd do so in a heartbeat, because it solves real problems that real users face every day; problems that are not even being thought about by anyone else outside of Apple..)
* Real-world example of the DHTML problems. Go to www.gatorade.com [gatorade.com] , and click on one of the product listings. There is some DHTML which causes a scroolbar to appear in the middle of the screen so that one can scrool down to see all of the products listed. This scrollbar does not work in Mozilla; it does work in IE.
[rant]What a horrible site - way too much bloat - I'm getting *really* tired of waiting too long for code bloat (Java, J/Java/ECMAscript, or the worst offender, the dreaded Flash) to load even over a very fast inet link.[/rant]
Even so, it may not work in Mozilla, but it works perfectly in Netscape 7.0 beta 1, on W98SE, which is what I'm running at the moment. Either this is a byug which was fixed by Netscape and not the Mozilla crowd, or it's platform-related (you don't say what your're running it on.) I know it may not be popular to say so here, but I've found Netscape 7.0b1 to be *considerably* more stable than Mozilla, which wants to create 4 TB bookmark files if given the chance...
The interesting thing, from a scientific perspective, is the sheer lack of study - and the reticence from the scientific community to question the party line.
Wow. Sounds just like evolution. What a coincidence. (Seriously, this isn't a troll (although I fear it will be moderated as one), but rather a sober observation that science is not often interested in investigating things that don't fit with the current body of popular opinion. Regardless of one's opinions on diets or evolution, there is clearly much more real science needing to be done before anyone should run around claiming an exclusive on the facts. In general that hardest thing for scientists to admit is that we simply don't know, even when that's the honest answer...)
FWIW, the Lucent, Cisco, and Buffalo 802.11 gear is excellent, and all are built by the same Japanese company. Here in the US, they use the Buffalo brand - take them apart, and you'll see that Buffalo, Cisco, and Lucent 802.11 gear is indeed all identical, except for a few questionable "value-added" features (which are pretty much non-standard, by definition) and, of course, the price. (I think they may not build the high-end Cisco units, but they're priced out of reason, anyhow.)
Buffalo has also been known to offer good group discounts to "wireless clubs" like the one here in Austin. BTW, Buffalo also has a wide range of antennas and antenna adapters/cables.
NTSC video is probably the safest bet. Make high-quality Beta or VHS copies of what you want to keep, and store them properly for maximum longevity. (Check with tape manufacturers about magnetic tape archival procedures for maximum life.) Because of the volume of such tapes from the 20th and 21st ceturies, this video will probably be usable for a very long time, decades or more.
:-)
Remember that with an analog format, degradation will be very slow, and after an initial loss in a few years, it will remain quite stable and usable for a very long time. In general, digital formats are likely to be obsolete and problematic long before you face the same problem with analog formats.
If you insist on digital, DV might be an acceptable option, but it is far from certain that DV will ever achieve anything like the ubiquity that plain old NTSC analog offers. Burning real DVDs may well work as well, since it looks like DVD's becoming a pretty standard format, and players are simple and cheap.
Of course, if you care about preserving the information, be sure to preserve a player (or two) for it, too. (It might even be a good idea to include a monitor or TV, as analog display devices could be hard to find down the road.)
You don't want to be in the awkward situation of a time capsule opened a few decades ago that contained a puzzing item: a spool of fine wire. I turned out to be from a wire recorder, although it took those opening the capsule some time to a) realize what it was, and b) find a working wire recorder on which to play it back. (Wire recorders were popular as business dictation devices in the 1930s and '40s, before Scotch 110 tape allowed magnetic tape to be a real option, creating the magnetic audio recording industry that eventually led Ampex to develop the first magnetic video tape recorder.) Interestingly, magnetic recording wire has one attribute that makes it well-suited for use in time capsules and such: since the magnetic patterns are larger and stored on a more magnetically reslient medium, the signal can be preserved for a very long time - much longer than the usable life of tape. Sorry, I don't know of anyone making video wire recorders, though...
It's amazing how ours may be the first generation in a century to have such a perishable history - our video, our snapshots, and such are all in very vulnerable digital formats, unlike the family photos, slides, and movies that are holding up pretty well after several decades, and do not require any Media-industry-approved equipment to view them. We somehow think burning these onto CDs will preserve them - it may, but not with anywhere near the reliability and robustness of real photographs...
It's really pretty hilarious to hear the /. crowd talking badly about people talking in buzzwords! Have a look around - most of the people here seem incapable of having a conversation with ordinary people. (Most of who are more intelligent, although less technically inclined that the self-righteous crowd that hangs out here.)
Believe it or not, there *is* a real world out there, and people live and work in it. It's that world that produces the surplus that gets wasted on salaries for ungrateful Stallman slaves.
It's crap like this that really makes me wonder if supporting Linux is a good idea. (after many years of supporting Linux, I am seriously pondering whether it is wise to continue to do so.)
If Ransome Love had any sense, he'd recognize this crowd of backstabbers for what it is and go pursure his plan based on BSD, where at least he has a chance at being successful. Linux is doomed to play along the edges of the market so long as the "community" insists on bludgeoning to death anyone that tries to bring it sucessfully into the real world.
If the charge is based on "connections", then what happens when Internet radion uses multicast like it should in the first place? Do the radio rules apply then? (Seems logical, since there's no way of knowing how many people are listening to what is now a true "broadcast".
Of course, this won't happen unless there's significant consumer demand for Mbone connections - enough that "broadband" (I hate the now-common misuse of that term...) ISPs start to offer it as a differentiator for their service.
Technology can make a difference here in the next few years: A near-ubiquitous IPv6 Mbone could change the game substantially. This won't happen as long as the Telcos are controlling the infrastructure, since many of them will face bankruptcy in the next few years even without these additional infrastructure investments. It will be interesting to see what happens, but expect the "content owners" to start making more plays to own the delivery infrastructure viz. Time Warner.
For a small company offsite backups are important too. Look at the USCG data that was lost inthe WTC with an offsite backup apparently in the other tower. If you are paranoid send of the hdd to your mother in arkansas.
I don't know about the USCG, but Cantor Fitzgerald's backup system did indeed copy to another set of servers in the other tower. I know the guy that worked for them to set this up (he's my account rep with one of our storage vendors.) He's a very sharp guy, and it seemed completely reasonable at the time, although he feels horrible now. He says they actually considered whether this was safe, but figured the only risk they ran was that of a monstrous tidal wave, and even that was iffy, so it seemed like a very safe bet.
Sadly, backups would have done little good for Cantor Fitzgerald, since the tragedy of thier loss was people - there was virtually no one left to use the backups even if they had been available...
A striking somewhat-related fact: More Americans are killed in abortion clinics every day than were killed in the Twin Towers Attack. Check it out for yourself - it's true.
A lot of people here are way too quick to dismiss the incredible usefulness of a device like this. Those of us that have owned and used the Tandy Model 100, 102, or the later, sleeker (but buggier) WP-2 know better. So do those millions of us for whom our Palm device is *by far* the most indispensible computer we own. (You can have my PC and 17" LCD monitor long before I turn loose of my Kyocera Smartphone or some other Palm device.)
While clearly not capable of doing everything a real laptop does, these things are in a usefulness class that's *far* beyond any laptop (and I think I speak with some authority, as a former program manager with responsibility for both Latitude and Inspiron at Dell.) Windows CE/PocketPC/HPC has tried to play here, but still doesn't "get it" - there's tremendous value in instant-on, simple applications that just allow the work to get done - fancy GUIs are actually a detriment in such form factors. Battery life measured in weeks or months of ordinary use is incredibly liberating. My dream is to one day own a computer with real desktop capabilities, that can operate through an entire day of hard use (conferences or client meetings), and that does not require luggage for all its life support equipment (rechargers, floppy & CD drives, etc.) Although some have come close (Some ThinkPads, and the new iBook), I've never seen such a thing actually make it to market. (Keep in mind that one of the largest target market for portable devices is the healthcare field, but they can really only consider devices that can make it through an entire shift without recharging - that pretty much eliminates all laptops, leaving you back with a Palm or CE choice. Sadly, too many are being decived by color screens and a "familiar" (not really) Windows interface.
This is a very interesting and useful gadget, but lacks one crucial capability to be a real, viable replacement for a laptop on a business trip: a graphical web browser. Sadly, the last time I checked, there were no native graphical web browsers for Palm devices (WAP is useless, text-based browsers are nearly so, and the TopGun derivatives require impractical transcoding proxies.)
Even so, this is a great tool for those that write a lot, either for deliverables, or simply notes to document the work they do. I can see lots of these in use by consultants, lawyers, and businessmen of all kinds. The main strategic error I see is one of physical form factor - my slim WP-2 is far easier to throw into my breifcase "just in case" than the Dana, with that wierd curved design that doubles its thickness. The second biggest erro is the lack of a real network connection (wired or unwired), but that's a Palm problem and not properly something that's the Dana's fault. (It is inexcusable that Palm hasn't fixed this in the years they've had to do so. It is the single biggest reason the CE devices are starting to kick their butt in vertical applications, but they seem oblivious, as usual.)
This sort of thing could be really interesting with the next-gen BeOS-derived environment. I can't wait. Buying one of these would be easier (and about the same cost) as hacking my WP-2 to run a uSimm/uCLinux card internally, although that would have the advantage of a real Ethernet interface...
Like Columbus' discovery of the new world, it's not news because it's first, it's news because it signals the beginning of commercialization of the concept.
The appearance of such a BIOS product on the market, especially when driven by an industry leader like Phoenix, may well change what's possible on bare metal. That's a very good thing, and a welcome return to computers that boot into a useful and usable state, like all old pre-PC computers (Apples, Commodores, TRS-80s, etc.) used to do. (Actually, the original IBM PCs (and early clones) had a socket for an optional BASIC ROM, since IBM realized most existing PC users expected the computer to be useful for something (even programming) at power-on. I've still got an old Xerox 1500 PC-clone with it's extra sockets filled with a BASIC ROM and the much-desired 8087 math coprocessor, which really sped up VisiCalc...)
In a few years, this may well be viewed as one of the most important announcements in the history of the PC. It literally has the potential to change everything. This is a big deal. Good job, Phoenix.
And the license cost for all of that from QNX would be far more than Phoenix can get awwy with charging even for such a rich BIOS.
Economics matter, and QNX has never really grasped that, which is why despite the elegance and power of their architecture and code, they are still a mere curiosity and a non-starter in the market. (I know, I used to work in set-top box architecture - QNX offered great capabilities, but we never even seriously considered them, because they made it quite clear that they expected license fees that would have scuttled the economic viability of the entire product. Sad.)
To quote a famous industry leader, "You have no security anyway. Get over it."
Seriously, this is probably far more secure than the way things are done today. I would love to have this now, if for no other reason than because it would make getting new drivers a breeze when doing the semi-annual Windows re-install.
Besides, outside a small percentage of geeks, any security that requires any user attention at all simply isn't going to get done. If you're worried about security, perhaps you shouldn't be using Linux at all (unless it's the NSA's SELinux), but should look at something like OpenBSD instead.
There's no reason to assume that Phoenix' FirstBIOS is in any way related to LANL's Linux BIOS project you point out. There have been several such Linux-based BIOs efforts, the first of which was, I believe, San Mehat's Linux BIOS for the Corel Netwinder. (I'm not sure whare San is these days - he was at VA Research, but now that they're out of the Linux biz, I suppose he's moved on somewhere else - perhaps he went to work for Phoenix?)
his box would be more comparable to some overpriced kit car that tries to look like a 1932 Ford, rather than an actual 1932 Ford.
A better comparison might be to a Cobra kit car as compared to an original Cobra. There definitely aren't enough of those to go around, and the ones that are left are ridiculously expensive. And to top it off, the better modern Cobras (say, Factory Five's, for example) are actually *better* than the originals. All the good of the original, with some of the stupidity filed off. (If you've ever ridden in a real big-block Cobra, you'll remember the funny smell of your shoes being roasted by their close proximity to the exhaust, shielded (or is that just re-radiated?) only by a bit of flimsy sheet metal...)
No this isn't for everyone. It's cool, though, and I wish them well. In fact, I wish I'd thought of building them myself.
We were robbed when they took away our front-panel switches and LEDs! Keyboard? We don't need no stinkin' keyboard...
Actually, "Software Automatic Mouth". I remember because I always thought it was a bit awkward...
I posted an article here only a few days ago lamenting the creativity and innovation we saw in the hacker community in those days. Maybe I'm just getting old, but back then, hacks were real, and really did something, they weren't just poser "hacks" consisting of case mods, neon lighting, and replacing the front panel LEDs with blue ones. Oooooh, that's impressive - NOT!
"It's about freakin' time! Now when are we going to replace these markedly inferior CD-ROMs and DVD-ROMs with the gloriously mellifluous LP-ROM?"
Actually, it's already been done. In my garage, I have a copy of Interface Age Magazine, one of the early PC mags (back when PC meant IMSAI 8080 or Altair or SWTPC 6800 boxes with S-100 buses, or if you were really flush with cash, the incredibly slick boxes from The Digital Group or CompuColor!)
This issue includes one of those LPs pressed on thin vinyl that were sometimes bound into magazines in the '70s. The interesting thing about this one (which they billed as the first "floppy ROM") is that instead of musical audio, it contained a BASIC interpreter (I think, I really don't remember what he program was, since I had no computer myself, I just read and dreamed) encoded on the disk in "Standard" Kansas City format - the same way data was commonly encoded on cassette tapes at the time. (Those of you that have never loaded a program from cassette have no concept of "slow" - I want no more griping about kernel build times on your 2 GHz P4s and Athlons!)
Therefore, you could just hook your turntable and stereo (line level out) up to the cassette interface on your computer, and voila!, you had a BASIC interpreter, dumped directly into your computer from a magazine delivered via regular mail.
This is so cool that it's one of the few magazines I kept from that era - I wish I had hung on to more, as the hacking was really far more adventurous back then than it is now... It's amazing how we're just now getting back to doing the things they were doing back then, things like voice recognition and such, which hasn't really improved all that much (perhaps 10-20x) despite CPU hardware that's 100,000 times faster.
Makes you wonder where we'd be if the vibrant CP/M hacker community that was dominant prior to the IBM PC had been able to somehow survive and continue to grow.
Well, while I was a fish in the dorms I changed a sign from "All Visitors Must Be Escorted" to "All Escorts Must Be Visited"... went unnoticed by most everyone all year long... gave me a chuckle every time I entered... guess I'm an artist too...
I love those signs with the removable letters and the clear, bubble-type covers. We had several of those at the Sun office in Houston, and as I saw them daily, it one day struck me that anagrams were in order. Stealing letters from other signs was not allowed - you had to play the hand you were dealt.
Two in particular I remember - One of them had said "OPEN DOOR SLOWLY" (it really was easy to cold-cock somebody with it otherwise) - that became "OPEN SLOW OLD ROY", not terribly clever, but not bad, and preserved the meaning acceptably.
The other was a junk room which at one time had actually been what it was labelled, "TAPE LIBRARY". I just coud not resist turning that one into "TRIPE RAY LAB". Both signs stayed in their modified form for the last year I was there, and presumably remained until Sun moved to another building a couple of years later....
I still get a chuckle out of the mental picture of a tripe ray, though...
That's one hell of a flashlight. I'll call mine "Little Boy". I promise to only use it in self defense. And to start small fires.
;-)
Yeah, I always wanted a flashlight strong enough to kick when you turn it on...
Many, many people have been killed (read any good history book) just for trying to organize. Remember that the next time you say "we don't want a union".
Sorry, I have to disagree here: Unions are inherently violent, corrupt, and the most egregious oppressors of the workers that have ever existed. As a young engineer at an aerospace company in California, I was given the job of watching for Union attacks from the top of the roof of the main assembly building. (This after the company acquiesced to all significant demands of the Union - it was later revealed that they struck "because we had the money in the strike fund".) After the mob of Union thugs turned over three Police cars and set fire to them outside the company gates, just "to make a point", they STARTED SHOOTING AT US later in the afternoon. No one was injured, but they could have been, and Union bulletin boards encouraged trying to take us out. (It gets darn cold on the roof in Riverside at night, BTW!)
Let there be no mistake about it: Unions are BAD!!! Nothing they've brought to American wokers is worth the continual price we pay in corruption, murder, and mayhem. Thank God I now live in a right-to-work state, where no one can be forced to join a union and have their money forcibly confiscated to pay for criminal activities or lobbying for causes they disagree with.
Fortunately, the American people are beginning to see Unions for what they really are, and Union membership is the lowest point in decades.
Now THAT is a huge boost to the cause of freedom!
Last year, I picked up a Xerox WorkCentre M940 for less than $100 at OfficeMax - not bad for a decent scanner, copier, and printer, especially one with indiviual ink taks for each color so you don't have to waste ink just because one color wears out.
I was a bit shocked when I went in to buy an extra set of cartridges at the same store: I was looking at almost $150 for a set of replacement cartridges (to be fair, that included the higher capacity black cartridge.)
Anyway, I balked and went online to find a better deal. I ordered from AcuJetUSA, since it looked like they had a high quality, professinal cartridge. Whn I finally tried them, I found that the AcuJet ink (these were replaceemnt cartridges, not some cheesy refill kit) was of a VASTLY DIFFERENT COMPOSITION from the Xerox originals. After replacing only the black cartridge, it's now impossible to print anything with multiple colors - the black ink bleeds almost 1/16" into any adjacent colored area, something that never happened with the Xerox ink. In addition, the AcuJet ink is "wetter", leaving the paper puckered and wavy from moisture even when printing an all-black page.
To AcuJet's credit, even though the normal 30-day return period was up, they've agreed to take all the cartridges back and return my money. (I'll send them back in a day or two, and we'll see.)
I will be buying geniune Xerox cartridges from now on, though. The hassle of returning these alone isn't worth the money I would have saved. Amazon seems to have about the best deal among the reliable vendors. (I don't want a refilled cartridge passed off as new...)