People seem to be spending a lot of time concentrating on this guy's motivation.
I really don't understand why. Why should I care what his motivation is? The point is that he got the USPTO to re-examine a dumb patent. If it did it for the publicity or to hawk his blog or because he thinks it'll speed up the Second Coming of Jesus, I don't really care. Hooray for him, in any case; the point is that the action he took was good, irrespective of the motive.
I'm not going to fault him because he was doing it for publicity. Maybe if a few more people did what he's doing -- for publicity or whatever reason -- we'd have a few less shitty patents floating around.
In fact, if you want to hurt a company these days, you're better spent going after their patents than trying to take them to court directly, especially small-claims. So if there's a company you don't like, and you can get the dough together to force a re-examination of their patents, by all means please do so. I won't question why you're doing it, the point is that it gets done.
As long as we're tossing around the Flamebait here, I'd like to disagree with this statement:
The worst ones are the fuckers that voted for him in the 2nd round, and now are all disappointed and disaproving.
No, the worst ones are all the people who voted for that mechanoid John Kerry in the Democratic primaries, over other people who might have actually -- oh, I don't know, had some sort of impact on the American public? Instead they picked a guy whose only strength was that he "wasn't named Bush," and made G.W. look well-spoken and animated by comparison.
The Democratic party snatched defeat from the jaws of victory in that race, by not having the balls to pick anyone other than Kerry.
Well, tape drives that have been around for a few years are cheap. It's new tape drives that are expensive.
For example, during one of my backup obsessions, I got a HP SureStore Tape 12000e for about $25 on eBay. It's a DDS-2 autoloader. Holds six 4GB (native, allegedly 8GB compressed) tapes in a magazine and changes them on command.
A few years ago, such a device would have set me back a few thousand bucks. Now, because people have moved to DDS-4 and higher, they're dirt cheap. There's nothing wrong with them, it's just an issue of capacity. (Though I think most people could probably fit their most important stuff in less than 8GB, counting documents only, no MP3s, warez, porn, applications, or system files, and backing up things like digital photos separately.)
Compared to hard drives of similar capacities, tape drives have always been expensive from a home user's perspective, at least in the recent past. I suspect they'll stay this way in the future: the economies of scale are simply not there to make tape drives as inexpensive to manufacture as HDs. However, it's not hard to find ones that are a generation or two old for rock-bottom prices and still have a lot of life left in them.
Except that they really don't. The cost differential between a comparably equipped Mac and PC is pretty small, and depending on how you spec things out (and the prices of commodity equipment at a particular time) sometimes the low-end Macs even come out ahead. I think when the Minis were first brought out, they were cheaper than a comparably-spec'd major-manufacturer PC. At any rate, when you take that increase in cost and amortize it out across the multi-year timespan that you're probably going to own a computer, it's fairly small. At least that's how I see it: every time I buy a computer, particularly my main desktop, I try to think about all the time I'm going to be spending in front of it over the next few years. It makes sense to buy what I'm going to be really happy with, even if it costs more initially.
I've used Macs for years, and there was a time when the cost difference was a lot more pronounced (right after the prices on commodity hardware started to collapse), but that's not the case so much anymore. While there will always be people who want to compare a $500 Mac to a $150 Wal-Mart PC (or back a few years ago, a $1k Mac to a $500 Office Max special), but that's not a valid comparison or argument.
Furthermore, if you're planning on making a living off of some type of computer skill, it doesn't make sense to cut corners on your tools.
It's Sun that would be putting their "weight" on Ubuntu, at least if you want to extrapolate from what TFA says, which really isn't all that convincing.
While we're on the subject, does anyone know what "Public key encryption support (RSA)" means, as a feature of a CPU?
Some sort of special on-chip hardware, or is it just a 'marketing feature' with no real relation to anything on the die?
A quick Google found only one reference to it, here, which suggests to me that might be for real, and has something to do with using it for SSL/HTTPS workloads. Anyone (someone who has perhaps read the Architecture Specifications, which I admit I have not) want to clue me in on the straight dope?
I use Ubuntu (actually Kubuntu) on my Linux desktop machine, but I use straight Debian on my headless server.
Can anyone tell me why a person would want to use Ubuntu on a server, as opposed to just using Debian?
It seems to me that most of the advantages of Ubuntu are on the GUI side of things, and this is the way that most of the software that's different for Ubuntu than Debian is aimed towards. Most of the server-type packages you'd probably be pulling from the Debian repositories anyway, so there's not much advantage and some things might not work, because Ubuntu doesn't follow the "Debian way" in everything (there are some file locations and paths that are different, I believe). Plus Debian has always seemed a bit better documented, although I admit that's arguable.
I'm glad to see Sun put its weight behind a Debian-based distro, but I don't quite get why Ubuntu and not just Debian, especially if it's for servers. The only reason I can think is that they don't want to get too close to Debian's leadership and philosophy, and find Ubuntu more palatable from a PR and customer-relations perspective. Still, it seems like an odd choice.
Idiotic moderation... your comment was the best explanation I've gotten in this thread so far as to what's going on and what is and isn't closed.
I can't really think of why anyone would want to run Darwin x86 without OS X either; we've already established that it's a worse server platform than Linux for most tasks, especially database ones, and headless servers are really the only place I think there'd be a market for Darwin. And it's not like there's any dearth of server OSes and distos these days anyway. The only other people are those who want to create a platform on which to do unauthorized ports of OS X onto commodity hardware (say by hacking the kernel to remove the hardware verification portions, and creating a foundation on which to run the proprietary portions of the Mac OS).
I figured this was inevitable all along. In fact, back when people were cooing over how folks had gotten OS X to boot on commodity hardware, I speculated that it was going to drive Apple to close up more and more of its OS, and I think if it continues, we're going to see a lot of phone-home type registration systems. To be perfectly honest, as someone who's always appreciated the fact that Mac OS has never had copy protection (because it depends on having a rather largish dongle, called a Mac), I would rather see Apple do what it needs to do to head off commodity ports with licensing than have them start to include obnoxious copy-protection in the OS itself that bothers legitimate users, a la Microsoft.
Normally I'm all for technological solutions rather than legal ones, but in this case the technological ones are going to be much more of a pain in my ass, so I'd appreciate it if they didn't. The day I have to type a serial number into Mac OS X so that it can phone home to Apple, I'm going to be pretty annoyed.
Why bother to be diplomatic? I mean, there have to be some empty seats on those CIA black-bag extradition flights; when you're taking terrorists to Russia to have their toenails ripped out, pay the right people and I'm sure you could get a few spammers to fill the seats on the return leg.
Criminals turn each other in for cash all the time. I'm sure if you threw around some hard currency you could get them to rat each other out, it probably wouldn't even cost that much compared to the actual cost of spam management in the U.S. All you'd need to do is round them up and take them to some place where sending viagra advertisements is a capital offense.
Or save the airfare and just shoot them and throw them into the North Sea, it doesn't matter a whole lot as long as other people in their organization know (or have a vague idea) what happened to them.
I'm joking here, but only partially. I think the sum of all the inconvenience that these people (spammers) cause greatly outweighs whatever their lives are possibly worth to humanity, and I don't see any reason why we should continue to allow them to suck down air.
I tend to wonder if the startup time of Office and its RAM requirements aren't being hidden because it's tied into the OS so deeply that it's really sharing resources with it.
Kind of like how IE seems to start up faster than Firefox, but only because a lot of the stuff that IE uses has already been preloaded by Windows, and therefore it has an unfair advantage (because nobody else but Microsoft can take advantage of this sort of thing).
The way Microsoft intertwines its products basically makes comparisons to anyone else's stuff very difficult, because it's hard to say whether you have half of the MS program already loaded just by turning the computer on.
Email is the killer app of the business world. I don't know any organization these days that doesn't live and breathe email.
Notes purports to do, among other things, email. And it does this poorly. So basically, in implementing Notes, you take away users' other email programs, and replace it with something that does a lot of stuff, but doesn't do the ONE THING that they want it to do, very well. Can you blame them for hating it? I sure can't.
The fact that you can develop great database-driven, collaborative applications on it doesn't matter one whit to 99% of its userbase, if the email features suck.
Any program which replaces users' existing email programs, must do email at least as well as what they replace, or it will be hated, regardless of what other features it offers. I don't care if the program makes coffee and massages my feet, I'd still hate it because it's my only option for email, and as an email client it's lame.
Oh Notes, how I hate thee, let me count the ways...
1. 90% of people use it for an email client, which is kind of like using Oracle's RDBMS to manage your grocery list. Or maybe a better analogy is, it's like using a Swiss Army knife to carve a marble sculpture. There are many things that swiss army knife is good for, but when all you want to do is carve marble, it's not the best tool for the job. Eventually it just leaves you wanting to find the person who gave you said pocketknife, and stab them with the little corkscrew, in the hopes that next time they'll give you a proper chisel.
2. Searching sucks. I mean, just sucks. I don't even bother using it, it's that bad. If you can search for stuff in the contents/body of emails, I've never figured out how, and even searching by subject/sender etc. is a PITA. You have to be in the main list view of emails for the folder you want to search, and then you have to click on the column of the term you want to search (e.g., subject). Then you do Find or start typing a few letters, which opens a dialog box into which you type your term. Then you press return. If you're lucky, an email will be highlighted in the view. If you're not lucky, nothing will appear to happen, but if you look down at the veeeeeery bottom of the window, they'll be a line that says "'foo' was not found." As far as I know, you can't use any sort of wildcards, expansions, or regexps in search, and it's totally literal: if you type 'Bob Jones', and the guy's name on the email is 'Bob T. Jones,' you won't find it. This is absolutely braindead for an email application, especially one that's as database-centric as Notes purports to be. I don't know about Exchange, but Apple Mail is better than this (stores everything as flat text files, too), as is GMail and every other webmail system I've ever used.
3. Everything runs in one gigantic window. If you can tear off message windows so that they're not all running as tabs, I've never figured out how. If you don't like it taking up your entire screen, your only option is to restore down, then resize the window and scroll. People really aren't that far off when they're joking that it seems like Lotus is about two lines of code away from becoming a full-fledged OS: it seems designed under the assumption that it's the only application you'll use. Ever. For everything. It's email, it's groupware, it's calendaring, it's a CMS/version-control system, hell, it's even a web browser. Why? No idea! And now it's going to be a word processor and office suite too.
4. It's not particularly stable. It's not that it's that unstable; it doesn't crash every hour on me or anything, but what annoys me is that when it crashes -- which it does -- it never goes down cleanly. Invariably, some little part of it is still running, and the only way you can get things working again is by restarting. And restarting is irritating. There's no reason a userland program should make you restart. Ever. Maybe this is a Windows thing, I don't know. It just sucks.
I could go on, but those are my major complaints. It's just a program that tries to do too much. There are obviously some very smart people working on it -- the whole database backend is pretty slick, and the way you can create local databases and keep them syncronized (replicated, in Notes-ese) to other computers and servers is neat. But everyone I know who uses it, spends 90% of their time with it just doing email, and it's a really crappy email client. Ten years ago, or even five years ago, it might have been competitive. But when you put it up next to Apple Mail or even GMail, using it is just painful.
What I do find somewhat uncomfortable is that you're always the "good guys" in America's Army, the enemy is always the terrorists no matter which side you're on.
You are aware who publishes America's Army, right?
I kinda bet that if Osama Bin Laden was in any position to make a FPS, the enemy would be "The Godless Infidels of the Great Satan" in every mission, too.
I'm 100% sure you can play as Germany in Close Combat 3 (and 90% sure you can do so in 1 & 2).
You definitely can play on the Axis side in original Close Combat. I've never played any of the other ones, but somewhere around I've still got the first one. One of the few decent products from Microsoft; I wonder where they bought it from.
A fun game, but not much of a "war simulator." In fact, none of those god-view RTS games are, much to the disappointment of actual military commanders everywhere, I'm sure. Of course, this is a direct result of the fact that the real experience isn't something that a whole lot of people would want to play (a real WWII command simulator would probably involve staring at a map table and listening to a lot of very confused radio broadcasts, and then spending 90% of your time trying to figure out what the hell is going on, and the other 10% reacting to it; you'd also do this for days on end without sleep).
Well I think it's probably because, up until 9/11, every time a plane had not responded to the tower, it had turned out to be some form of equipment malfunction, pilot error ("oops -- you wanted us on what frequency?"), or an honest emergency.
There are enough stories about small planes wandering off course and into no-fly zones; pre-9/11, just shooting down an airliner full of people because the pilot flew it too close to a military installation would have been really bad PR. I suspect that whatever standing orders allowed such a shootdown would only ever be used once -- after that, the political fallout would be so bad that they'd be immediately rescinded and replaced with a chain of command that included everyone from NORAD to the Junior Congressman from Hawaii's masseuse's dog. In other words, it would turn into a complete goatfuck, and quite possibly make the future situation worse by slowing responsiveness.
Post 9/11, of course it seems obvious that a shoot-first-ask-questions-later policy is the only way to go when every passenger plane is one box-cutter away from becoming a giant guided missile, but that would definitely not have been the public perception pre-9/11, I don't think.
Telling people "we had to shoot the plane down, it might have been terrorism," as an excuse falls pretty flat when there are the charred bodies of passengers shot down by their government all over CNN.
The public doesn't make rational decisions, it makes emotional ones, and such a policy wouldn't have been viable without the giant slap in the face that 9/11 was.
And build a datacenter and run some atm lines to them for nightly backups.
Seems like building a datacenter is sort of putting all their eggs in one basket.
They're banks. Assumedly, they have vaults. I'm going to assume also that they have internet connections--if not, then they should. Rather than centralizing the backups, have each bank back up to a server located at another bank branch every night, then make physical-media snapshots and put them in the vault. Best would be to have each bank's partner be one that's geographically isolated from it, and/or for each bank to back up to two other branches (if you can afford the storage).
That way each branch gets at least one off-site backup, but you also avoid having everyone's off-site backups being in the same place. Even if some sort of disaster wipes out a few branches, it's not cataclysmic. You can lose any branch and not lose any data, or lose any two branches and only have a (1/n) chance of losing data, where n is the number of branches.
I'm just making this up quickly, but I think the odds of losing data, if you have b backup sets of each branch's data, L branches destroyed, and n total branches, is: ( ((1/b) * L) / (b * n) ) Eh, on second thought I don't think that quite works. I was trying to test the scenario of 6 branches each of which backs up to 2 other branches, and then randomly destroying 3 branches, to come up with the odds of data loss. I think it might be 1 in 3, and the formula gives 1 in 8.
Quick question... you mentioned using OpenVPN to do the remote-to-central backups. Why not just use rsync? Seems like it would be easier than opening a VPN connection, mounting or otherwise connecting to the server, and then syncronizing the files to be backed up (which you'd need to use other utilities for anyway). With rsync, it's all done for you and the security is still there, since it's done over SSH. Keeping a remote mirror is as easy as one line in crontab (plus setting up the required certificates), and snapshots aren't much harder.
After SSH itself, rsync is one of the most useful little utilities that I couldn't live without. It just works. About the only thing it doesn't do is true bidirectional syncronization, but this isn't as much a limitation for making backups as it is for situations where people are going to be changing things on both ends.
Anyway, I thought the rest of your post was right on, I just thought the SSL VPN thing was the hard way.
I think what you said about biometrics can't be said enough.
I can't tell you how many times I've heard lately about biometrics and how they're going to be the "next big thing," and how they're "so secure." A few times, I've even heard the dreaded P-word come up. The one you never hear from anyone who knows what they're talking about in regards to system security: "perfect."
People think because they use their thumb-print to access their computer, that somehow it's impossible for anyone without their thumb to get in. It's ridiculous, but that's literally what people think. My theory is that this stems from a general lack of understanding by the public (and PHBs) about how systems get hacked: they believe that it happens because passwords get guessed, and never for any other reason. Therefore, they reason, if you remove the passwords, there's nothing to guess, so it's perfect.
It just never seems to occur to people that to a computer, a thumb is just a bunch of numbers. A biometric ID, at the end of the day, is just a (hopefully very long) password. And there will always be a place somewhere in the authentication system where that raw data can be extracted or injected (before it's hashed). Right at the scanner, if nowhere else. Once you've grabbed the numeric equivalent to their thumbprint, you can use it over and over. Unlike stealing a password, which can be changed, the victim is now quite screwed.
The only biometric systems I've ever seen that I think deal with these issues are the ones that require everyone to carry around their own scanner/biometric reader, which would have enough onboard logic to do a public key exchange with the system (after using your thumbprint to decrypt the stored secret key, perhaps). That way no sensitive information ever touches the system that's being authenticated into: you never touch your thumb (eye, voice, whatever) to anyone's but your own scanner. This type of system, of course, removes the dreamy goal that people have for biometrics: that basically you could walk around and never have to carry or remember anything.
People get so blinded by the "Get Smart"/sci-fi aspect of biometrics that they don't consider that for many applications, they may in fact be a step down from current password-based schemes (barring people keeping their passwords on Post-Its on their terminals). While biometrics are a fine for one part of a multiple-factor scheme, or as a replacement for a unique but nonsecure identifier, (e.g., if your current system is "employee ID" plus password, you could replace "employee ID" with a biometric), I wouldn't want them to be used as the only authentication secret.
Regardless, there is little reason beyond the philosophical for them to be open source except to ensure that your games will play on all platforms going forward provided someone will take the time to make necessary tweaks so that it will compile.
I think that's a pretty important reason. One of my personal reasons for liking open source is that it limits the amount of abandonware that can't ever be re-used (not to say that projects don't get abandoned in the OSS world, but they at least can be restarted by someone else if there's a desire).
On the other hand, I understand that game companies want to make money, and that perhaps not everyone can survive on the open engine plus closed content model.
What I really like as a compromise is something like what Bungie ended up doing with Marathon: after the game wasn't bringing them in any money anymore, they open sourced the engine and basically turned it over to the fan community. Actually Bungie goes above and beyond the call of duty IMO they maintain a domain (and maybe the server too, I'm not sure) for the GPL effort at http://source.bungie.org/ and they released the game content files for download, so you can grab the GPL engine and the content files, and play the retail version of the game on your modern machine.
There are a lot of games I remember playing and enjoying that I wish had gotten this treatment, but instead just fell off the face of the earth.
Somewhere I read an article/posting where someone described a concept they called "foreverware." Basically it's commercial software, games in partiuclar, which is sold with the promise (in writing) that after a certain amount of time, the source code will be released to the public. I think compromises like this might do a lot of good, in the markets where OSS has so far achieved limited penetration. I could even imagine situations where escrow organizations hold the source code and other documentation, so that even if the company were to go bankrupt, their promise to release would still be valid.
With more and more software being built for a defined life-cycle (e.g. "we're going to support this for 5 years, after that it's done,") I think that there could be a significant demand for software that would open itself up after a certain amount of time, while also giving the maker time to realize profits and stay ahead of the competition in releasing the next version, by not giving up too much of an edge.
Interesting; I'll have to see if I can find the episode around somewhere. Sounds like maybe the deflection angle is important: maybe that causes the modern bullets to spin and shatter, while a round ball is basically immune to damage by spinning (since its cross section is the same from all angles).
I wonder what a musket firing a Minie ball would do.
If you inflation adjust TVs, vcrs, dvd players, stereos, computers, or ANY other consumer electronic device you will see a violently declining price.
That's kind of the point here. If you look at the graph, the price isn't declining all that violently. They declined pretty sharply in the late 70s, 80s, and early 90s, then climbed in the mid 90s, collapsed with the debut of the N64 and Dreamcast, and are now going back up again in inflation-adjusted (real) dollars.
From a perspective of someone interested in a PS3 is that Sony has priced its product into a range previously occupied over the past 20 years by the Neo Geo, CDi, and 3DO -- none of which were terribly successful commercially, at least when compared to less-expensive platforms like the NES, SNES, and Playstion.
So while experience in other technology sectors indicates that consoles ought to be dirt cheap right now, Sony is still trying to charge a 1982 price for a 2006 product. It remains to be seen whether consumers will be wooed by the technology into shelling out that much dough. Frankly, I'm skeptical. But then again, I've never bought a new console in my life: I wait until they're one generation out from new and I get can get one used, complete with a mod chip and other goodies that early adopters have to live without.
Slashdot Mirror
People seem to be spending a lot of time concentrating on this guy's motivation.
I really don't understand why. Why should I care what his motivation is? The point is that he got the USPTO to re-examine a dumb patent. If it did it for the publicity or to hawk his blog or because he thinks it'll speed up the Second Coming of Jesus, I don't really care. Hooray for him, in any case; the point is that the action he took was good, irrespective of the motive.
I'm not going to fault him because he was doing it for publicity. Maybe if a few more people did what he's doing -- for publicity or whatever reason -- we'd have a few less shitty patents floating around.
In fact, if you want to hurt a company these days, you're better spent going after their patents than trying to take them to court directly, especially small-claims. So if there's a company you don't like, and you can get the dough together to force a re-examination of their patents, by all means please do so. I won't question why you're doing it, the point is that it gets done.
But is it open source?
Sure -- it's published under the NSA Open Source License. In short, "We tell you, but then we'd have to kill you."
As long as we're tossing around the Flamebait here, I'd like to disagree with this statement:
The worst ones are the fuckers that voted for him in the 2nd round, and now are all disappointed and disaproving.
No, the worst ones are all the people who voted for that mechanoid John Kerry in the Democratic primaries, over other people who might have actually -- oh, I don't know, had some sort of impact on the American public? Instead they picked a guy whose only strength was that he "wasn't named Bush," and made G.W. look well-spoken and animated by comparison.
The Democratic party snatched defeat from the jaws of victory in that race, by not having the balls to pick anyone other than Kerry.
Well, tape drives that have been around for a few years are cheap. It's new tape drives that are expensive.
For example, during one of my backup obsessions, I got a HP SureStore Tape 12000e for about $25 on eBay. It's a DDS-2 autoloader. Holds six 4GB (native, allegedly 8GB compressed) tapes in a magazine and changes them on command.
A few years ago, such a device would have set me back a few thousand bucks. Now, because people have moved to DDS-4 and higher, they're dirt cheap. There's nothing wrong with them, it's just an issue of capacity. (Though I think most people could probably fit their most important stuff in less than 8GB, counting documents only, no MP3s, warez, porn, applications, or system files, and backing up things like digital photos separately.)
Compared to hard drives of similar capacities, tape drives have always been expensive from a home user's perspective, at least in the recent past. I suspect they'll stay this way in the future: the economies of scale are simply not there to make tape drives as inexpensive to manufacture as HDs. However, it's not hard to find ones that are a generation or two old for rock-bottom prices and still have a lot of life left in them.
Except that they really don't. The cost differential between a comparably equipped Mac and PC is pretty small, and depending on how you spec things out (and the prices of commodity equipment at a particular time) sometimes the low-end Macs even come out ahead. I think when the Minis were first brought out, they were cheaper than a comparably-spec'd major-manufacturer PC. At any rate, when you take that increase in cost and amortize it out across the multi-year timespan that you're probably going to own a computer, it's fairly small. At least that's how I see it: every time I buy a computer, particularly my main desktop, I try to think about all the time I'm going to be spending in front of it over the next few years. It makes sense to buy what I'm going to be really happy with, even if it costs more initially.
I've used Macs for years, and there was a time when the cost difference was a lot more pronounced (right after the prices on commodity hardware started to collapse), but that's not the case so much anymore. While there will always be people who want to compare a $500 Mac to a $150 Wal-Mart PC (or back a few years ago, a $1k Mac to a $500 Office Max special), but that's not a valid comparison or argument.
Furthermore, if you're planning on making a living off of some type of computer skill, it doesn't make sense to cut corners on your tools.
Him:
I sure wouldn't want to have to tell a client that their work is delayed because the WLAN at my Howard Johnson shit the bed.
You:
I stay in a lot of business-class hotels (4* and above).
Note subtle difference. (HoJo's != 4-star Business Hotel.) That might be part of your difference in experience.
He was making fun on Sun, not Ubuntu, I think.
It's Sun that would be putting their "weight" on Ubuntu, at least if you want to extrapolate from what TFA says, which really isn't all that convincing.
While we're on the subject, does anyone know what "Public key encryption support (RSA)" means, as a feature of a CPU?
Some sort of special on-chip hardware, or is it just a 'marketing feature' with no real relation to anything on the die?
A quick Google found only one reference to it, here, which suggests to me that might be for real, and has something to do with using it for SSL/HTTPS workloads. Anyone (someone who has perhaps read the Architecture Specifications, which I admit I have not) want to clue me in on the straight dope?
I use Ubuntu (actually Kubuntu) on my Linux desktop machine, but I use straight Debian on my headless server.
Can anyone tell me why a person would want to use Ubuntu on a server, as opposed to just using Debian?
It seems to me that most of the advantages of Ubuntu are on the GUI side of things, and this is the way that most of the software that's different for Ubuntu than Debian is aimed towards. Most of the server-type packages you'd probably be pulling from the Debian repositories anyway, so there's not much advantage and some things might not work, because Ubuntu doesn't follow the "Debian way" in everything (there are some file locations and paths that are different, I believe). Plus Debian has always seemed a bit better documented, although I admit that's arguable.
I'm glad to see Sun put its weight behind a Debian-based distro, but I don't quite get why Ubuntu and not just Debian, especially if it's for servers. The only reason I can think is that they don't want to get too close to Debian's leadership and philosophy, and find Ubuntu more palatable from a PR and customer-relations perspective. Still, it seems like an odd choice.
Idiotic moderation ... your comment was the best explanation I've gotten in this thread so far as to what's going on and what is and isn't closed.
I can't really think of why anyone would want to run Darwin x86 without OS X either; we've already established that it's a worse server platform than Linux for most tasks, especially database ones, and headless servers are really the only place I think there'd be a market for Darwin. And it's not like there's any dearth of server OSes and distos these days anyway. The only other people are those who want to create a platform on which to do unauthorized ports of OS X onto commodity hardware (say by hacking the kernel to remove the hardware verification portions, and creating a foundation on which to run the proprietary portions of the Mac OS).
I figured this was inevitable all along. In fact, back when people were cooing over how folks had gotten OS X to boot on commodity hardware, I speculated that it was going to drive Apple to close up more and more of its OS, and I think if it continues, we're going to see a lot of phone-home type registration systems. To be perfectly honest, as someone who's always appreciated the fact that Mac OS has never had copy protection (because it depends on having a rather largish dongle, called a Mac), I would rather see Apple do what it needs to do to head off commodity ports with licensing than have them start to include obnoxious copy-protection in the OS itself that bothers legitimate users, a la Microsoft.
Normally I'm all for technological solutions rather than legal ones, but in this case the technological ones are going to be much more of a pain in my ass, so I'd appreciate it if they didn't. The day I have to type a serial number into Mac OS X so that it can phone home to Apple, I'm going to be pretty annoyed.
Why bother to be diplomatic? I mean, there have to be some empty seats on those CIA black-bag extradition flights; when you're taking terrorists to Russia to have their toenails ripped out, pay the right people and I'm sure you could get a few spammers to fill the seats on the return leg.
Criminals turn each other in for cash all the time. I'm sure if you threw around some hard currency you could get them to rat each other out, it probably wouldn't even cost that much compared to the actual cost of spam management in the U.S. All you'd need to do is round them up and take them to some place where sending viagra advertisements is a capital offense.
Or save the airfare and just shoot them and throw them into the North Sea, it doesn't matter a whole lot as long as other people in their organization know (or have a vague idea) what happened to them.
I'm joking here, but only partially. I think the sum of all the inconvenience that these people (spammers) cause greatly outweighs whatever their lives are possibly worth to humanity, and I don't see any reason why we should continue to allow them to suck down air.
If there was an anonymous, untraceable way to send money to someone who would use it to kill spammers, I'd probably contribute.
Seriously, it's that annoying.
Maybe Sealand wants to start a Special Forces unit or something.
I tend to wonder if the startup time of Office and its RAM requirements aren't being hidden because it's tied into the OS so deeply that it's really sharing resources with it.
Kind of like how IE seems to start up faster than Firefox, but only because a lot of the stuff that IE uses has already been preloaded by Windows, and therefore it has an unfair advantage (because nobody else but Microsoft can take advantage of this sort of thing).
The way Microsoft intertwines its products basically makes comparisons to anyone else's stuff very difficult, because it's hard to say whether you have half of the MS program already loaded just by turning the computer on.
What is NOT - it is NOT e-mail client.
And that is Notes' biggest problem.
Email is the killer app of the business world. I don't know any organization these days that doesn't live and breathe email.
Notes purports to do, among other things, email. And it does this poorly. So basically, in implementing Notes, you take away users' other email programs, and replace it with something that does a lot of stuff, but doesn't do the ONE THING that they want it to do, very well. Can you blame them for hating it? I sure can't.
The fact that you can develop great database-driven, collaborative applications on it doesn't matter one whit to 99% of its userbase, if the email features suck.
Any program which replaces users' existing email programs, must do email at least as well as what they replace, or it will be hated, regardless of what other features it offers. I don't care if the program makes coffee and massages my feet, I'd still hate it because it's my only option for email, and as an email client it's lame.
Oh Notes, how I hate thee, let me count the ways...
1. 90% of people use it for an email client, which is kind of like using Oracle's RDBMS to manage your grocery list. Or maybe a better analogy is, it's like using a Swiss Army knife to carve a marble sculpture. There are many things that swiss army knife is good for, but when all you want to do is carve marble, it's not the best tool for the job. Eventually it just leaves you wanting to find the person who gave you said pocketknife, and stab them with the little corkscrew, in the hopes that next time they'll give you a proper chisel.
2. Searching sucks. I mean, just sucks. I don't even bother using it, it's that bad. If you can search for stuff in the contents/body of emails, I've never figured out how, and even searching by subject/sender etc. is a PITA. You have to be in the main list view of emails for the folder you want to search, and then you have to click on the column of the term you want to search (e.g., subject). Then you do Find or start typing a few letters, which opens a dialog box into which you type your term. Then you press return. If you're lucky, an email will be highlighted in the view. If you're not lucky, nothing will appear to happen, but if you look down at the veeeeeery bottom of the window, they'll be a line that says "'foo' was not found." As far as I know, you can't use any sort of wildcards, expansions, or regexps in search, and it's totally literal: if you type 'Bob Jones', and the guy's name on the email is 'Bob T. Jones,' you won't find it. This is absolutely braindead for an email application, especially one that's as database-centric as Notes purports to be. I don't know about Exchange, but Apple Mail is better than this (stores everything as flat text files, too), as is GMail and every other webmail system I've ever used.
3. Everything runs in one gigantic window. If you can tear off message windows so that they're not all running as tabs, I've never figured out how. If you don't like it taking up your entire screen, your only option is to restore down, then resize the window and scroll. People really aren't that far off when they're joking that it seems like Lotus is about two lines of code away from becoming a full-fledged OS: it seems designed under the assumption that it's the only application you'll use. Ever. For everything. It's email, it's groupware, it's calendaring, it's a CMS/version-control system, hell, it's even a web browser. Why? No idea! And now it's going to be a word processor and office suite too.
4. It's not particularly stable. It's not that it's that unstable; it doesn't crash every hour on me or anything, but what annoys me is that when it crashes -- which it does -- it never goes down cleanly. Invariably, some little part of it is still running, and the only way you can get things working again is by restarting. And restarting is irritating. There's no reason a userland program should make you restart. Ever. Maybe this is a Windows thing, I don't know. It just sucks.
I could go on, but those are my major complaints. It's just a program that tries to do too much. There are obviously some very smart people working on it -- the whole database backend is pretty slick, and the way you can create local databases and keep them syncronized (replicated, in Notes-ese) to other computers and servers is neat. But everyone I know who uses it, spends 90% of their time with it just doing email, and it's a really crappy email client. Ten years ago, or even five years ago, it might have been competitive. But when you put it up next to Apple Mail or even GMail, using it is just painful.
That was one of the best posts I've read in a long time. I normally refrain from contentless "mod parent up" posts, but I'll break my rule here.
Sadly, I don't get mod points anymore (apparently somebody didn't like my moderations too much), but if I did, you'd have gotten some.
What I do find somewhat uncomfortable is that you're always the "good guys" in America's Army, the enemy is always the terrorists no matter which side you're on.
You are aware who publishes America's Army, right?
I kinda bet that if Osama Bin Laden was in any position to make a FPS, the enemy would be "The Godless Infidels of the Great Satan" in every mission, too.
I'm 100% sure you can play as Germany in Close Combat 3 (and 90% sure you can do so in 1 & 2).
You definitely can play on the Axis side in original Close Combat. I've never played any of the other ones, but somewhere around I've still got the first one. One of the few decent products from Microsoft; I wonder where they bought it from.
A fun game, but not much of a "war simulator." In fact, none of those god-view RTS games are, much to the disappointment of actual military commanders everywhere, I'm sure. Of course, this is a direct result of the fact that the real experience isn't something that a whole lot of people would want to play (a real WWII command simulator would probably involve staring at a map table and listening to a lot of very confused radio broadcasts, and then spending 90% of your time trying to figure out what the hell is going on, and the other 10% reacting to it; you'd also do this for days on end without sleep).
Well I think it's probably because, up until 9/11, every time a plane had not responded to the tower, it had turned out to be some form of equipment malfunction, pilot error ("oops -- you wanted us on what frequency?"), or an honest emergency.
There are enough stories about small planes wandering off course and into no-fly zones; pre-9/11, just shooting down an airliner full of people because the pilot flew it too close to a military installation would have been really bad PR. I suspect that whatever standing orders allowed such a shootdown would only ever be used once -- after that, the political fallout would be so bad that they'd be immediately rescinded and replaced with a chain of command that included everyone from NORAD to the Junior Congressman from Hawaii's masseuse's dog. In other words, it would turn into a complete goatfuck, and quite possibly make the future situation worse by slowing responsiveness.
Post 9/11, of course it seems obvious that a shoot-first-ask-questions-later policy is the only way to go when every passenger plane is one box-cutter away from becoming a giant guided missile, but that would definitely not have been the public perception pre-9/11, I don't think.
Telling people "we had to shoot the plane down, it might have been terrorism," as an excuse falls pretty flat when there are the charred bodies of passengers shot down by their government all over CNN.
The public doesn't make rational decisions, it makes emotional ones, and such a policy wouldn't have been viable without the giant slap in the face that 9/11 was.
And build a datacenter and run some atm lines to them for nightly backups.
Seems like building a datacenter is sort of putting all their eggs in one basket.
They're banks. Assumedly, they have vaults. I'm going to assume also that they have internet connections--if not, then they should. Rather than centralizing the backups, have each bank back up to a server located at another bank branch every night, then make physical-media snapshots and put them in the vault. Best would be to have each bank's partner be one that's geographically isolated from it, and/or for each bank to back up to two other branches (if you can afford the storage).
That way each branch gets at least one off-site backup, but you also avoid having everyone's off-site backups being in the same place. Even if some sort of disaster wipes out a few branches, it's not cataclysmic. You can lose any branch and not lose any data, or lose any two branches and only have a (1/n) chance of losing data, where n is the number of branches.
I'm just making this up quickly, but I think the odds of losing data, if you have b backup sets of each branch's data, L branches destroyed, and n total branches, is:
( ((1/b) * L) / (b * n) )
Eh, on second thought I don't think that quite works. I was trying to test the scenario of 6 branches each of which backs up to 2 other branches, and then randomly destroying 3 branches, to come up with the odds of data loss. I think it might be 1 in 3, and the formula gives 1 in 8.
If anyone wants to correct me I'd be interested.
Quick question ... you mentioned using OpenVPN to do the remote-to-central backups. Why not just use rsync? Seems like it would be easier than opening a VPN connection, mounting or otherwise connecting to the server, and then syncronizing the files to be backed up (which you'd need to use other utilities for anyway). With rsync, it's all done for you and the security is still there, since it's done over SSH. Keeping a remote mirror is as easy as one line in crontab (plus setting up the required certificates), and snapshots aren't much harder.
After SSH itself, rsync is one of the most useful little utilities that I couldn't live without. It just works. About the only thing it doesn't do is true bidirectional syncronization, but this isn't as much a limitation for making backups as it is for situations where people are going to be changing things on both ends.
Anyway, I thought the rest of your post was right on, I just thought the SSL VPN thing was the hard way.
I think what you said about biometrics can't be said enough.
I can't tell you how many times I've heard lately about biometrics and how they're going to be the "next big thing," and how they're "so secure." A few times, I've even heard the dreaded P-word come up. The one you never hear from anyone who knows what they're talking about in regards to system security: "perfect."
People think because they use their thumb-print to access their computer, that somehow it's impossible for anyone without their thumb to get in. It's ridiculous, but that's literally what people think. My theory is that this stems from a general lack of understanding by the public (and PHBs) about how systems get hacked: they believe that it happens because passwords get guessed, and never for any other reason. Therefore, they reason, if you remove the passwords, there's nothing to guess, so it's perfect.
It just never seems to occur to people that to a computer, a thumb is just a bunch of numbers. A biometric ID, at the end of the day, is just a (hopefully very long) password. And there will always be a place somewhere in the authentication system where that raw data can be extracted or injected (before it's hashed). Right at the scanner, if nowhere else. Once you've grabbed the numeric equivalent to their thumbprint, you can use it over and over. Unlike stealing a password, which can be changed, the victim is now quite screwed.
The only biometric systems I've ever seen that I think deal with these issues are the ones that require everyone to carry around their own scanner/biometric reader, which would have enough onboard logic to do a public key exchange with the system (after using your thumbprint to decrypt the stored secret key, perhaps). That way no sensitive information ever touches the system that's being authenticated into: you never touch your thumb (eye, voice, whatever) to anyone's but your own scanner. This type of system, of course, removes the dreamy goal that people have for biometrics: that basically you could walk around and never have to carry or remember anything.
People get so blinded by the "Get Smart"/sci-fi aspect of biometrics that they don't consider that for many applications, they may in fact be a step down from current password-based schemes (barring people keeping their passwords on Post-Its on their terminals). While biometrics are a fine for one part of a multiple-factor scheme, or as a replacement for a unique but nonsecure identifier, (e.g., if your current system is "employee ID" plus password, you could replace "employee ID" with a biometric), I wouldn't want them to be used as the only authentication secret.
Regardless, there is little reason beyond the philosophical for them to be open source except to ensure that your games will play on all platforms going forward provided someone will take the time to make necessary tweaks so that it will compile.
I think that's a pretty important reason. One of my personal reasons for liking open source is that it limits the amount of abandonware that can't ever be re-used (not to say that projects don't get abandoned in the OSS world, but they at least can be restarted by someone else if there's a desire).
On the other hand, I understand that game companies want to make money, and that perhaps not everyone can survive on the open engine plus closed content model.
What I really like as a compromise is something like what Bungie ended up doing with Marathon: after the game wasn't bringing them in any money anymore, they open sourced the engine and basically turned it over to the fan community. Actually Bungie goes above and beyond the call of duty IMO they maintain a domain (and maybe the server too, I'm not sure) for the GPL effort at http://source.bungie.org/ and they released the game content files for download, so you can grab the GPL engine and the content files, and play the retail version of the game on your modern machine.
There are a lot of games I remember playing and enjoying that I wish had gotten this treatment, but instead just fell off the face of the earth.
Somewhere I read an article/posting where someone described a concept they called "foreverware." Basically it's commercial software, games in partiuclar, which is sold with the promise (in writing) that after a certain amount of time, the source code will be released to the public. I think compromises like this might do a lot of good, in the markets where OSS has so far achieved limited penetration. I could even imagine situations where escrow organizations hold the source code and other documentation, so that even if the company were to go bankrupt, their promise to release would still be valid.
With more and more software being built for a defined life-cycle (e.g. "we're going to support this for 5 years, after that it's done,") I think that there could be a significant demand for software that would open itself up after a certain amount of time, while also giving the maker time to realize profits and stay ahead of the competition in releasing the next version, by not giving up too much of an edge.
Interesting; I'll have to see if I can find the episode around somewhere. Sounds like maybe the deflection angle is important: maybe that causes the modern bullets to spin and shatter, while a round ball is basically immune to damage by spinning (since its cross section is the same from all angles).
I wonder what a musket firing a Minie ball would do.
If you inflation adjust TVs, vcrs, dvd players, stereos, computers, or ANY other consumer electronic device you will see a violently declining price.
That's kind of the point here. If you look at the graph, the price isn't declining all that violently. They declined pretty sharply in the late 70s, 80s, and early 90s, then climbed in the mid 90s, collapsed with the debut of the N64 and Dreamcast, and are now going back up again in inflation-adjusted (real) dollars.
From a perspective of someone interested in a PS3 is that Sony has priced its product into a range previously occupied over the past 20 years by the Neo Geo, CDi, and 3DO -- none of which were terribly successful commercially, at least when compared to less-expensive platforms like the NES, SNES, and Playstion.
So while experience in other technology sectors indicates that consoles ought to be dirt cheap right now, Sony is still trying to charge a 1982 price for a 2006 product. It remains to be seen whether consumers will be wooed by the technology into shelling out that much dough. Frankly, I'm skeptical. But then again, I've never bought a new console in my life: I wait until they're one generation out from new and I get can get one used, complete with a mod chip and other goodies that early adopters have to live without.