I think maybe he was confusing MUA and MTA, but I'm not sure.
The encryption really needs to happen at the MUA level, which theoretically could either be a desktop (traditional) MUA or a very smart web app, but I'm not aware of any web apps that will do encryption in the browser. (Not SSL, but actually load up a copy of GPG and encrypt the message.) I talked a little more about this in another post further down in this thread, but what I think the world needs is some sort of lightweight GPG implementation that could run as a Java or JS applet. That way you'd be able to use GPG even through a webmail interface. (You'd still need some sort of private-key management system, not sure how you'd do that. USB stick, maybe?)
But without that, using encrypted email basically means using a traditional MUA, or trusting that your webmail provider won't sell you out to the authorities (e.g. like Hushmail), if the encryption is being done on their server or via untrusted software.
wow! Run one app at a time that boots instantly! Reminds me of my Apple 2, the only technology that 1-12th graders still need. Agreed. I've often felt that, in terms of sheer usability, computers have mostly been downhill from the Apple II.
You put the disk in, you turn it on. When you're done, you save and turn it off. If you want to use a different program, you wait for the light on the disk drive to be out, then you swap disks and give it the three-finger salute.
I've seen kindergarteners and first-graders working Apple IIs without any problems, supervision, or assistance at all. All you have to do is reinforce to them that they're not to take the disk out of the drive when the light is on, and it's pretty tough to mess anything up.
In the past I would have recommended Hushmail (http://www.hushmail.com/), but apparently they've cowed to U.S. law enforcement (due to Canadian law) lately, too. Although I wonder whether by cooperating with the Feds they turned over just the encrypted messages, or the actual plaintext? I can't seem to find any details on exactly what level of cooperation they offer.
Really, what you need to find is a web-based service where the encryption/decryption is performed in the user's browser via a JS or Java applet, and thus no unencrypted information is ever transferred to the server. With nothing to turn over, they could happily cooperate. However, I'm not sure this is how Hushmail works -- it sounds like they may do the actual GPG encryption on their server, which of course ruins the entire security model.
The GPG binary is 4MB, so it's a bit large to be using in an applet directly, but I wonder if you couldn't reimplement it in Java or pare it down somehow so that you could embed it?
One of the problems with email is that, to get real security, you want to push the encryption as far out to the edges as possible. But for convenience, people are actually moving towards heavily centralized systems (Gmail, etc.) so that they can access their mail from anywhere. I think the only way you can have both is if you put the encryption engine in the browser somehow.
Don't keep it in your local mailstore (your 'Inbox') unencrypted? When people talk about 'end to end' email security, they don't mean just securing the over-the-wire transmission (e.g. by using SSL); they mean user-to-user encryption.
Most decent MUAs that support encryption leave the messages encrypted, and only decrypt them for the purposes of viewing the message. When you move away from the message or close the window, the ephemeral decrypted copy is destroyed and the only thing on the disk is the encrypted version. (If you're really paranoid, you should probably make sure you have encrypted virtual memory turned on, and that you're using a Tempest-resistant font/display, etc. But even if you don't, you've significantly raised the bar for someone to snoop on you.)
Of course, if you're using an OS and/or software that you don't trust, then you have a problem, since it's entirely feasible that you could get rooted, or maybe the OS or MUA would just be backdoored. More reason for using software that, if it's not practical to actually review it yourself, is at least reviewed by people whose reputations you trust.
And then you have to think about hardware security -- would you know if someone came and swapped your keyboard for an identical model while you were away from your desk? If the answer to that question isn't a sure 'yes,' then you're screwed.
So you have no data you wish to keep private and off of other people's computers? There are lots of ways to achieve this besides actually storing the bits locally. (And honestly, even if you're keeping the data stored locally, if you're using even one network-enabled application that you haven't personally combed over the code for, you have no idea where your data really is, anyway.)
You could use a smart terminal that contains decryption hardware and keep your keys stored on a smartcard; that way all the data stored on the server, and traveling down the wire, would be encrypted, and you'd just have to worry about the small amount of hardware and microcode at your end of things. That's probably much better security than keeping your data stored locally on a platform that's so complicated that you can't ever trust it completely. (You'd probably have to do some of the application-level processing on the client side too, to avoid sending unencrypted data back to the server, e.g. if you wanted to search a document.)
Also, I suspect the majority of PCs in the world are used in workplaces where their owners -- not necessarily their users -- would prefer that all the data be stored in a centralized location. It makes it easier to backup, easier to share, and more difficult for employees to steal or hold hostage.
Taking care of your sick and poor makes sense from both humanitarian and monetary viewpoint.
If the monetary side of things was really there and could be supported by hard evidence, you'd probably see a lot less opposition to the idea in the U.S.
However, that's not the way the debate is framed. It's all about mushy philosophical issues, and there's nothing that makes businesspeople scream "bullshit" and run the other way faster than hearing about what they ought to be doing out of charity or 'compassion.'
Start talking numbers and you might have a case. The people who are going to be swayed by moral arguments already have been; the rest of the country doesn't give a damn.
The "dying" market simply represents the maturity of the technology, and that people are finally getting off the upgrade treadmill. For the past decade or two, although most notably in the past few years, it's been necessary to continually upgrade your machine if you wanted to stay on top of current capabilities. Except for people who play games, I think that's beginning to become no longer true. Or perhaps the new capabilities offered by possible hardware upgrades are no longer as compelling as they once were.
The market has gotten saturated (I suspect moreso in Japan than here in the U.S., where you do occasionally find people who still don't own a computer), and people are not upgrading as quickly. That doesn't mean the PC is "dying," it just means that sales are going to stabilize at the replacement rate.
Personally, I think this is a really good thing for the PC and for software. For years software has been alternately leading and chasing hardware, and there haven't been too many periods where the architecture and capabilities were stable enough to allow for many resources to be devoted to optimization and good design. If the upgrade pace slows, maybe we'll finally get a focus on quality and design, since lazy programmers won't be able to count on hardware increases to make their same bloated code get faster every year.
I expect there'll be a shakeout, eventually, both in the hardware and software side. (The hardware side mostly shook the dead weight out during the dot-bomb, but I think there's a lot of people who are going to get squashed when sales in the established markets plateau and they can't move to other markets or get into maintenance-mode fast enough.) But computers, in general, are here to stay, and the desktop/laptop PC performs too many functions for it to easily be replaced by lots of discrete devices. It's just no longer the frontier.
I really don't understand what all the hate is about for Computer Modern. I think it's a fine font.
Of course, it doesn't look nice on the screen when viewed at 100%. But that's what you get for viewing something at 72 (or 96 or whatever Windows uses) dpi, that's designed to be viewed on paper at 300dpi.
If you blow Computer Modern up to 150% or so, which in my experience tends to be what happens if you fit the width of a document to a good-sized monitor, I think it looks pretty good. But at 10 or 12pt at 100% magnification on a low-resolution device like a computer monitor, you lose all the fine detail that you need.
I guess I can see how there's a need for an alternative for people who are doing all-digital workflows, but if you're going to print the paper out at the end, there's nothing wrong with Computer Modern. It certainly beats the pants off of Times New Roman, IMHO, if you have a good 300+ dpi laser to use as an output device.
buying from a union shop makes it just about impossible for you to be supporting illegal immigrants working in this country for illegally low wages without paying taxes
If buying from a union shop was a guarantee that you weren't supporting illegal labor, I'd be a lot more inclined to buy union. But it isn't, and in fact quite a few of the unions are involved in legalization schemes because they think all the illegal workers will increase their political clout, if they can get them unionized.
In effect, the union organizations are selling out their rank-and-file (who are the ones who really get hit by the wage depression as a result of all the illegal workers) in order to bring in lots of new members and make themselves more powerful. I've seen this to most obvious effect in the unions that have gotten involved in the service industries, but it's pretty widespread if you look for it. Very few of the unions seemed to be putting up much of a fight when push came to shove earlier this year on the immigration issue.
At least the corporations generally are upfront about screwing you; the union leaders seem to enjoy pretending that they're on the side of legitimate workers while doing the same thing.
Off the top of my head, open source typically carries an ideology with it. It would be like a public school requiring students to do missionary for the Mormons - you're forcing people who may not agree to further your beliefs. Ideology and religion are two separate things.
Public education rams a lot of ideology down its students' throats. That's not an accident -- it's by design. It takes whatever happens to be the majority viewpoint in a particular area, usually as determined by the local school board in conjunction with whatever's written in the textbooks, and impresses that on the students.
Now, generally, that's desirable. There are some ideologies that I think we can probably agree we want rammed down students throats; or, if not that, we can at least agree are necessary to do so, in order to preserve our society. (If you're cynical, you could say that it's necessary in order to produce the next generation of suckers who will slave under the system in order to keep it going.) Things like democracy, tolerance, the rule of law, etc.
There's nothing that inherently prohibits public education from being ideologically loaded. It's prohibited from teaching the specific ideologies that comprise religion, however, by law. But almost anything else is fair game. Outside of religion, it's all a popularity contest.
Although I think some people take the Free Software thing a bit far, I'm not sure it's achieved the status of a true religion just yet. (Though when Stallman dies, all bets are off.) If you could show that there was some social utility in ramming Free Software down students throats, in the same way that the schools treat gender equality and democracy -- both equally artificial concepts -- then there's no real issue.
As much as I like Linux, the modern distros are terribly sluggish on this particular machine. I think that's really because of the mentality of some of the distro packers rather than anything architectural about Linux itself. The big-name Linux distros are mostly playing the same eye candy war as Vista and Leopard; they're developed by people who have relatively new, high-powered machines, for people with the same equipment.
However, there are distros that offer less crap and have the same core, like Xubuntu, or even by just installing Debian and then pulling down your favorite desktop environment (I like Ratpoison, which is designed to work without a mouse). And unlike W2k, which really doesn't offer you much in the way of an upgrade path anymore, at least a Linux install lets you run a modern kernel, just without the fluff.
(As a datapoint, I'm running a modern Debian Stable kernel on a Pentium 133 with 48MB of RAM. I don't normally use any GUI on it at all, but it's fine for ncurses-based apps and for working as an SSH/VPN endpoint. It'd be tough to do that with most commercial OSes, because it's so old it wouldn't run current versions, and thus wouldn't be secure.)
You had a 20GB drive in 1990? I think you mean 20MB, unless you stole it from the time machine research lab. Or you had your entire basement and several million dollars dedicated to it.:)
I remember when Iomega came out with the 100MB Zip in 1994, and I thought that was cool. I was still using a 80MB drive up until then, although it had started to get really unpleasant. (I had an organizational system for 1.44MB HD floppies that still seems like a ridiculous level of anal-retentiveness today, but it was the only way I could find anything.)
To be honest I find the increase in storage capacity and decrease in cost even more interesting and important in its impact than the increase in processor speeds over the same period.
Yeah, but don't try and run Open Office on that machine. Gonna suck. Not that MS Office 2007 would run great either, but for a modern machine, it's still pretty pathetic. It's about at the year 2001 level. Yeah, and as we all know, prior to 2001, the office suite market was an absolute wasteland. It's amazing we got any work done at all.
In 1990 or 91 it would have been a supercomputer, too.
1991 was pre-Pentium; hot stuff then would have been a 68040 if you were a Mac user, and some sort of 486 if you used PCs.
I don't get what people's problem is with the performance of this thing. Granted, I don't upgrade very quickly (I have a dual-1GHz machine and an ancient headless Pentium-133 box under my desk, and the dual-GHz box only replaced a 400MHz one last year), but that's more than enough power for everything except gaming.
Frankly I'd like to see consumer hardware plateau so we can get off of the upgrade treadmill, and the software people can start groveling around retirement communities to find someone who knows what 'optimization' means. It's absolutely ridiculous that people think you can't do typical productivity tasks on a 1.5GHz machine.
I'm not sure I understand your point. The drug market doesn't really respond very well to threats of punishment, because many of the substances involved are physically addictive. Thus there's always a demand, regardless of how hard the government cracks down on it. The drug dealers are probably motivated by threats of punishment (in that if the threat is higher, they'll demand more compensation to take the risks, thus driving the cost of drugs higher), but the consumers definitely aren't.
Spam-sending seems like it would be far more sensitive to cost. If you executed everyone who sent spam, and you had an apparatus that was fairly good at catching people who sent it (you got a non-trivial percentage, anyway), the remaining people sending spam would have to be pretty desperate for the money. That would mean they'd charge more, and that would mean that the firms paying for the spam in the first place would have a disincentive to purchase their services.
An online pharmacy isn't addicted to spam in the same way that a junkie is addicted to smack; if there's a method of advertising that's more effective for the price than spam, the online pharmacy is going to do it. (Or, put differently, if spamming is no longer cost-effective, they're not going to pay for it any longer.)
Spam has a very low response rate anyway; if you pushed the cost to send it above what generally comes back from a mailing, it would stop. There's nothing really magical or biologically addictive about it -- it's just an obnoxious economic practice.
A couple of times I've gone to wikipedia to look something up - perhaps trivia in someones view, but of interest to me at the moment and found the page "marked for possible deletion", or I've found a page I remember reading had been removed. One locally relevant example was on "slashdot trolls". Was this earthshakingly important information? No. But it did collect in one place a bunch of data that might well have been useful to someone looking at the sociology of trolling (and similar phenonema) on the internet. The deletion of that page annoyed me, too. So much so that when I did find it again, I mirrored it under my WP user page. You can find it here. Of course, it's GFDL, so you can download or further mirror it to your heart's content.
I guess that's the other thing that really gets me about WP deletions. Wikipedia is probably the biggest repository of free, GFDLed content around -- it's stuff that could be useful to a lot of other projects, besides Wikipedia. Every time they permanently delete (and they really seem to expunge with gusto -- not even the Internet Archive seems to retain copies of deleted Wikipedia pages) an article for not meeting the notability guidelines, that's a little bit of free content that's being destroyed forever, to say nothing of the labor and effort that went into producing it.
Just as an example, there's an article on the queue for deletion today, which consists of a massive table of character names and their katakana translations. I think it has something to do with a RPG or trading-card name, I'm not sure -- and to be honest I don't really care. I archived it not out of any particular interest in the subject matter but just because it was a ton of information to just destroy, when it's obvious that a lot of work went into its compilation, and it might be of interest to somebody. To me, deleting something like that without first making sure it's archived somewhere, ranks with burning a rare book because it's in a language you can't read: just because it's not interesting or useful to one person doesn't mean it's without value and should be withheld from the world as a whole. There's something offensive about that on a fundamental level.
That's not a flaw in Wikipedia, it's a flaw in people's search heuristics because they're unused to dealing with databases that contain lots of stuff.
You can either fix it through user education, or by making the search program smarter (e.g., more like Google, which when you type in "Alice", tells you that there are 52 million results but gives you the game "Alice" as #1 and "Alice in Wonderland" as #5, because it takes popularity and other factors besides a literal string-match into account).
Removing information from the database in order to make a stupid search program better at going through it is counterproductive.
could be a problem if we want to improve the credibility of Wikipedia by having a reasonable minimum quality standard. I acknowledge this as a possibility, but I don't think it's as serious a problem as you're probably thinking. Very esoteric articles, by their nature, probably wouldn't be accessed very often, so even if they weren't up to the same standard as the frequently-accessed articles, it wouldn't have nearly as large an impact.
Unless your metric for quality was to click the "Random Article" button and look at the results (which is a stupid way of doing it), infrequently-accessed articles would have minimal effect on overall perceived quality. Particularly if they were marked as questionable in some way. (I've always been a fan of marking pages that have either only been in existence a short time or have only been worked on by a small number of authors/editors.)
I would rather have greater depth and breadth of information, with high quality at the middle and perhaps somewhat lower (but constantly improving!) quality at the edges, than simply destroy everything that doesn't meet some arbitrary and culturally elitist standard. After all, if you do that, you're not even giving the authors of the 'less notable' pages an opportunity to improve.
And finally, deleting articles doesn't just eliminate articles, it disenfranchises people who might otherwise have been contributing members of the community. What does it say to you, if you spent a few days working on an article, only to have some admin decide it's not important enough for Wikipedia and speedy-delete it? Chances are, you're not coming back. By taking a more inclusive approach, Wikipedia would probably have more resources to throw at the problem, because it wouldn't constantly be turning away the people who have the interest, dedication, and expertise to expend on niche subjects.
Frankly, he's a damned looney nutter, and so no, his opinion is not valid. In my own opinion, of course. Sure. But I think the same thing about a lot of people on mainstream TV and in print, too.
Again though you speak of Wikipedia as if it has a fixed volume. You can't "fill it up" with useless trivia -- it has no (effective) size limitation.
Yes, you have to be concerned about pushing the S/N ratio too low, but that could be remedied without constant purging based on subjective guidelines. If an article starts to accrue a lot of cruft or trivia, either just rewrite it more cleanly (preserving the other information, if anyone wants it, in the older versions), or move the trivia to a sub-page. There's no reason why you can't have a page for 'foo' and then a separate page for 'foo trivia' or 'foo in popular culture', if those sections are starting to get out of hand. That lets the people who want to find that information find it, while presenting a concise summary on the main namespace page.
More information is always better; the only bad information is unorganized information. If WP admins were as aggressive about shuffling non-essential stuff into sub-articles and keeping the main namespace clear, it would be fine, and Wikipedia would be broader and deeper as a result.
They pretty much ruined the Dock in this one. And the Finder. (Cover Flow? Seriously? And that iTunes-ish left-side pane thing? It was bad enough in a music jukebox, but now they've got it everywhere. To say nothing about the destruction of spatial viewing.)
There are some neat features but nothing totally compelling. Certainly nothing that would overwhelm the usability problems they've created by fucking around with things that worked fine.
Hopefully there will be an outcry and they'll un-break things in a point release, or in 10.5.1. But if they don't, I think I'll just hold out on this upgrade, particularly since they've made it harder for third-party customization apps to do their thing and repair the damage.
As an Apple customer and Mac user, I lived through the 'lean years' of the mid 90s. And although they did occasionally turn out something controversial, I don't recall them ever boning something up quite this obviously or badly, and then touting it as a feature.
I really love the Mac but jesus they make it hard sometimes. (head in hands)
It's a lot faster to slam my mouse to the bottom of the screen and then pick the application I want out of the Dock than to type it in and pick it out of a Spotlight search result.
I have basically everything that I use in a typical day on my Mac in the Dock (20 applications, not counting the Finder), arranged left to right basically by frequency-of-use (Firefox, Mail, and Terminal are at the far left, Calculator is at the right).
I think it works pretty well, although I have a tremendous hatred for Dashboard (even though I've shut down the process, the damn thing still sits down there in my Dock, taking up space). I was really mixed on the idea of the Dock back around 10.0, because it seemed Windows-ish and foreign, and I still think that it wouldn't be nearly as necessary if they had a functional spatial Finder like they used to (the further destruction of which is one of the reasons why I won't be upgrading to Leopard anytime soon), but it's a lot more intuitive for me than having to search for everything.
I think there's a big difference between people who like choosing options from lists and might prefer a search and list-based interface. I am definitely not in that camp. I don't remember names, I remember icons and spatial relationships (where something was in a window, where it was in relation to other objects, etc.), so Spotlight is unhelpful for me. I use it perhaps a few times a month when I'm looking for some really old document.
Very odd. The official keyboard reference page says that the Firefox Mac shortcut for "next tab" is Cmd-Opt-Tab, but this doesn't work on my system at all. Neither does Cmd-PageDn, which is supposedly the alternative.
Command-Option-[Left|Right]Arrow seem to do it, though.
Honestly, what does it really matter? Information is information, and I thought the goal behind Wikipedia was to centralize as much of it as possible. So long as it's accurate, why does it matter if it's deemed "important"? Importance is hugely subjective - if I were in charge of deciding what articles are important enough to keep in WP, you'd see a whole lot less about Hollywood entertainment, for example. Yet Hollywood information stays - I can go check out Hally Barre's bio if I'm so inclined. Why shouldn't I be able to dig up information on some obscure webcomic, too? I've always thought the same thing. But there are people on Wikipedia who seem to treat the whole project as if bits were a limited resource desperately in need of preservation.
Part of the reason why Wikipedia is cool is because of the sometimes-bizarre breadth and depth of the information in there. Have you ever looked at some of the TV show pages? I won't name names, because I don't want some overzealous admin going in and burning them all, but there are some long-running shows that have pages for every one of hundreds of episodes, that get into incredible minutiea and detail. And I think that's great. That's what makes Wikipedia superior to any other 'encyclopedia' -- every other encyclopedia that's ever been written has been forced to cut and compress content due to the nature of paper-based printing. Wikipedia doesn't, but it sure seems like some people are still thinking that way.
If an article is well-written and the content in it is factual and referenced, I think it's ridiculous to delete it on "notability" grounds, particularly when the 'notability' criteria tend to be debatable and subjective.
Wikipedia is, despite all these things, a good project. But it's sometimes painful to watch because it could be so much more, if it wasn't held back by people quibbling over what "encyclopediac" means. If Wikipedia just kept going and didn't look back, it would redefine what an 'encyclopedia' meant. It could own that word, rather than be shackled by it.
alt-tab works pretty much the same under OS X as it does under Windows I agree that the GP is going out of his way to make work for himself, but Alt-Tab/Cmd-Tab in the Mac OS and in Windows do not work the same way.
In Windows, Alt-Tab cycles between windows. In Mac OS, Cmd-Tab cycles through applications. (Cmd-tilde cycles through an application's windows on the Mac.)
Slashdot Mirror
I think maybe he was confusing MUA and MTA, but I'm not sure.
The encryption really needs to happen at the MUA level, which theoretically could either be a desktop (traditional) MUA or a very smart web app, but I'm not aware of any web apps that will do encryption in the browser. (Not SSL, but actually load up a copy of GPG and encrypt the message.) I talked a little more about this in another post further down in this thread, but what I think the world needs is some sort of lightweight GPG implementation that could run as a Java or JS applet. That way you'd be able to use GPG even through a webmail interface. (You'd still need some sort of private-key management system, not sure how you'd do that. USB stick, maybe?)
But without that, using encrypted email basically means using a traditional MUA, or trusting that your webmail provider won't sell you out to the authorities (e.g. like Hushmail), if the encryption is being done on their server or via untrusted software.
You put the disk in, you turn it on. When you're done, you save and turn it off. If you want to use a different program, you wait for the light on the disk drive to be out, then you swap disks and give it the three-finger salute.
I've seen kindergarteners and first-graders working Apple IIs without any problems, supervision, or assistance at all. All you have to do is reinforce to them that they're not to take the disk out of the drive when the light is on, and it's pretty tough to mess anything up.
In the past I would have recommended Hushmail (http://www.hushmail.com/), but apparently they've cowed to U.S. law enforcement (due to Canadian law) lately, too. Although I wonder whether by cooperating with the Feds they turned over just the encrypted messages, or the actual plaintext? I can't seem to find any details on exactly what level of cooperation they offer.
Really, what you need to find is a web-based service where the encryption/decryption is performed in the user's browser via a JS or Java applet, and thus no unencrypted information is ever transferred to the server. With nothing to turn over, they could happily cooperate. However, I'm not sure this is how Hushmail works -- it sounds like they may do the actual GPG encryption on their server, which of course ruins the entire security model.
The GPG binary is 4MB, so it's a bit large to be using in an applet directly, but I wonder if you couldn't reimplement it in Java or pare it down somehow so that you could embed it?
One of the problems with email is that, to get real security, you want to push the encryption as far out to the edges as possible. But for convenience, people are actually moving towards heavily centralized systems (Gmail, etc.) so that they can access their mail from anywhere. I think the only way you can have both is if you put the encryption engine in the browser somehow.
Don't keep it in your local mailstore (your 'Inbox') unencrypted? When people talk about 'end to end' email security, they don't mean just securing the over-the-wire transmission (e.g. by using SSL); they mean user-to-user encryption.
Most decent MUAs that support encryption leave the messages encrypted, and only decrypt them for the purposes of viewing the message. When you move away from the message or close the window, the ephemeral decrypted copy is destroyed and the only thing on the disk is the encrypted version. (If you're really paranoid, you should probably make sure you have encrypted virtual memory turned on, and that you're using a Tempest-resistant font/display, etc. But even if you don't, you've significantly raised the bar for someone to snoop on you.)
Of course, if you're using an OS and/or software that you don't trust, then you have a problem, since it's entirely feasible that you could get rooted, or maybe the OS or MUA would just be backdoored. More reason for using software that, if it's not practical to actually review it yourself, is at least reviewed by people whose reputations you trust.
And then you have to think about hardware security -- would you know if someone came and swapped your keyboard for an identical model while you were away from your desk? If the answer to that question isn't a sure 'yes,' then you're screwed.
You could use a smart terminal that contains decryption hardware and keep your keys stored on a smartcard; that way all the data stored on the server, and traveling down the wire, would be encrypted, and you'd just have to worry about the small amount of hardware and microcode at your end of things. That's probably much better security than keeping your data stored locally on a platform that's so complicated that you can't ever trust it completely. (You'd probably have to do some of the application-level processing on the client side too, to avoid sending unencrypted data back to the server, e.g. if you wanted to search a document.)
Also, I suspect the majority of PCs in the world are used in workplaces where their owners -- not necessarily their users -- would prefer that all the data be stored in a centralized location. It makes it easier to backup, easier to share, and more difficult for employees to steal or hold hostage.
If the monetary side of things was really there and could be supported by hard evidence, you'd probably see a lot less opposition to the idea in the U.S.
However, that's not the way the debate is framed. It's all about mushy philosophical issues, and there's nothing that makes businesspeople scream "bullshit" and run the other way faster than hearing about what they ought to be doing out of charity or 'compassion.'
Start talking numbers and you might have a case. The people who are going to be swayed by moral arguments already have been; the rest of the country doesn't give a damn.
Agreed. The PC is definitely here to stay.
The "dying" market simply represents the maturity of the technology, and that people are finally getting off the upgrade treadmill. For the past decade or two, although most notably in the past few years, it's been necessary to continually upgrade your machine if you wanted to stay on top of current capabilities. Except for people who play games, I think that's beginning to become no longer true. Or perhaps the new capabilities offered by possible hardware upgrades are no longer as compelling as they once were.
The market has gotten saturated (I suspect moreso in Japan than here in the U.S., where you do occasionally find people who still don't own a computer), and people are not upgrading as quickly. That doesn't mean the PC is "dying," it just means that sales are going to stabilize at the replacement rate.
Personally, I think this is a really good thing for the PC and for software. For years software has been alternately leading and chasing hardware, and there haven't been too many periods where the architecture and capabilities were stable enough to allow for many resources to be devoted to optimization and good design. If the upgrade pace slows, maybe we'll finally get a focus on quality and design, since lazy programmers won't be able to count on hardware increases to make their same bloated code get faster every year.
I expect there'll be a shakeout, eventually, both in the hardware and software side. (The hardware side mostly shook the dead weight out during the dot-bomb, but I think there's a lot of people who are going to get squashed when sales in the established markets plateau and they can't move to other markets or get into maintenance-mode fast enough.) But computers, in general, are here to stay, and the desktop/laptop PC performs too many functions for it to easily be replaced by lots of discrete devices. It's just no longer the frontier.
I really don't understand what all the hate is about for Computer Modern. I think it's a fine font.
Of course, it doesn't look nice on the screen when viewed at 100%. But that's what you get for viewing something at 72 (or 96 or whatever Windows uses) dpi, that's designed to be viewed on paper at 300dpi.
If you blow Computer Modern up to 150% or so, which in my experience tends to be what happens if you fit the width of a document to a good-sized monitor, I think it looks pretty good. But at 10 or 12pt at 100% magnification on a low-resolution device like a computer monitor, you lose all the fine detail that you need.
I guess I can see how there's a need for an alternative for people who are doing all-digital workflows, but if you're going to print the paper out at the end, there's nothing wrong with Computer Modern. It certainly beats the pants off of Times New Roman, IMHO, if you have a good 300+ dpi laser to use as an output device.
buying from a union shop makes it just about impossible for you to be supporting illegal immigrants working in this country for illegally low wages without paying taxes
If buying from a union shop was a guarantee that you weren't supporting illegal labor, I'd be a lot more inclined to buy union. But it isn't, and in fact quite a few of the unions are involved in legalization schemes because they think all the illegal workers will increase their political clout, if they can get them unionized.
In effect, the union organizations are selling out their rank-and-file (who are the ones who really get hit by the wage depression as a result of all the illegal workers) in order to bring in lots of new members and make themselves more powerful. I've seen this to most obvious effect in the unions that have gotten involved in the service industries, but it's pretty widespread if you look for it. Very few of the unions seemed to be putting up much of a fight when push came to shove earlier this year on the immigration issue.
At least the corporations generally are upfront about screwing you; the union leaders seem to enjoy pretending that they're on the side of legitimate workers while doing the same thing.
Public education rams a lot of ideology down its students' throats. That's not an accident -- it's by design. It takes whatever happens to be the majority viewpoint in a particular area, usually as determined by the local school board in conjunction with whatever's written in the textbooks, and impresses that on the students.
Now, generally, that's desirable. There are some ideologies that I think we can probably agree we want rammed down students throats; or, if not that, we can at least agree are necessary to do so, in order to preserve our society. (If you're cynical, you could say that it's necessary in order to produce the next generation of suckers who will slave under the system in order to keep it going.) Things like democracy, tolerance, the rule of law, etc.
There's nothing that inherently prohibits public education from being ideologically loaded. It's prohibited from teaching the specific ideologies that comprise religion, however, by law. But almost anything else is fair game. Outside of religion, it's all a popularity contest.
Although I think some people take the Free Software thing a bit far, I'm not sure it's achieved the status of a true religion just yet. (Though when Stallman dies, all bets are off.) If you could show that there was some social utility in ramming Free Software down students throats, in the same way that the schools treat gender equality and democracy -- both equally artificial concepts -- then there's no real issue.
However, there are distros that offer less crap and have the same core, like Xubuntu, or even by just installing Debian and then pulling down your favorite desktop environment (I like Ratpoison, which is designed to work without a mouse). And unlike W2k, which really doesn't offer you much in the way of an upgrade path anymore, at least a Linux install lets you run a modern kernel, just without the fluff.
(As a datapoint, I'm running a modern Debian Stable kernel on a Pentium 133 with 48MB of RAM. I don't normally use any GUI on it at all, but it's fine for ncurses-based apps and for working as an SSH/VPN endpoint. It'd be tough to do that with most commercial OSes, because it's so old it wouldn't run current versions, and thus wouldn't be secure.)
You had a 20GB drive in 1990? I think you mean 20MB, unless you stole it from the time machine research lab. Or you had your entire basement and several million dollars dedicated to it. :)
I remember when Iomega came out with the 100MB Zip in 1994, and I thought that was cool. I was still using a 80MB drive up until then, although it had started to get really unpleasant. (I had an organizational system for 1.44MB HD floppies that still seems like a ridiculous level of anal-retentiveness today, but it was the only way I could find anything.)
To be honest I find the increase in storage capacity and decrease in cost even more interesting and important in its impact than the increase in processor speeds over the same period.
In 1990 or 91 it would have been a supercomputer, too.
1991 was pre-Pentium; hot stuff then would have been a 68040 if you were a Mac user, and some sort of 486 if you used PCs.
I don't get what people's problem is with the performance of this thing. Granted, I don't upgrade very quickly (I have a dual-1GHz machine and an ancient headless Pentium-133 box under my desk, and the dual-GHz box only replaced a 400MHz one last year), but that's more than enough power for everything except gaming.
Frankly I'd like to see consumer hardware plateau so we can get off of the upgrade treadmill, and the software people can start groveling around retirement communities to find someone who knows what 'optimization' means. It's absolutely ridiculous that people think you can't do typical productivity tasks on a 1.5GHz machine.
Huh?
I'm not sure I understand your point. The drug market doesn't really respond very well to threats of punishment, because many of the substances involved are physically addictive. Thus there's always a demand, regardless of how hard the government cracks down on it. The drug dealers are probably motivated by threats of punishment (in that if the threat is higher, they'll demand more compensation to take the risks, thus driving the cost of drugs higher), but the consumers definitely aren't.
Spam-sending seems like it would be far more sensitive to cost. If you executed everyone who sent spam, and you had an apparatus that was fairly good at catching people who sent it (you got a non-trivial percentage, anyway), the remaining people sending spam would have to be pretty desperate for the money. That would mean they'd charge more, and that would mean that the firms paying for the spam in the first place would have a disincentive to purchase their services.
An online pharmacy isn't addicted to spam in the same way that a junkie is addicted to smack; if there's a method of advertising that's more effective for the price than spam, the online pharmacy is going to do it. (Or, put differently, if spamming is no longer cost-effective, they're not going to pay for it any longer.)
Spam has a very low response rate anyway; if you pushed the cost to send it above what generally comes back from a mailing, it would stop. There's nothing really magical or biologically addictive about it -- it's just an obnoxious economic practice.
I guess that's the other thing that really gets me about WP deletions. Wikipedia is probably the biggest repository of free, GFDLed content around -- it's stuff that could be useful to a lot of other projects, besides Wikipedia. Every time they permanently delete (and they really seem to expunge with gusto -- not even the Internet Archive seems to retain copies of deleted Wikipedia pages) an article for not meeting the notability guidelines, that's a little bit of free content that's being destroyed forever, to say nothing of the labor and effort that went into producing it.
Just as an example, there's an article on the queue for deletion today, which consists of a massive table of character names and their katakana translations. I think it has something to do with a RPG or trading-card name, I'm not sure -- and to be honest I don't really care. I archived it not out of any particular interest in the subject matter but just because it was a ton of information to just destroy, when it's obvious that a lot of work went into its compilation, and it might be of interest to somebody. To me, deleting something like that without first making sure it's archived somewhere, ranks with burning a rare book because it's in a language you can't read: just because it's not interesting or useful to one person doesn't mean it's without value and should be withheld from the world as a whole. There's something offensive about that on a fundamental level.
That's not a flaw in Wikipedia, it's a flaw in people's search heuristics because they're unused to dealing with databases that contain lots of stuff.
You can either fix it through user education, or by making the search program smarter (e.g., more like Google, which when you type in "Alice", tells you that there are 52 million results but gives you the game "Alice" as #1 and "Alice in Wonderland" as #5, because it takes popularity and other factors besides a literal string-match into account).
Removing information from the database in order to make a stupid search program better at going through it is counterproductive.
Unless your metric for quality was to click the "Random Article" button and look at the results (which is a stupid way of doing it), infrequently-accessed articles would have minimal effect on overall perceived quality. Particularly if they were marked as questionable in some way. (I've always been a fan of marking pages that have either only been in existence a short time or have only been worked on by a small number of authors/editors.)
I would rather have greater depth and breadth of information, with high quality at the middle and perhaps somewhat lower (but constantly improving!) quality at the edges, than simply destroy everything that doesn't meet some arbitrary and culturally elitist standard. After all, if you do that, you're not even giving the authors of the 'less notable' pages an opportunity to improve.
And finally, deleting articles doesn't just eliminate articles, it disenfranchises people who might otherwise have been contributing members of the community. What does it say to you, if you spent a few days working on an article, only to have some admin decide it's not important enough for Wikipedia and speedy-delete it? Chances are, you're not coming back. By taking a more inclusive approach, Wikipedia would probably have more resources to throw at the problem, because it wouldn't constantly be turning away the people who have the interest, dedication, and expertise to expend on niche subjects.
Again though you speak of Wikipedia as if it has a fixed volume. You can't "fill it up" with useless trivia -- it has no (effective) size limitation.
Yes, you have to be concerned about pushing the S/N ratio too low, but that could be remedied without constant purging based on subjective guidelines. If an article starts to accrue a lot of cruft or trivia, either just rewrite it more cleanly (preserving the other information, if anyone wants it, in the older versions), or move the trivia to a sub-page. There's no reason why you can't have a page for 'foo' and then a separate page for 'foo trivia' or 'foo in popular culture', if those sections are starting to get out of hand. That lets the people who want to find that information find it, while presenting a concise summary on the main namespace page.
More information is always better; the only bad information is unorganized information. If WP admins were as aggressive about shuffling non-essential stuff into sub-articles and keeping the main namespace clear, it would be fine, and Wikipedia would be broader and deeper as a result.
Yep, that's about right.
They pretty much ruined the Dock in this one. And the Finder. (Cover Flow? Seriously? And that iTunes-ish left-side pane thing? It was bad enough in a music jukebox, but now they've got it everywhere. To say nothing about the destruction of spatial viewing.)
There are some neat features but nothing totally compelling. Certainly nothing that would overwhelm the usability problems they've created by fucking around with things that worked fine.
Hopefully there will be an outcry and they'll un-break things in a point release, or in 10.5.1. But if they don't, I think I'll just hold out on this upgrade, particularly since they've made it harder for third-party customization apps to do their thing and repair the damage.
As an Apple customer and Mac user, I lived through the 'lean years' of the mid 90s. And although they did occasionally turn out something controversial, I don't recall them ever boning something up quite this obviously or badly, and then touting it as a feature.
I really love the Mac but jesus they make it hard sometimes. (head in hands)
It's a lot faster to slam my mouse to the bottom of the screen and then pick the application I want out of the Dock than to type it in and pick it out of a Spotlight search result.
I have basically everything that I use in a typical day on my Mac in the Dock (20 applications, not counting the Finder), arranged left to right basically by frequency-of-use (Firefox, Mail, and Terminal are at the far left, Calculator is at the right).
I think it works pretty well, although I have a tremendous hatred for Dashboard (even though I've shut down the process, the damn thing still sits down there in my Dock, taking up space). I was really mixed on the idea of the Dock back around 10.0, because it seemed Windows-ish and foreign, and I still think that it wouldn't be nearly as necessary if they had a functional spatial Finder like they used to (the further destruction of which is one of the reasons why I won't be upgrading to Leopard anytime soon), but it's a lot more intuitive for me than having to search for everything.
I think there's a big difference between people who like choosing options from lists and might prefer a search and list-based interface. I am definitely not in that camp. I don't remember names, I remember icons and spatial relationships (where something was in a window, where it was in relation to other objects, etc.), so Spotlight is unhelpful for me. I use it perhaps a few times a month when I'm looking for some really old document.
Very odd. The official keyboard reference page says that the Firefox Mac shortcut for "next tab" is Cmd-Opt-Tab, but this doesn't work on my system at all. Neither does Cmd-PageDn, which is supposedly the alternative.
Command-Option-[Left|Right]Arrow seem to do it, though.
Part of the reason why Wikipedia is cool is because of the sometimes-bizarre breadth and depth of the information in there. Have you ever looked at some of the TV show pages? I won't name names, because I don't want some overzealous admin going in and burning them all, but there are some long-running shows that have pages for every one of hundreds of episodes, that get into incredible minutiea and detail. And I think that's great. That's what makes Wikipedia superior to any other 'encyclopedia' -- every other encyclopedia that's ever been written has been forced to cut and compress content due to the nature of paper-based printing. Wikipedia doesn't, but it sure seems like some people are still thinking that way.
If an article is well-written and the content in it is factual and referenced, I think it's ridiculous to delete it on "notability" grounds, particularly when the 'notability' criteria tend to be debatable and subjective.
Wikipedia is, despite all these things, a good project. But it's sometimes painful to watch because it could be so much more, if it wasn't held back by people quibbling over what "encyclopediac" means. If Wikipedia just kept going and didn't look back, it would redefine what an 'encyclopedia' meant. It could own that word, rather than be shackled by it.
In Windows, Alt-Tab cycles between windows. In Mac OS, Cmd-Tab cycles through applications. (Cmd-tilde cycles through an application's windows on the Mac.)