When I was in college (graduated 1998), I took a software-only degree.
If I had it to do over again, I would have taken both hardware and software.
There have been times I have wanted to know what happens at the hardware level, and have been baffled by it. There is also the practical side of it: I don't like staring at a broken circuit board and not knowing how to troubleshoot and repair it.
Also, as someone else mentioned, there's the practical nature of hardware: it can't easily be outsourced, as it must often be babysat in person....
It's easier (and friendlier) to download the RealPlayer, but it results in the same spyware-infested client. What does it say about RealPlayer when their biggest clients are threatening to rebel, because of frustrations caused by their listeners trying to use Real's spyware-infested software? It's sad that Real is following the KaZaA "business model" by selling out to spyware.
(Can't resist a link to my original rant about RealPlayer 8. I'm glad that version was their nadir, and they've been (slowly) improving since then.)
As far as MP3 hitting critical mass goes, isn't it already there?
Doesn't everybody already have some kind of hardware or software these days that can play the MP3 format?
What I want to see is audiobooks sold in MP3 CD format. Currently, audiobooks take a huge number of tapes (or conventional CD's) to hold audio, and because it's spoken word, it would compress very easily while still remaining high quality.
I've converted the Harry Potter books to MP3 CD's, and they sound great. The first two books even fit together on a single CD! All this was from a large stack of grossly inefficient standard audio CD's. No need to store spoken word at 44100x16x2 uncompressed!
I'm hoping the makers of audiobooks will realize that MP3 has already hit critical mass for years now....
I have a Swords Of Fury machine. It only has 3 balls installed, for a 3-ball multiball.
The flippers at the bottom aren't unusually close. Maybe what you're thinking of is the post that's placed dead center to protect balls from draining straight down the middle (as often they drop from the center ramp).
Never did find an infinite-life bug, if there was one, and the flippers never were powerful enough to actually get over the walls (I've never heard of this happening).
Swords Of Fury is considered a very easy game, though, and it's fairly easy to play for as long as you want (if you can keep getting multiballs)....
I was thinking about the recent conflict between municipal and commercial Internet access providers. Government-owned Internet access that blankets an area, whether it be via wireless or fiber-optic cable or something equally useful, is a great thing to have from a property-values and quality-of-life standpoint (hence the city's motivation to build it). However, it obviously competes against commercial ISP's, which (rightfully) feel undercut by the government.
How about this compromise:
Government-owned ISP's focus on doing what they do best: building out infrastructure and reaching through the unprofitable "last mile" to get all customers -- not just the most desirable large corporations or dense urban populations -- hooked up. The purpose of this infrastructure would be to transparently pass packets through, acting as a pipe between their customers and the Internet.
Commercial ISP's focus on providing content and configuring the network to deliver that content. IP address assignment (DHCP), provisioning of subnets, useful servers (email, web hosting, newsgroups), etc. would be handled by competing commercial ISP's. All the various levels and varieties of access found today would still be available: customers could choose to pay extra for a static IP, or certain premium content, or whatever else strikes their fancy. All of this content would be delivered via the municipal infrastructure! Both commercial and government play useful parts here.
Do you think this compromise would actually work in practice? I'd love to see it given a chance somewhere....
Local powerful CPU still good for grid
on
The PC Is Not Dead
·
· Score: 1
Yes, the primary reason for the desirability of thin clients in the workplace is their lack of an administration burden.
Being able to save money by purchasing less powerful hardware isn't even close: over the useful life of the system, the cost of administration more than outweighs the one-time expense of the hardware. In fact, bizarre CPU-less hardware (some dedicated thin client "terminals") can cost *more* than an equivalent standard PC, because of lack of an economy of scale!
Most corporations just ride the price-performance curve and purchase whatever standard PC is currently in the middle of the market. These days, that gets a decently powerful CPU. Thin clients become thin due to the use of network booting and non-local file storage, not due to the lack of a CPU! When properly configured, a standard PC makes a good thin client.
What would be a wonderful use of resources is to form some kind of cluster/grid out of all of the thin clients! The powerful local CPU could then be used for jobs that are submitted over the network, giving a powerful large cluster for heavy computation.
The user would use the PC as a thin client, running remote applications, and any unused local CPU power would still go to good use, powering the cluster! To me, this seems like a win-win situation....
Enjoy! (To my knowledge, Knoppix is 100% Open Source, so all these links are legal)
Kensington Expert Mouse does this somewhat already
on
Is Horse the New Mouse?
·
· Score: 2, Interesting
My Kensington Expert Mouse does this somewhat already.
It's a large trackball (the size of a billiard ball), and your three middle fingers do curve over it to reach the buttons during normal use. It does have a thumb scrollwheel, going around the circumference of the trackball! This is a very handy feature, and lets you dial through pages extremely fast (faster than you could wiggle your middle finger using an ordinary mouse wheel). It's optical, so it's precise and doesn't have the sticky-wheels problem older trackballs used to have.
Disadvantages:
* No place to rest a wrist (the provided wrist rest is a small little joke). A folded-up old sock fixes this.
* Dodgy Windows driver. (It's marked as "beta", but really is the only choice, since the officially released driver is absolutely ancient.) It really hates my switchbox, and doesn't have any way of regaining synchronization short of rebooting the machine. Works fine in Linux, though, but all the buttons aren't recognized (there are 4 buttons).
* The trackball doesn't track fast enough when rolled at high speed, making it useless for certain applications.
Still, I like it because it does fit my fingers better, and has some of the advantages of this new "horse" mouse.
It is 128-bit. It might be IPv6-compatible on an isolated network, but I don't believe it adheres to the IPv6 global hierarchy.
InfiniBand cards are commonly multiport. In addition to the card having a unique 64-bit ID, each *port* on the card has its own 64-bit ID! These addresses are combined to form the 128-bit GID. (This is supposedly to help with such things as bonding, failover, and so on.)
More annoying than floaters are forced-registration sites. The poster child of forced-registration is nytimes, of course.
Here's a clever approach: http://www.bugmenot.com/
It's a database of voluntarily-shared registration accounts. The idea is that people don't have to give up their identity, or spend time making up false information and resubmitting until it passes validation, in order to visit these sites!
1) The == can easily be mistyped as =, introducing a nasty bug into your program.
It's harder to read, but when comparing with the == operator, I try to make sure the value on the left isn't a lvalue. For instance, (42==lue) instead of (lue==42). That way, if I mistype == as =, the compiler catches it, because you can't assign to something that's not a lvalue. Unfortunately, this won't work when both sides are lvalues.
2) The NULL resolves to a constant (0), which might get inserted into your program's code and used during the comparison. An unnecessary size bloat.
It's the same reason you want to write a loop as for(;;) instead of while(TRUE): you avoid the insertion of the TRUE constant, and unnecessary comparison thereof, into your program.
A waste - Cyrix VALUABLE for pinball
on
Cyrix Hotplate Howto
·
· Score: 2, Interesting
Bad idea to just burn up old Cyrix chips: if you have a Cyrix motherboard of just the right make and model, you can flip it on Ebay as it will be quite valuable.
The reason is that Williams Pinball made their final two pinball machines with Cyrix motherboards, before going out of business in 1999: "Revenge From Mars" and "Star Wars Episode I".
Because they went out of business before completing their plans to make the game software more portable to newer motherboards, these pinball machines work ONLY with these certain Cyrix motherboards!
The motherboard is CyrixMediaGX, BAT form factor, with the CX5520 bridge. Not CX5510, and not CX5530. CPU speed should be 233 Mhz (33x7), but 266 and 200 are also rumoured to work.
A motherboard that matches this description is quite rare these days, and sells for $300 or more -- ironically, twice the price of that motherboard when it was new!
So, if you have an old Cyrix motherboard sitting around, it just might be a gold mine, think of that before melting those chips onto a hotplate....
This happened to me too, but only for interview
on
Google Fires Blogger?
·
· Score: 3, Interesting
This happened to me too, but only for a job interview at Google. That's as far as I made it: I never worked for them or signed any NDA.
My resume was submitted, and I made it as far as the first phone screen. It was one of the best interviews I have ever had in my life! Everything went 100% great, better than I had ever hoped for. I felt we had really clicked. Then, it turns out that I lost the interview, because the interviewer read my blog.
He didn't like me talking about my job search or my experiences with Google's hiring process. He especially didn't like the way I described the interview, perhaps because it would have given future interviewees tips on what to expect. He valued his ability to "surprise" people with trick logic questions, and my description of the involved thought processes might have tipped his hand. (I've since edited my blog to remove the spoilers, as per his implied request.)
Google and Microsoft share similar cultures, evidently. Both select for candidates who are good at discovering the "a-ha" moment that enables them to see through a tricky logic puzzle and solve it. I'm not good at logic puzzles or riddles in general, but in this case, I was able to relate the puzzle to a real-life problem I faced (when writing a simulator for a particular mechanism of a pinball machine).
Lesson learned. The culture at Google is one of paranoid security, as others have confirmed with me. When interviewing (or working) there, don't reveal anything about the process. Merely mentioning the fact that you are interviewing/working might raise eyebrows. When in doubt, don't.
The good news is that the interviewer liked me, and encouraged me to re-apply. Since I seemed to learn my lesson well, he told me he wouldn't put me on the blacklist, so I've another chance. I believe the cut-off period for previous failed applicants is a year and a day.
During the time, I found a job I'm happy with now, and I'll definitely stick with it. I won't be jumping ship, in case you're reading this posting there and wondering:) The free food at Google is tempting, though....
Why not just skip directly from existing DVD to this new holographic HVD technology?
Moving from floppy to CD/DVD was a breakthrough for the computer industry, as moving from videotape to DVD was for the movie industry.
However, moving from DVD to BluRay or HDVD doesn't seem to be that much of an improvement. Consumers, especially those who are not early adopters of new technology, do tire of replacing their media every few years when a new standard disc is declared! I fear that BluRay/HDVD won't be that great of a breakthrough (HDTV movies, big deal, that can be already done by tweaking the compression on an existing double-layer DVD).
If a working HVD prototype has already been demonstrated, then why not skip the impending format war between BluRay and HDVD, and move directly to an HVD standard? It would save consumers a lot of hassle in the long run, and get us all a cool new holographic storage technology:)
I think that Linux has a unique advantage for PC games.
The entire OS could be bundled along with the game, because it's free, doesn't require hard drive installation, and doesn't take up much space. The PC game CD (or DVD) could be made self-booting, and it would automatically load and run a working Linux system from the disc, like Knoppix! It would not need to touch the hard drive at all.
Use a bootable CD, and Linux support for USB thumbdrives for game saves: instant game console:)
Advantages:
* Ability to try out the game without needing to install it first
* OS can autodetect and configure all hardware automatically, so user doesn't need to spend hours getting OpenGL working
* Developers have complete control over the OS image (no danger of Microsoft introducing DirectX incompatibilities with version updates, for instance)
* Brain-dead easy for computer novices to get the game running (just put in the CD and reboot, just like a game console)
* Allows game to be played on systems where users aren't allowed to install software or change things on the system (businesses, schools, hotels, etc.).
* Works without regard to whatever Linux distribution is installed on the system, since it doesn't touch the hard disk
* Disc can be given to children and they can play the game, without needing to install stuff all over Dad's computer
* Game state can still be saved, as Linux supports USB thumbdrives and memory cards
* As with all Linux software, no royalties or code signing needed in order to get a bootable disc that can be commercially sold
I'd love to see Linux game discs be bootable like this, so that I have the option of running the game directly from CD in addition to installing it on my hard drive. I'm probably not the only one who thinks this would be a boon to the Linux community....
Speaking of running Windows viruses... is there an existing tool to run VBS scripts under Linux?
I know, the last thing Linux needs is another scripting language, and VBS is such a horrible abomination.
The one non-viral use of VBS scripts I have seen is Visual Pinball.
There has been some talk of making a similar pinball simulation for Linux, but one of the major obstacles is that there are already hundreds of pinball games already simulated in Visual Pinball, by dozens of authors, and they all use VBS scripting (as required by Visual Pinball). This is required to interact with the PinMAME emulator (for modern games) or to implement the entire logic of the game (for old games that did not use a computer). It would be nearly impossible to get everybody to rewrite their tables again from scratch!
If Visual Pinball is ever ported to Linux, it would be necessary to write a VBS script interpreter for Linux (assuming one doesn't already exist). Heavily sandboxed, of course!
I miss PointCast. It was my grandfather's favorite application, and the "killer app" that motivated him to get a PC and dialup Internet access, back in the mid-1990's.
Having everything in one place is a great feature for those who appreciate some handholding as they use the Internet. It is also a great boon to those with slow/unreliable connections: you can just hit "Update", and it goes out and grabs all new content (auto-retrying as often as necessary). Then, you can read the content offline, taking as long as you need, without needing to be online -- a plus when in an old farmhouse with only one telephone line!
PointCast provided just enough Internet functionality to be useful: reading web pages, gathering preselected web pages from many known-good content sources, and providing the ability to track stocks and investments as well. If PointCast had included an email feature, that would be icing on the cake: my grandfather would have never used any other application.
The screensaver feature of PointCast, while not needed in this case, was also some good eye candy. It made him feel like his computer was powerful and well-connected to the world, sitting back and watching the headlines and stock prices flying by.
What I liked about PointCast is the content was always presented in a professional way: instead of the constant pop-up advertisements and demeaning "Hey, click here now!" headlines of AOL, PointCast simply presented you with everything that was available, and let you discover/consume the content at your own pace. The small square advertisement in the upper right corner was acceptable and unobtrusive. A single Post-It note, strategically placed on the monitor, did wonders:)
It would be wonderful to see another service like PointCast today, customizable via RSS feeds. It would gather information from various sites on the Internet, instead of one central clearinghouse, so it would scale better and be cheaper to run. No central server would be required, as the client would be capable of going out and connecting to the various websites directly on its own. Maybe this would be a good idea for an open source project?
That's a good idea, but won't it lose the auto-restart feature of init?
That might be actually a desirable feature, since if GDM crashes, that might be a sign of a deeper problem (misconfigured X server, etc.), and perhaps it would be best just to remain in text mode instead of repeatedly trying to restart GDM.
Back in the day, if the X server was configured, I've heard that monitors used to be damaged by Red Hat's default behaviour: GDM (and X) starts up, video card enters graphics mode, something goes wrong, X stops, video card returns to text mode, and init repeats the entire process: cheap monitors used to click and hum noisily as they were rapidly forced to change modes in an infinite loop!
My laptop takes an unacceptable five full minutes to boot up to the default Fedora Core 3 desktop. I often hit Ctrl-Alt-F1 to get a text console and do some useful work there, but even this can't be done until two-or-so minutes have elapsed (the time it takes Fedora to run all the boot scripts before the getty(8) processes are started).
The reason is that each boot script is run serially. The Red Hat boot process hasn't really changed since Red Hat 6.0 or so. The init(8) runs/etc/rc.d/rc.sysinit, then/etc/rc.d/rc for the chosen runlevel, then that script iterates through all of the scripts in/etc/rc.d/rc?.d/S??servicename, running each in turn.
This is well-understood and (reasonably) stable, but has one huge disadvantage. It's SLOW!
Because everything must be done in strict sequential order, blocking operations (like sending DHCP requests to get an IP address, or querying NTP to set the time, and so on) slow down the entire boot process.
Knoppix has a good idea, backgrounding the DHCP query process! Unfortunately, some services depending on knowing their IP address at startup time, so this wouldn't work for systems that run servers.
I'd love to see a more general method used to run as many setup tasks in parallel as possible. I would love to see a keyword-based provides/requires system in the startup scripts, similar to what is commonly used elsewhere to represent dependencies (Makefile targets, RPM keywords, and so on).
If each of the startup scripts would declare capabilities they provide, as well as capabilities they require to be present already, then the startup process can be made smarter, running as many scripts in parallel as possible. It would be similar to how make(1) runs with the -j option, paralellizing, but something other than the actual "make" program would need to be used, to be as error-tolerant as possible (since the goal is to provide you with a useable system, not to detect build failures during software development).
I've never used BeOS but heard that it made strides towards improving boot time, by doing as much in parallel as possible. Maybe there are some good ideas there that we could pick up.
I've often thought about how a "StreamTorrent" protocol would work. This technique goes a long way towards solving the problem, but still has a way to go.
What of live radio/TV?
An "ordered BitTorrent" design, as described in this paper, could work if it's a static file being streamed. People watching late in the clip would send data they've already seen to newcomers who are watching early in the clip. However, this would not work for live content. There would be no "early" or "late" listeners, as everybody would be receiving content at the same point in time!
Also, there is another concern. If content is being broadcast live, it would most likely originate from a single source. It will become desirable to get as close as possible to the source, to get data that has passed through as few hands as possible, for reasons of timeliness and reliability. Getting and holding onto a low-generation connection would be desirable. BitTorrent's ranking/evaluation system would need to be extended in order to sort out who gets the scarce low-generation connections.
There are other issues, but these two come to the forefront. Perhaps long-term buffering (TiVo-style) could be used to allieviate some of these concerns. Listeners of live content, if their connection is judged inadequate to receive a live stream (or they just didn't get in early enough), would receive a "tape-delayed" stream from other listeners. Clients would be required to save the past hour or so of live content, so that an appropriate delay could be assigned to each new listener, giving the data a chance to percolate throughout the system. Content would eventually fan out throughout the network, going from low-generation to high-generation connections.
It's a toss-up: receiving live data, as direct as posssible from the source, is the best for timely content (and less chance of getting a degraded connection caused by somebody else refusing to send you data), but receiving time-buffered content is best for overall reliability (giving you a long time to retry and successfully re-download the data from different sources before your playback buffer starts to drop out).
Surprisingly, the large streaming vendors (RealPlayer, Windows Media Player, etc.) are surprisingly stuck in the mud: their recent innovations in playback protocols are slow or nonexistent. For the most part, they're still using a single basic TCP or UDP connection! A heroic third-party plugin will have to be the answer. I look forward to seeing "StreamTorrent" show up in RealAudio 20.0 or Windows Media Player 30.0 or so:)
To check out the current (sad) state of streaming online radio, check my signature below, and try to listen to Dr. Demento on Sunday nights when the servers overload....:(
I've owned nearly a dozen Pentium-II systems over the years (they just keep falling into my hands somehow), and I must say they seem to represent the height of PC stability and reliability. The Intel 440BX chipset has been rock-solid.
Motherboard designs around the time of the Pentium-II were at a fairly stable point in PC history. PCI and AGP were established, SDRAM modules were getting standardized and more affordable, and USB was getting ironed out. I've seen newer systems crash more often, as they seem to be made less robust (stricter thermal requirements, newer designs, mismatched components, and so on).
The Pentium-II CPU and Intel 440BX chipset seem to be a stable rock-solid combination, and this platform provides just enough speed to be useful for casual purposes (web browser, email, word processor, MP3 player, and so on). It makes an excellent "second PC" to have around the house as a spare. So, these systems should continue to be common and perform well for years to come!
1) Any vehicle with an ignition system that can't be turned off the same way it was turned on -- from Renault's new smartcard ignition system, to an old-style lawnmower with a pull cord -- needs a KILL SWITCH! Pressing the kill switch shuts off the flow of gas/spark/whatever to the engine, causing it to stall, and would not have any other side effects such as re-engaging the steering column lock.
2) Pedals usually have different shapes: a horizontal rectangle for the brake (or clutch), and a vertical rectangle for the throttle. These are shaped differently on purpose: when strongly pushed, you should feel the difference in your feet as your shoes bend around the pedals differently! Often, the pedals will even be at different heights: notice the gas pedal is lower than the brake, on most cars.
3) Say what you will about our overzealous product safety/testing/defect laws, but there's a reason Renault cars aren't sold anymore in the US:)
This would be great for a subnotebook PC, like the Toshiba Libretto.
It's good progress, but still not what I ideally want: a device that is completely a standard PC, but in the form factor of a Gameboy Advance SP! The GBA SP is an ideal design: it is small, and has a flip-up screen that is sturdy and locks into a convenient position. If small buttons were used to form a fully-functional keyboard, as on the BlackBerry, and a small touchpad nearby for the mouse, this would be wonderful to have! The PC would have decent 3D graphics for games compatibility, a micro hard drive similar to the iPod, and of course Wi-Fi and GPRS wireless data support.
Does any company want to merge this pocketful of present-day devices into a future dream machine that could do it all?
Yes, it does. This is endemic of the latest version of Macrovision's "SafeDisc" protection scheme: blacklist certain programs, and refuse to load the game if any of them are detected.
What's a shame is that this only hurts legitimate CD mastering programs, such as Nero and Easy CD Creator! DAEMON Tools, the CD emulator most often used to play pirated games, comes out with a new version within days (hours?) of the game being released -- it has slightly changed strings within the executable/driver/etc. to foil the blacklist, so the only people hurt by Macrovision's "protection" are the legitimate customers who also have commercial CD mastering software installed on their PC!
This is ridiculous, and another good example of why copy protection is a failure.
I get 3000 spams/day with my catchall address (krellan.com)!
I will soon be putting in a whitelist of allowed usernames, and bouncing everything else, in hopes of reducing this ridiculously high spamcount.
I get only 100 spams/day correctly addressed to my real addresses that I use on that domain.
This is still too high, but a combination of SpamAssassin on the server and Bayesian filtering on the client (Mozilla Thunderbird) help reduce the number of spams I actually see to almost none! (For safety and in case of false positives, all emails are still archived.)
Slashdot Mirror
When I was in college (graduated 1998), I took a software-only degree.
If I had it to do over again, I would have taken both hardware and software.
There have been times I have wanted to know what happens at the hardware level, and have been baffled by it. There is also the practical side of it: I don't like staring at a broken circuit board and not knowing how to troubleshoot and repair it.
Also, as someone else mentioned, there's the practical nature of hardware: it can't easily be outsourced, as it must often be babysat in person....
Wow, nice! I wish I had known about this, before I installed the normal RealPlayer....
The audio program CarTalk issued a similar ultimatum several months ago, and that resulted in a different RealPlayer download link.
http://www.real.com/freeplayer/?rppr=cartalk.com
It's easier (and friendlier) to download the RealPlayer, but it results in the same spyware-infested client.
What does it say about RealPlayer when their biggest clients are threatening to rebel, because of frustrations caused by their listeners trying to use Real's spyware-infested software? It's sad that Real is following the KaZaA "business model" by selling out to spyware.
(Can't resist a link to my original rant about RealPlayer 8. I'm glad that version was their nadir, and they've been (slowly) improving since then.)
As far as MP3 hitting critical mass goes, isn't it already there?
Doesn't everybody already have some kind of hardware or software these days that can play the MP3 format?
What I want to see is audiobooks sold in MP3 CD format. Currently, audiobooks take a huge number of tapes (or conventional CD's) to hold audio, and because it's spoken word, it would compress very easily while still remaining high quality.
I've converted the Harry Potter books to MP3 CD's, and they sound great. The first two books even fit together on a single CD! All this was from a large stack of grossly inefficient standard audio CD's. No need to store spoken word at 44100x16x2 uncompressed!
I'm hoping the makers of audiobooks will realize that MP3 has already hit critical mass for years now....
I have a Swords Of Fury machine. It only has 3 balls installed, for a 3-ball multiball.
The flippers at the bottom aren't unusually close. Maybe what you're thinking of is the post that's placed dead center to protect balls from draining straight down the middle (as often they drop from the center ramp).
Never did find an infinite-life bug, if there was one, and the flippers never were powerful enough to actually get over the walls (I've never heard of this happening).
Swords Of Fury is considered a very easy game, though, and it's fairly easy to play for as long as you want (if you can keep getting multiballs)....
Maybe you were thinking of another game?
I was thinking about the recent conflict between municipal and commercial Internet access providers. Government-owned Internet access that blankets an area, whether it be via wireless or fiber-optic cable or something equally useful, is a great thing to have from a property-values and quality-of-life standpoint (hence the city's motivation to build it). However, it obviously competes against commercial ISP's, which (rightfully) feel undercut by the government.
How about this compromise:
Government-owned ISP's focus on doing what they do best: building out infrastructure and reaching through the unprofitable "last mile" to get all customers -- not just the most desirable large corporations or dense urban populations -- hooked up. The purpose of this infrastructure would be to transparently pass packets through, acting as a pipe between their customers and the Internet.
Commercial ISP's focus on providing content and configuring the network to deliver that content. IP address assignment (DHCP), provisioning of subnets, useful servers (email, web hosting, newsgroups), etc. would be handled by competing commercial ISP's. All the various levels and varieties of access found today would still be available: customers could choose to pay extra for a static IP, or certain premium content, or whatever else strikes their fancy. All of this content would be delivered via the municipal infrastructure! Both commercial and government play useful parts here.
Do you think this compromise would actually work in practice? I'd love to see it given a chance somewhere....
Yes, the primary reason for the desirability of thin clients in the workplace is their lack of an administration burden.
Being able to save money by purchasing less powerful hardware isn't even close: over the useful life of the system, the cost of administration more than outweighs the one-time expense of the hardware. In fact, bizarre CPU-less hardware (some dedicated thin client "terminals") can cost *more* than an equivalent standard PC, because of lack of an economy of scale!
Most corporations just ride the price-performance curve and purchase whatever standard PC is currently in the middle of the market. These days, that gets a decently powerful CPU. Thin clients become thin due to the use of network booting and non-local file storage, not due to the lack of a CPU! When properly configured, a standard PC makes a good thin client.
What would be a wonderful use of resources is to form some kind of cluster/grid out of all of the thin clients! The powerful local CPU could then be used for jobs that are submitted over the network, giving a powerful large cluster for heavy computation.
The user would use the PC as a thin client, running remote applications, and any unused local CPU power would still go to good use, powering the cluster! To me, this seems like a win-win situation....
Yes, it IS on BitTorrent now.
3 .8CeBitTorrent/KNOPPIX_V3.8-2005-02-28-CeBIT_Editi on.iso.torrent
3 .8CeBitTorrent
i xx/KNOPPIX_V3.8-2005-02-28-CeBIT_Edition.iso
http://stacktrace.org/index_html/20050313-Knoppix
Here's the webpage this link comes from:
http://stacktrace.org/index_html/20050313-Knoppix
And, here is a (very slow, may be down now) conventional download:
http://dl1.gamerznet.org/vollversionen/unix/knopp
Here's the webpage that it came from (beware, heavy advertisements/popups, better get Adblock for your Firefox):
http://www.winfuture.de/news,19477.html
Enjoy! (To my knowledge, Knoppix is 100% Open Source, so all these links are legal)
My Kensington Expert Mouse does this somewhat already.
It's a large trackball (the size of a billiard ball), and your three middle fingers do curve over it to reach the buttons during normal use. It does have a thumb scrollwheel, going around the circumference of the trackball! This is a very handy feature, and lets you dial through pages extremely fast (faster than you could wiggle your middle finger using an ordinary mouse wheel). It's optical, so it's precise and doesn't have the sticky-wheels problem older trackballs used to have.
Disadvantages:
* No place to rest a wrist (the provided wrist rest is a small little joke). A folded-up old sock fixes this.
* Dodgy Windows driver. (It's marked as "beta", but really is the only choice, since the officially released driver is absolutely ancient.) It really hates my switchbox, and doesn't have any way of regaining synchronization short of rebooting the machine. Works fine in Linux, though, but all the buttons aren't recognized (there are 4 buttons).
* The trackball doesn't track fast enough when rolled at high speed, making it useless for certain applications.
Still, I like it because it does fit my fingers better, and has some of the advantages of this new "horse" mouse.
InfiniBand uses something called a "GID" for a hardware address, similar to a MAC address.
1 internettech2.html
http://www.iw.com/magazine.php?inc=060101/06.01.0
It is 128-bit. It might be IPv6-compatible on an isolated network, but I don't believe it adheres to the IPv6 global hierarchy.
InfiniBand cards are commonly multiport. In addition to the card having a unique 64-bit ID, each *port* on the card has its own 64-bit ID! These addresses are combined to form the 128-bit GID. (This is supposedly to help with such things as bonding, failover, and so on.)
More annoying than floaters are forced-registration sites. The poster child of forced-registration is nytimes, of course.
:)
Here's a clever approach:
http://www.bugmenot.com/
It's a database of voluntarily-shared registration accounts. The idea is that people don't have to give up their identity, or spend time making up false information and resubmitting until it passes validation, in order to visit these sites!
BTW, my nytimes account is nospam2/nospam2
Problems with using (ptr==NULL):
1) The == can easily be mistyped as =, introducing a nasty bug into your program.
It's harder to read, but when comparing with the == operator, I try to make sure the value on the left isn't a lvalue. For instance, (42==lue) instead of (lue==42). That way, if I mistype == as =, the compiler catches it, because you can't assign to something that's not a lvalue. Unfortunately, this won't work when both sides are lvalues.
2) The NULL resolves to a constant (0), which might get inserted into your program's code and used during the comparison. An unnecessary size bloat.
It's the same reason you want to write a loop as for(;;) instead of while(TRUE): you avoid the insertion of the TRUE constant, and unnecessary comparison thereof, into your program.
Bad idea to just burn up old Cyrix chips: if you have a Cyrix motherboard of just the right make and model, you can flip it on Ebay as it will be quite valuable.
The reason is that Williams Pinball made their final two pinball machines with Cyrix motherboards, before going out of business in 1999: "Revenge From Mars" and "Star Wars Episode I".
Because they went out of business before completing their plans to make the game software more portable to newer motherboards, these pinball machines work ONLY with these certain Cyrix motherboards!
The motherboard is Cyrix MediaGX, BAT form factor, with the CX5520 bridge. Not CX5510, and not CX5530. CPU speed should be 233 Mhz (33x7), but 266 and 200 are also rumoured to work.
A motherboard that matches this description is quite rare these days, and sells for $300 or more -- ironically, twice the price of that motherboard when it was new!
So, if you have an old Cyrix motherboard sitting around, it just might be a gold mine, think of that before melting those chips onto a hotplate....
This happened to me too, but only for a job interview at Google. That's as far as I made it: I never worked for them or signed any NDA.
:) The free food at Google is tempting, though....
My resume was submitted, and I made it as far as the first phone screen. It was one of the best interviews I have ever had in my life! Everything went 100% great, better than I had ever hoped for. I felt we had really clicked. Then, it turns out that I lost the interview, because the interviewer read my blog.
He didn't like me talking about my job search or my experiences with Google's hiring process. He especially didn't like the way I described the interview, perhaps because it would have given future interviewees tips on what to expect. He valued his ability to "surprise" people with trick logic questions, and my description of the involved thought processes might have tipped his hand. (I've since edited my blog to remove the spoilers, as per his implied request.)
Google and Microsoft share similar cultures, evidently. Both select for candidates who are good at discovering the "a-ha" moment that enables them to see through a tricky logic puzzle and solve it. I'm not good at logic puzzles or riddles in general, but in this case, I was able to relate the puzzle to a real-life problem I faced (when writing a simulator for a particular mechanism of a pinball machine).
Lesson learned. The culture at Google is one of paranoid security, as others have confirmed with me. When interviewing (or working) there, don't reveal anything about the process. Merely mentioning the fact that you are interviewing/working might raise eyebrows. When in doubt, don't.
The good news is that the interviewer liked me, and encouraged me to re-apply. Since I seemed to learn my lesson well, he told me he wouldn't put me on the blacklist, so I've another chance. I believe the cut-off period for previous failed applicants is a year and a day.
During the time, I found a job I'm happy with now, and I'll definitely stick with it. I won't be jumping ship, in case you're reading this posting there and wondering
Why not just skip directly from existing DVD to this new holographic HVD technology?
:)
Moving from floppy to CD/DVD was a breakthrough for the computer industry, as moving from videotape to DVD was for the movie industry.
However, moving from DVD to BluRay or HDVD doesn't seem to be that much of an improvement. Consumers, especially those who are not early adopters of new technology, do tire of replacing their media every few years when a new standard disc is declared! I fear that BluRay/HDVD won't be that great of a breakthrough (HDTV movies, big deal, that can be already done by tweaking the compression on an existing double-layer DVD).
If a working HVD prototype has already been demonstrated, then why not skip the impending format war between BluRay and HDVD, and move directly to an HVD standard? It would save consumers a lot of hassle in the long run, and get us all a cool new holographic storage technology
I think that Linux has a unique advantage for PC games.
:)
The entire OS could be bundled along with the game, because it's free, doesn't require hard drive installation, and doesn't take up much space. The PC game CD (or DVD) could be made self-booting, and it would automatically load and run a working Linux system from the disc, like Knoppix! It would not need to touch the hard drive at all.
Use a bootable CD, and Linux support for USB thumbdrives for game saves: instant game console
Advantages:
* Ability to try out the game without needing to install it first
* OS can autodetect and configure all hardware automatically, so user doesn't need to spend hours getting OpenGL working
* Developers have complete control over the OS image (no danger of Microsoft introducing DirectX incompatibilities with version updates, for instance)
* Brain-dead easy for computer novices to get the game running (just put in the CD and reboot, just like a game console)
* Allows game to be played on systems where users aren't allowed to install software or change things on the system (businesses, schools, hotels, etc.).
* Works without regard to whatever Linux distribution is installed on the system, since it doesn't touch the hard disk
* Disc can be given to children and they can play the game, without needing to install stuff all over Dad's computer
* Game state can still be saved, as Linux supports USB thumbdrives and memory cards
* As with all Linux software, no royalties or code signing needed in order to get a bootable disc that can be commercially sold
I'd love to see Linux game discs be bootable like this, so that I have the option of running the game directly from CD in addition to installing it on my hard drive. I'm probably not the only one who thinks this would be a boon to the Linux community....
Speaking of running Windows viruses... is there an existing tool to run VBS scripts under Linux?
I know, the last thing Linux needs is another scripting language, and VBS is such a horrible abomination.
The one non-viral use of VBS scripts I have seen is Visual Pinball.
There has been some talk of making a similar pinball simulation for Linux, but one of the major obstacles is that there are already hundreds of pinball games already simulated in Visual Pinball, by dozens of authors, and they all use VBS scripting (as required by Visual Pinball). This is required to interact with the PinMAME emulator (for modern games) or to implement the entire logic of the game (for old games that did not use a computer). It would be nearly impossible to get everybody to rewrite their tables again from scratch!
If Visual Pinball is ever ported to Linux, it would be necessary to write a VBS script interpreter for Linux (assuming one doesn't already exist). Heavily sandboxed, of course!
I miss PointCast. It was my grandfather's favorite application, and the "killer app" that motivated him to get a PC and dialup Internet access, back in the mid-1990's.
:)
Having everything in one place is a great feature for those who appreciate some handholding as they use the Internet. It is also a great boon to those with slow/unreliable connections: you can just hit "Update", and it goes out and grabs all new content (auto-retrying as often as necessary). Then, you can read the content offline, taking as long as you need, without needing to be online -- a plus when in an old farmhouse with only one telephone line!
PointCast provided just enough Internet functionality to be useful: reading web pages, gathering preselected web pages from many known-good content sources, and providing the ability to track stocks and investments as well. If PointCast had included an email feature, that would be icing on the cake: my grandfather would have never used any other application.
The screensaver feature of PointCast, while not needed in this case, was also some good eye candy. It made him feel like his computer was powerful and well-connected to the world, sitting back and watching the headlines and stock prices flying by.
What I liked about PointCast is the content was always presented in a professional way: instead of the constant pop-up advertisements and demeaning "Hey, click here now!" headlines of AOL, PointCast simply presented you with everything that was available, and let you discover/consume the content at your own pace. The small square advertisement in the upper right corner was acceptable and unobtrusive. A single Post-It note, strategically placed on the monitor, did wonders
It would be wonderful to see another service like PointCast today, customizable via RSS feeds. It would gather information from various sites on the Internet, instead of one central clearinghouse, so it would scale better and be cheaper to run. No central server would be required, as the client would be capable of going out and connecting to the various websites directly on its own. Maybe this would be a good idea for an open source project?
RIP PointCast and RIP my grandfather....
That's a good idea, but won't it lose the auto-restart feature of init?
That might be actually a desirable feature, since if GDM crashes, that might be a sign of a deeper problem (misconfigured X server, etc.), and perhaps it would be best just to remain in text mode instead of repeatedly trying to restart GDM.
Back in the day, if the X server was configured, I've heard that monitors used to be damaged by Red Hat's default behaviour: GDM (and X) starts up, video card enters graphics mode, something goes wrong, X stops, video card returns to text mode, and init repeats the entire process: cheap monitors used to click and hum noisily as they were rapidly forced to change modes in an infinite loop!
My laptop takes an unacceptable five full minutes to boot up to the default Fedora Core 3 desktop. I often hit Ctrl-Alt-F1 to get a text console and do some useful work there, but even this can't be done until two-or-so minutes have elapsed (the time it takes Fedora to run all the boot scripts before the getty(8) processes are started).
/etc/rc.d/rc.sysinit, then /etc/rc.d/rc for the chosen runlevel, then that script iterates through all of the scripts in /etc/rc.d/rc?.d/S??servicename, running each in turn.
The reason is that each boot script is run serially. The Red Hat boot process hasn't really changed since Red Hat 6.0 or so. The init(8) runs
This is well-understood and (reasonably) stable, but has one huge disadvantage. It's SLOW!
Because everything must be done in strict sequential order, blocking operations (like sending DHCP requests to get an IP address, or querying NTP to set the time, and so on) slow down the entire boot process.
Knoppix has a good idea, backgrounding the DHCP query process! Unfortunately, some services depending on knowing their IP address at startup time, so this wouldn't work for systems that run servers.
I'd love to see a more general method used to run as many setup tasks in parallel as possible. I would love to see a keyword-based provides/requires system in the startup scripts, similar to what is commonly used elsewhere to represent dependencies (Makefile targets, RPM keywords, and so on).
If each of the startup scripts would declare capabilities they provide, as well as capabilities they require to be present already, then the startup process can be made smarter, running as many scripts in parallel as possible. It would be similar to how make(1) runs with the -j option, paralellizing, but something other than the actual "make" program would need to be used, to be as error-tolerant as possible (since the goal is to provide you with a useable system, not to detect build failures during software development).
I've never used BeOS but heard that it made strides towards improving boot time, by doing as much in parallel as possible. Maybe there are some good ideas there that we could pick up.
I've often thought about how a "StreamTorrent" protocol would work. This technique goes a long way towards solving the problem, but still has a way to go.
What of live radio/TV?
An "ordered BitTorrent" design, as described in this paper, could work if it's a static file being streamed. People watching late in the clip would send data they've already seen to newcomers who are watching early in the clip. However, this would not work for live content. There would be no "early" or "late" listeners, as everybody would be receiving content at the same point in time!
Also, there is another concern. If content is being broadcast live, it would most likely originate from a single source. It will become desirable to get as close as possible to the source, to get data that has passed through as few hands as possible, for reasons of timeliness and reliability. Getting and holding onto a low-generation connection would be desirable. BitTorrent's ranking/evaluation system would need to be extended in order to sort out who gets the scarce low-generation connections.
There are other issues, but these two come to the forefront. Perhaps long-term buffering (TiVo-style) could be used to allieviate some of these concerns. Listeners of live content, if their connection is judged inadequate to receive a live stream (or they just didn't get in early enough), would receive a "tape-delayed" stream from other listeners. Clients would be required to save the past hour or so of live content, so that an appropriate delay could be assigned to each new listener, giving the data a chance to percolate throughout the system. Content would eventually fan out throughout the network, going from low-generation to high-generation connections.
It's a toss-up: receiving live data, as direct as posssible from the source, is the best for timely content (and less chance of getting a degraded connection caused by somebody else refusing to send you data), but receiving time-buffered content is best for overall reliability (giving you a long time to retry and successfully re-download the data from different sources before your playback buffer starts to drop out).
Surprisingly, the large streaming vendors (RealPlayer, Windows Media Player, etc.) are surprisingly stuck in the mud: their recent innovations in playback protocols are slow or nonexistent. For the most part, they're still using a single basic TCP or UDP connection! A heroic third-party plugin will have to be the answer. I look forward to seeing "StreamTorrent" show up in RealAudio 20.0 or Windows Media Player 30.0 or so :)
To check out the current (sad) state of streaming online radio, check my signature below, and try to listen to Dr. Demento on Sunday nights when the servers overload.... :(
I've owned nearly a dozen Pentium-II systems over the years (they just keep falling into my hands somehow), and I must say they seem to represent the height of PC stability and reliability. The Intel 440BX chipset has been rock-solid.
Motherboard designs around the time of the Pentium-II were at a fairly stable point in PC history. PCI and AGP were established, SDRAM modules were getting standardized and more affordable, and USB was getting ironed out. I've seen newer systems crash more often, as they seem to be made less robust (stricter thermal requirements, newer designs, mismatched components, and so on).
The Pentium-II CPU and Intel 440BX chipset seem to be a stable rock-solid combination, and this platform provides just enough speed to be useful for casual purposes (web browser, email, word processor, MP3 player, and so on). It makes an excellent "second PC" to have around the house as a spare. So, these systems should continue to be common and perform well for years to come!
Lessons learned:
:)
1) Any vehicle with an ignition system that can't be turned off the same way it was turned on -- from Renault's new smartcard ignition system, to an old-style lawnmower with a pull cord -- needs a KILL SWITCH! Pressing the kill switch shuts off the flow of gas/spark/whatever to the engine, causing it to stall, and would not have any other side effects such as re-engaging the steering column lock.
2) Pedals usually have different shapes: a horizontal rectangle for the brake (or clutch), and a vertical rectangle for the throttle. These are shaped differently on purpose: when strongly pushed, you should feel the difference in your feet as your shoes bend around the pedals differently! Often, the pedals will even be at different heights: notice the gas pedal is lower than the brake, on most cars.
3) Say what you will about our overzealous product safety/testing/defect laws, but there's a reason Renault cars aren't sold anymore in the US
This would be great for a subnotebook PC, like the Toshiba Libretto.
It's good progress, but still not what I ideally want: a device that is completely a standard PC, but in the form factor of a Gameboy Advance SP! The GBA SP is an ideal design: it is small, and has a flip-up screen that is sturdy and locks into a convenient position. If small buttons were used to form a fully-functional keyboard, as on the BlackBerry, and a small touchpad nearby for the mouse, this would be wonderful to have! The PC would have decent 3D graphics for games compatibility, a micro hard drive similar to the iPod, and of course Wi-Fi and GPRS wireless data support.
Does any company want to merge this pocketful of present-day devices into a future dream machine that could do it all?
Yes, it does. This is endemic of the latest version of Macrovision's "SafeDisc" protection scheme: blacklist certain programs, and refuse to load the game if any of them are detected.
What's a shame is that this only hurts legitimate CD mastering programs, such as Nero and Easy CD Creator! DAEMON Tools, the CD emulator most often used to play pirated games, comes out with a new version within days (hours?) of the game being released -- it has slightly changed strings within the executable/driver/etc. to foil the blacklist, so the only people hurt by Macrovision's "protection" are the legitimate customers who also have commercial CD mastering software installed on their PC!
This is ridiculous, and another good example of why copy protection is a failure.
I get 3000 spams/day with my catchall address (krellan.com)!
I will soon be putting in a whitelist of allowed usernames, and bouncing everything else, in hopes of reducing this ridiculously high spamcount.
I get only 100 spams/day correctly addressed to my real addresses that I use on that domain.
This is still too high, but a combination of SpamAssassin on the server and Bayesian filtering on the client (Mozilla Thunderbird) help reduce the number of spams I actually see to almost none! (For safety and in case of false positives, all emails are still archived.)