I'm not saying you're wrong, but as you don't link to evidence, and as I've experienced an alternative, I thought I'd point out the other explanation.
In Britain, until 1984, British Telecommunications Plc (better known as "BT") was part of the Post Office, and essentially had a monopoly over all telecommunications. The company was privatized, and the laws that had previously governed telecommunications were very slowly liberalized.
Now, to give you some idea of the scale of what we're talking about, the state of the law in 1983 meant that you couldn't actually wire up a large building with a private phone system. You had to get the Post Office to do it for you. The Post Office argued that it had the right to insist on this, as anything else was a blatent attempt to circumvent its monopoly. Worse, if the system was to be connected to the PSTN (via a PBX or something), then miswiring could cause the entire telephone network to be destroyed.
(But lightning strikes wouldn't cause the same damage. Go figure.)
When the law was liberalized, it remained the case the government had an absolute monopoly on all telecommunications. What they did then was grant two types of license, special telecommunications licenses for entities like BT and Mercury (the half-arsed attempt by Cable and Wireless to build a competitor), and "class licenses" that applied to anyone trying to do a certain type of thing. For example, a class license exists (and does so today, I'm not making this up) that allows you to run telephone wires in your own home.
For a long time, until the EU told them they couldn't do it, you had to get government approval for every piece of equipment you wanted to hook up to the PSTN. Checking for approval meant looking for a device with a green BABT stcker on it. If it didn't have the sticker, you weren't allowed to hook it up. Nobody took it particularly seriously, even BT, the company supposedly most at risk from faulty equipment, ignored violations.
I don't know if Australia's phone service was initially a state monopoly, but if it was, I'd be very surprised if the reason for this type of law isn't similar to the nonsense above. A monopoly is picked apart and an industry deregulated, and little things like the above are left behind, startling anyone who comes across them.
Why this is modded at +5 I'm not sure. What you describe as Open Source is actually what is considered "Shared Source."
Open Source is, from a legal point of view, practically identical to Free Software. The group of people who created the OSI intended the term be used to "sell" Free Software as a concept to businesses by extolling the virtues of a community development process enabled by licenses that gave developers free use of the code created.
While Open Source, in practice, is less about freedom and more about methodologies (sometimes to the detriment of freedom, unlike the FSF, the OSI doesn't urge developers to avoid licenses that are technically free but actually create genuine practical problems), you most certainly are granted more than just the right to see the code.
(Repost, as original has disappeared for some reason. Still appears in my comment list but doesn't appear on the page. Wierd.)
FWIW, OpenBSD was "built with security in mind." *BSD is just a family of operating systems spawned by a long term project that started with some University hackers in the mid-seventies. It's like a boy-band, OpenBSD is the tough one, NetBSD is the slightly wierd one that gets on with everyone, FreeBSD into whose pants most teenage girls want to be.
Well, ok, scrub that: FreeBSD is, near as damn it, the latest version of BSD "proper", NetBSD is the portable one, and OpenBSD is the "secure" one. But none are based upon anything "built with security in mind", far from it: Unix was never built with security in mind, and the major thing the OBSD people have been doing has been to lock down the default distribution so it has as few back-doors as possible, together with encouraging use of protocols like SSH so passwords are transmitted in clear text form a little less often.
How does this relate to Mac OS X? Well, not a lot, to be honest. OS X is the latest version of NeXTStep, and the developers essentially updated the userland of the lower levels of NeXTStep, then BSD based too, with the latest from modern BSDs. The result is less bugs and more compatability. Less bugs = slightly more security, but if Apple's other layers have bugs, then that doesn't mean a whole lot. The underlying operating system is still based upon Unix's old security model.
Mac OS X has a few things that operating systems like Windows do not that arguably make it less secure. The default settings of the default browser, Safari, for example, will save and open many types of file automatically without even prompting the user where to save the thing. An extreme example are.sit files. When expanded, any executable programs within them automatically become registered with the system (unlike Windows, the mere presense of an application on your hard disk is enough for Mac OS and Mac OS X to feel it can use it - run it, open files associated with it, etc.) Recent versions of Panther have at least tried to fix the obvious hole there by requiring a user confirm that a program should be run the first time they run it. But it's not exactly fool proof, as a million Windows users who have said yes, they do want to install ActiveSpyWare to view the content on this webpage, can demonstrate.
Panther users on dial-up, for whom downloading a multi-megabyte update isn't really something they want to do, as well as Jaguar users, are still vulnerable, obviously.
I've said it before and I'll say it again: there's absolutely nothing inherently more secure about Mac OS X compared to modern versions of Windows. That is, (NT based Windows, rather than DOS based versions. With proper memory management and a decent rights-based security system. It's a shame the default config of XP et al isn't inherently secure, but alas, that will remain the case as long as usability is valued over security. And anyone who thinks that it isn't the case that usability is valued over security in OS X needs a good kicking...) In both systems, the default user has admin privileges. Both rely, for the most part, on users making informed decisions that frequently they don't have the information for. Both get regular security updates from their vendors, patching holes. OS X hasn't been targetted by malware and virus writers, but it's a matter of time. Until very recently, OS X simply wasn't in the hands of the bulk of malware and virus writers. An OS with less than 2% of the market just isn't worth wasting your time on.
This isn't a bad thing. It's given Apple the chance to put in hacks to fix holes like the above that, otherwise, black hats would have driven trains through. But it remains the case: Anyone who claims that OS X is "inherently more secure" is doing nobody any favours. At some point, it is going to be a big enough target. People will start attacking it. Those who've made these arguments will ensure the problems are far more serious than they would have been otherwise.
It's kind of funny. I see people making claims all the time about Mac OS X that I'm pretty sure Theo De Raadt wouldn't make about OpenBSD.
there's absolutely nothing inherently more secure about Mac OS X compared to Windows
That should, of course, mean:
there's absolutely nothing inherently more secure about Mac OS X compared to modern versions of Windows.
(NT based Windows, rather than DOS based versions. With proper memory management and a decent rights-based security system. It's a shame the default config of XP et al isn't inherently secure, but alas, that will remain the case as long as usability is valued over security. And anyone who thinks that it isn't the case that usability is valued over security in OS X needs a good kicking...)
It's kind of funny. I see people making claims all the time about Mac OS X that I'm pretty sure Theo De Raadt wouldn't make about OpenBSD.
FWIW, OpenBSD was "built with security in mind." *BSD is just a family of operating systems spawned by a long term project that started with some University hackers in the mid-seventies. It's like a boy-band, OpenBSD is the tough one, NetBSD is the slightly wierd one that gets on with everyone, FreeBSD into whose pants most teenage girls want to be.
Well, ok, scrub that: FreeBSD is, near as damn it, the latest version of BSD "proper", NetBSD is the portable one, and OpenBSD is the "secure" one. But none are based upon anything "built with security in mind", far from it: Unix was never built with security in mind, and the major thing the OBSD people have been doing has been to lock down the default distribution so it has as few back-doors as possible, together with encouraging use of protocols like SSH so passwords are transmitted in clear text form a little less often.
How does this relate to Mac OS X? Well, not a lot, to be honest. OS X is the latest version of NeXTStep, and the developers essentially updated the userland of the lower levels of NeXTStep, then BSD based too, with the latest from modern BSDs. The result is less bugs and more compatability. Less bugs = slightly more security, but if Apple's other layers have bugs, then that doesn't mean a whole lot. The underlying operating system is still based upon Unix's old security model.
Mac OS X has a few things that operating systems like Windows do not that arguably make it less secure. The default settings of the default browser, Safari, for example, will save and open many types of file automatically without even prompting the user where to save the thing. An extreme example are.sit files. When expanded, any executable programs within them automatically become registered with the system (unlike Windows, the mere presense of an application on your hard disk is enough for Mac OS and Mac OS X to feel it can use it - run it, open files associated with it, etc.) Recent versions of Panther have at least tried to fix the obvious hole there by requiring a user confirm that a program should be run the first time they run it. But it's not exactly fool proof, as a million Windows users who have said yes, they do want to install ActiveSpyWare to view the content on this webpage, can demonstrate.
Panther users on dial-up, for whom downloading a multi-megabyte update isn't really something they want to do, as well as Jaguar users, are still vulnerable, obviously.
I've said it before and I'll say it again: there's absolutely nothing inherently more secure about Mac OS X compared to Windows. In both systems, the default user has admin privileges. Both rely, for the most part, on users making informed decisions that frequently they don't have the information for. Both get regular security updates from their vendors, patching holes. OS X hasn't been targetted by malware and virus writers, but it's a matter of time. Until very recently, OS X simply wasn't in the hands of the bulk of malware and virus writers. An OS with less than 2% of the market just isn't worth wasting your time on.
This isn't a bad thing. It's given Apple the chance to put in hacks to fix holes like the above that, otherwise, black hats would have driven trains through. But it remains the case: Anyone who claims that OS X is "inherently more secure" is doing nobody any favours. At some point, it is going to be a big enough target. People will start attacking it. Those who've made these arguments will ensure the problems are far more serious than they would have been otherwise.
For a website which is devoted to shoveling up information for the most elitist of all computer-literate people [including some bright individuals], you'd think that somehow, a better system could be put into place than "bomb websites with loads of traffic, indiscriminantly".
That's what Coral is supposed to fix. You could at least be consistant, either tell office workers they shouldn't complain about not being able to use Coral, and rejoice in the overloading of servers this results in, or support Coral changing to a more regular port.
s/commercial/proprietary/g, and even after that it's still wrong.
RMS can't force you to do anything as long as you're the copyright holder, as has been pointed out many times before. If you write some code, and add it to a GPL'd work, you make a work that's derived from that GPL'd work and the combination (but not your original) is licensed under the GPL, which will restrict usage of the combination in proprietary products. (But not commercial projects, people are still free to sell copies, sell support, sell programming time to add custom features, use within a business as an engine for a critical process that makes the company money, etc.)
But your original code is still copyrighted by you, and you can still use it as part of proprietary products when you don't combine it with GPL'd works.
You can even release it on its own under the GPL, and offer companies that want to buy rights to release proprietary versions custom licenses, in exchange for hard cash, of course.
Nothing RMS will ever do will change that. It's not really possible to modify the GPL to remove rights from the copyright holder. A license is something that grants rights to someone who didn't previously have them.
The best RMS can do is be elected President, have the rest of the FSF take over Congress, and abolish copyright, and if that happens, "trusting RMS" as far as new versions of the GPL goes will not be your problem.
Well, my counter to that is it most certainly should be meant to improve usability. I know that Seth's reason for posting the article is to say "Wheee! Look at this, look at what we can do!", but without context "what we can do" is useless. The context here is that the techniques are designed to improve UIs in various ways.
While a lot of Slashdotters and other geeks find a lot of pleasure in eye-candy without regard to usability, I think it's refreshing that Seth actually did post some examples of techniques used where they had an intuitively obvious improvement on usability. If he hadn't, I'd have ignored the demonstrations, or even flamed them. If everything had been like the initial wobbly windows effect, I'd have put it down as yet another thing that'll pointlessly bloat applications in a year or two in order to satisfy the "Ooo look, pretty colours!" mob.
Context is important. You can't really demonstrate a technique without showing that it's potentially useful. I think Seth, for the most part, wobbly windows aside, did a great job doing just that.
There's some nice ideas in there, and some not so nice ones. The wobbly windows thing looks completely unnecessary (worse still, I get it for free when I try to drag opaque windows on a slow machine;-), and it's hard to see how it can actually improve usability.
On the other hand, the similar effect applied to drop down menus did make some sense. It made the menu appearing more obvious and anyone glancing at an unrelated part of the screen and accidentally activating the menu would be more aware of their mistake with this kind of heavily animated approach. It also looked like it wouldn't get in the way, the way it was implemented.
I also liked the translucent file selector. That's the first time I've seen translucency done in a relevant, useful, manner. Yes, I do want to see the window underneath, damn it! Combined with Apple's "attaching selectors to the window they came from" philosophy, you could have quite a massive improvement in usability.
It's nice to see some of the techniques developed largely as eye-candy actually find uses where they have functional, not just subjectively aesthetic, justification.
The other side of the coin is: if you're an employer, and you don't like unions, don't give your employees any reason to join one.
BTW, the laziness you cite is common throughout the industry, it's just anti-union kooks notice it more with unionized companies. One of the most absurdly anti-union companies in the US at the moment is Wal*Mart.
Tell us what you think about the average Wal*Mart employee's attitude to work.
They certainly can! I mean, games are a prime example of the type of stuff that can be outsourced to a group of generic coders, as opposed to developed by a hand picked team of people with specific skills in specific areas. Games, by and large:
Are written in collaborative languages like COBOL, C#, and Java
Require little or optimization
Require no imagination or development of new ideas. Games are usually simple black boxes, easily spec'd, with no unusual programming techniques required
Require no specialized art beyond simple UI icons and existing corporate logos
Whoever heard of a game with audio? Only programmers need to be involved in the development of new games
Require similar skillsets amongst all developers. A programmer working on one aspect of a game's design is almost certainly suited to developing every other part. So different programmers can be swapped in and out of development as needed, and to increase development speed, all you have to do is throw in more programmers
I've used a wide variety of games in my time. Most, of course, were bespoke, developed for specific giant corporations, to manage their payrolls for example, or online ordering systems. Just because greedy people like John Carmack have made their millions through automated bank statement printers (thanks Carmack's "Quake 3 Arena" for our outrageous banking fees!) and supermarket inventory control systems, doesn't mean we need people like them developing the next generation
So don't tell me that you can't out-source games development! It's just a matter of firing everyone and sending the specs of any new games you need developed to an outside agency.
(Yes, it's sarcasm)
(Yes, I'm aware there's probably a lot of talented people all over the world. But that's not what outsourcing is. Outsourcing is about making use of shared pools of programmers operating according to specs that have travelled half way around the world, who can program more cheaply than the people they're replacing. If a job requires talent, that's not possible. You can, obviously, open an EA office in India and headhunt the best hardware hackers, artists, etc, but that's not exactly something you can do overnight.)
Anti-Free Market perhaps, but not anti-Capitalism. Can you invest in an enterprise and make a profit from doing so with copyrights in place?
Most certainly. Indeed, the more arbitrary monopolies you allow people to make for themselves, the easier it is to create opportunities for that investment.
An example, methinks, of how the Free Market isn't always the same thing as Capitalism.
It's a little more than that. It's not just that bad.site is treated as identical to good.site, it's that good.site is potentially removed from Google. "302" means "temporary redirect", which gives Google the false idea that good.site isn't a permanent website.
Whether it actually removes good.site from the index has to do with, apparently, the PageRank of both sites.
It really wasn't until I read a full explanation and they covered that bit that the whole thing clicked for me.
IP addresses contain network information, not physical location information. Even if an ISP can tell you what POP an IP address is assigned to, it's not uncommon for larger ISPs to have phone companies provide "VPOP"s (virtual POPs) where calls in a large calling area are all routed to a central bank of modems. At an extreme, almost every ISP in Britain did this in the late nineties with every single connection being routed via local numbers (or a "local rate" area code) being routed by the phone companies to a central office for that ISP. Why? Because it's cheaper than having an office in every local calling area. And it's more efficient in terms of use of IP addresses and modems.
The best you can have is to ask ISPs to support some form of protocol that would take an IP address and return some kind of geographic information related to the user of that IP address at that very instant. Presumably dialup users would take that from the CLI of the phone calling the modem bank and assigned to that IP address. DSL and Cable would have to be based on the account user's information. It's relatively hairy, there are a lot of ways in which it can fail to provide reliable information.
If Vonage is pitching itself as a viable replacement for landline telephone service, it should be responsible for this education itself. Without information, people can't educate themselves.
As a general rule, I'm a strong believer that any argument that consumers must be responsible has to be balanced by responsibility from the producers of the products they consume. That means clear warnings about how a product or service might differ from expectations, critical safety issues, etc. If an informed consumer in a genuinely competitive market then chooses a product unsuited to them, then that's their problem.
Looking at Vonage's product page, it's clear 911 is advertised as a benefit, not an optional extra that isn't a real 911 service. While the information that it's actually an optional (albeit free) feature is a click away, there's no actual reason in the way 911 service is presented to a potential customer to encourage such a person to assume there would be more to it, and that clicking on that link actually provides important information.
The link, in some ways, is discredited by what it links to. "911 Dialing" isn't a "great benefit" of Vonage, it's optional, it's not a real 911 service ("Your Call Will Go To A General Access Line at the Public Safety Answering Point (PSAP). This is different from the 911 Emergency Response Center where traditional 911 calls go."), and in some circumstances will not work anyway.
I think the more something becomes a critical safety issue, the more important it is for a provider to be upfront and straightforward and tell the truth - and not just do so by avoiding lying or using turns of phrase it knows will be misunderstood - but to actually say "If you get this product, you should know that XYZ will have the following limitations".
Vonage isn't doing this. I think they should. And I think, be it through lawsuits or FCC actions, they should be forced to.
It's the very first "Great benefit!" on their products page. Now, if you click on that link (and why would you? Most people will assume saying that 911 is a "great benefit" of your phone service means exactly that), it then explains how the 911 system works and how it's actually optional (albeit free), not a standard part of the service at all.
I personally wouldn't have followed the link had I not read this article. It all looks a little more flaky than I thought was the case.
Piracy has meant copyright infringement (and illegal radio transmitter use) now for many decades, if not, in the former case, centuries. Complaining that it dishonours victims of "real" pirates (sea mobsters) to use it in this way is a little like complaining that using the word "rape" to describe the yellow crop farmed to create canola oil dishonours victims of the violent offense also associated with that word.
Like it or not, piracy - in the English language - has more than one meaning.
Most of the eighties, at least in Britain, and from what I can Google, in the US and Europe too. Games consoles rose during the seventies, and sales then disappeared rapidly in the early eighties when the home computing boom was in full swing. People prefered Commodore 64s and Sinclair Spectrums as they were "full computers", as opposed to Atari, et al's, best efforts at the time.
Wikipedia has an interesting article about the sudden crash of many manufacturers in 1983. One of the few survivors was Atari who cancelled the 7800 (bringing it back later in the eighties but still to no avail.) I didn't really see people buying consoles, or see them taken seriously, until the early nineties, when Nintendo started to make serious headway in the US and Sega in Europe, the latter after several false starts. Both only really succeeded because of the strength of the Japanese market during the eighties.
Games consoles seem to be doing remarkably well right now despite the fact that just about any off-the-shelf PC has better hardware once you put in a modern graphics card. Indeed, I don't think any of the big 3 were ahead of the PC when they first came out. About the only way in which they were/are superior was in price.
I agree. In that sense, the PC will never be dead. Different markets will have different needs, and while individuals have need of computing, they'll have personal computers. These may be highly optimized platforms like games consoles (eg optimized to one type of application) or more universal systems. Time has told us that people are never happy with a single, limited, box - when games consoles went up against home computers, the latter won. Games consoles only came back when it became normal to have both a computer and a console.
It's a li[tt]le like the ocean. You have your sharks and dolphins (big businesses and little businesses, with specific business needs), and you have your regular fish - clownfish, for example, for those who liked "Finding Nemo", and cod. While they all may swim in the same ocean and have similar needs, the fact these needs aren't identical means they end up eating different things. Sharks, for example, will happily eat seals, not so cod. What you end up with is a different style, sharks will not even hunt for their food in the same way as smaller fish. An algae-eater, for example, will constantly be feeding on the walls of coral and other areas where algae may hang out.
In the same way, centralised computer systems may make sense for businesses. But for individuals, families, and other households, they're just inappropriate. A large business can eat a seal and not have to feed again for a while, but an algae-eating games player needs localised power at their fingertips to provide them with the game playing environment they crave. Grandma, wanting to surf the net or write email, will want the computing equivalent of plankton, power available when she needs it, localized to her.
Personal computing will simply never die. It will go through periods of being more or less application specific, but I suspect if you were to draw an image of the average household in 2015, then, like it was in 1995, will you see a PC in every home. Just as you do more with your PC today - manage MP3 archives, view remote web pages, etc - than you did in '95, so will the PC of 2015 be a more sophisticated, more important, engine.
Because it's difficult to access those links when you're stuck behind corporate firewalls. Coral uses port 8090, a non-standard port that most firewalls are unaware of and block.
It's like a restaurant. You're stuck with the menu the restaurant has. Now, it's not that you can't necessarily get the kitchen to do a ham and cheese, but you have to do it in terms it understands (for example, you can order a burger that has ham and cheese, and order it without the beefburger, salad, etc), kind of like h[tt]p:, which runs on port 80. You can do it via the firewall, but it has to look like an HTTP request, which means running it on port 80. You can then say "Ohh, it's not really a burger, it's a ham and cheese sandwich" but as far as the kitchen's concerned, it's just one of their regular burgers. You might look at port 8090 as the ham - they're likely to have cheese burgers, but a ham, cheese, and beef burger? Not likely. So you can't have your ham and cheese because you haven't come up with a sandwich that really works within the framework you're given.
The only option is to leave the restaurant, and cook your own sandwich, but that's not always an option, especially if you actually work at the restaurant so can't leave until 5pm, but you're a waiter or you work at the bar or you greet people or wash up or something so you can't actually make the sandwich yourself (well, not in a unionized restaurant anyway. A union-free restaurant might allow it, but you don't want to upset the staff, and it's probably going against company policy.)
Port 8090 isn't supported by most corporate firewalls, so making all URLs point at it would just prevent Slashdot's working readers (the vast majority) from "eating their ham and cheese sandwich" - or, in other words, accessing the website. This would damage Slashdot long term as people would just stop reading it except for a few people at Universities and in Cybercafes, neither of which are appealing to Slashdot's advertisers.
As far as the first idea, you're underestimating the availability of competent ISPs;-) Even those that offer static IP addresses often put the addresses in the DUL anyway.
I don't know if Earthlink, my erstwhile ISP (don't have DSL available in my new home), does it, but they do block outgoing port 25 anyway, so I might as well have been on the DUL.
Damn it!
I think virtual servers are becoming increasingly the only way to get certain things done that ought to be perfectly fine but aren't because ISPs put dumb things in their T&Cs, thanks to the antics of the extremists amongst the anti-spammers.
Ah, so it's another thing designed to annoy anyone who wants to manage their own email, but whose choice of local ISPs is restricted to those that arbitrarily sign up to the DUL, for always-on and dial-up connections alike.
Great. Fantastic. *throws up*
(Sorry, see my most recent JE for the explanation of that last bit. Nothing to do with this. I think, anyway.)
Slashdot Mirror
In Britain, until 1984, British Telecommunications Plc (better known as "BT") was part of the Post Office, and essentially had a monopoly over all telecommunications. The company was privatized, and the laws that had previously governed telecommunications were very slowly liberalized.
Now, to give you some idea of the scale of what we're talking about, the state of the law in 1983 meant that you couldn't actually wire up a large building with a private phone system. You had to get the Post Office to do it for you. The Post Office argued that it had the right to insist on this, as anything else was a blatent attempt to circumvent its monopoly. Worse, if the system was to be connected to the PSTN (via a PBX or something), then miswiring could cause the entire telephone network to be destroyed.
(But lightning strikes wouldn't cause the same damage. Go figure.)
When the law was liberalized, it remained the case the government had an absolute monopoly on all telecommunications. What they did then was grant two types of license, special telecommunications licenses for entities like BT and Mercury (the half-arsed attempt by Cable and Wireless to build a competitor), and "class licenses" that applied to anyone trying to do a certain type of thing. For example, a class license exists (and does so today, I'm not making this up) that allows you to run telephone wires in your own home.
For a long time, until the EU told them they couldn't do it, you had to get government approval for every piece of equipment you wanted to hook up to the PSTN. Checking for approval meant looking for a device with a green BABT stcker on it. If it didn't have the sticker, you weren't allowed to hook it up. Nobody took it particularly seriously, even BT, the company supposedly most at risk from faulty equipment, ignored violations.
I don't know if Australia's phone service was initially a state monopoly, but if it was, I'd be very surprised if the reason for this type of law isn't similar to the nonsense above. A monopoly is picked apart and an industry deregulated, and little things like the above are left behind, startling anyone who comes across them.
Open Source is, from a legal point of view, practically identical to Free Software. The group of people who created the OSI intended the term be used to "sell" Free Software as a concept to businesses by extolling the virtues of a community development process enabled by licenses that gave developers free use of the code created.
While Open Source, in practice, is less about freedom and more about methodologies (sometimes to the detriment of freedom, unlike the FSF, the OSI doesn't urge developers to avoid licenses that are technically free but actually create genuine practical problems), you most certainly are granted more than just the right to see the code.
FWIW, OpenBSD was "built with security in mind." *BSD is just a family of operating systems spawned by a long term project that started with some University hackers in the mid-seventies. It's like a boy-band, OpenBSD is the tough one, NetBSD is the slightly wierd one that gets on with everyone, FreeBSD into whose pants most teenage girls want to be.
Well, ok, scrub that: FreeBSD is, near as damn it, the latest version of BSD "proper", NetBSD is the portable one, and OpenBSD is the "secure" one. But none are based upon anything "built with security in mind", far from it: Unix was never built with security in mind, and the major thing the OBSD people have been doing has been to lock down the default distribution so it has as few back-doors as possible, together with encouraging use of protocols like SSH so passwords are transmitted in clear text form a little less often.
How does this relate to Mac OS X? Well, not a lot, to be honest. OS X is the latest version of NeXTStep, and the developers essentially updated the userland of the lower levels of NeXTStep, then BSD based too, with the latest from modern BSDs. The result is less bugs and more compatability. Less bugs = slightly more security, but if Apple's other layers have bugs, then that doesn't mean a whole lot. The underlying operating system is still based upon Unix's old security model.
Mac OS X has a few things that operating systems like Windows do not that arguably make it less secure. The default settings of the default browser, Safari, for example, will save and open many types of file automatically without even prompting the user where to save the thing. An extreme example are .sit files. When expanded, any executable programs within them automatically become registered with the system (unlike Windows, the mere presense of an application on your hard disk is enough for Mac OS and Mac OS X to feel it can use it - run it, open files associated with it, etc.) Recent versions of Panther have at least tried to fix the obvious hole there by requiring a user confirm that a program should be run the first time they run it. But it's not exactly fool proof, as a million Windows users who have said yes, they do want to install ActiveSpyWare to view the content on this webpage, can demonstrate.
Panther users on dial-up, for whom downloading a multi-megabyte update isn't really something they want to do, as well as Jaguar users, are still vulnerable, obviously.
I've said it before and I'll say it again: there's absolutely nothing inherently more secure about Mac OS X compared to modern versions of Windows. That is, (NT based Windows, rather than DOS based versions. With proper memory management and a decent rights-based security system. It's a shame the default config of XP et al isn't inherently secure, but alas, that will remain the case as long as usability is valued over security. And anyone who thinks that it isn't the case that usability is valued over security in OS X needs a good kicking...) In both systems, the default user has admin privileges. Both rely, for the most part, on users making informed decisions that frequently they don't have the information for. Both get regular security updates from their vendors, patching holes. OS X hasn't been targetted by malware and virus writers, but it's a matter of time. Until very recently, OS X simply wasn't in the hands of the bulk of malware and virus writers. An OS with less than 2% of the market just isn't worth wasting your time on.
This isn't a bad thing. It's given Apple the chance to put in hacks to fix holes like the above that, otherwise, black hats would have driven trains through. But it remains the case: Anyone who claims that OS X is "inherently more secure" is doing nobody any favours. At some point, it is going to be a big enough target. People will start attacking it. Those who've made these arguments will ensure the problems are far more serious than they would have been otherwise.
It's kind of funny. I see people making claims all the time about Mac OS X that I'm pretty sure Theo De Raadt wouldn't make about OpenBSD.
It's kind of funny. I see people making claims all the time about Mac OS X that I'm pretty sure Theo De Raadt wouldn't make about OpenBSD.
Well, ok, scrub that: FreeBSD is, near as damn it, the latest version of BSD "proper", NetBSD is the portable one, and OpenBSD is the "secure" one. But none are based upon anything "built with security in mind", far from it: Unix was never built with security in mind, and the major thing the OBSD people have been doing has been to lock down the default distribution so it has as few back-doors as possible, together with encouraging use of protocols like SSH so passwords are transmitted in clear text form a little less often.
How does this relate to Mac OS X? Well, not a lot, to be honest. OS X is the latest version of NeXTStep, and the developers essentially updated the userland of the lower levels of NeXTStep, then BSD based too, with the latest from modern BSDs. The result is less bugs and more compatability. Less bugs = slightly more security, but if Apple's other layers have bugs, then that doesn't mean a whole lot. The underlying operating system is still based upon Unix's old security model.
Mac OS X has a few things that operating systems like Windows do not that arguably make it less secure. The default settings of the default browser, Safari, for example, will save and open many types of file automatically without even prompting the user where to save the thing. An extreme example are .sit files. When expanded, any executable programs within them automatically become registered with the system (unlike Windows, the mere presense of an application on your hard disk is enough for Mac OS and Mac OS X to feel it can use it - run it, open files associated with it, etc.) Recent versions of Panther have at least tried to fix the obvious hole there by requiring a user confirm that a program should be run the first time they run it. But it's not exactly fool proof, as a million Windows users who have said yes, they do want to install ActiveSpyWare to view the content on this webpage, can demonstrate.
Panther users on dial-up, for whom downloading a multi-megabyte update isn't really something they want to do, as well as Jaguar users, are still vulnerable, obviously.
I've said it before and I'll say it again: there's absolutely nothing inherently more secure about Mac OS X compared to Windows. In both systems, the default user has admin privileges. Both rely, for the most part, on users making informed decisions that frequently they don't have the information for. Both get regular security updates from their vendors, patching holes. OS X hasn't been targetted by malware and virus writers, but it's a matter of time. Until very recently, OS X simply wasn't in the hands of the bulk of malware and virus writers. An OS with less than 2% of the market just isn't worth wasting your time on.
This isn't a bad thing. It's given Apple the chance to put in hacks to fix holes like the above that, otherwise, black hats would have driven trains through. But it remains the case: Anyone who claims that OS X is "inherently more secure" is doing nobody any favours. At some point, it is going to be a big enough target. People will start attacking it. Those who've made these arguments will ensure the problems are far more serious than they would have been otherwise.
RMS can't force you to do anything as long as you're the copyright holder, as has been pointed out many times before. If you write some code, and add it to a GPL'd work, you make a work that's derived from that GPL'd work and the combination (but not your original) is licensed under the GPL, which will restrict usage of the combination in proprietary products. (But not commercial projects, people are still free to sell copies, sell support, sell programming time to add custom features, use within a business as an engine for a critical process that makes the company money, etc.)
But your original code is still copyrighted by you, and you can still use it as part of proprietary products when you don't combine it with GPL'd works.
You can even release it on its own under the GPL, and offer companies that want to buy rights to release proprietary versions custom licenses, in exchange for hard cash, of course.
Nothing RMS will ever do will change that. It's not really possible to modify the GPL to remove rights from the copyright holder. A license is something that grants rights to someone who didn't previously have them. The best RMS can do is be elected President, have the rest of the FSF take over Congress, and abolish copyright, and if that happens, "trusting RMS" as far as new versions of the GPL goes will not be your problem.
While a lot of Slashdotters and other geeks find a lot of pleasure in eye-candy without regard to usability, I think it's refreshing that Seth actually did post some examples of techniques used where they had an intuitively obvious improvement on usability. If he hadn't, I'd have ignored the demonstrations, or even flamed them. If everything had been like the initial wobbly windows effect, I'd have put it down as yet another thing that'll pointlessly bloat applications in a year or two in order to satisfy the "Ooo look, pretty colours!" mob.
Context is important. You can't really demonstrate a technique without showing that it's potentially useful. I think Seth, for the most part, wobbly windows aside, did a great job doing just that.
On the other hand, the similar effect applied to drop down menus did make some sense. It made the menu appearing more obvious and anyone glancing at an unrelated part of the screen and accidentally activating the menu would be more aware of their mistake with this kind of heavily animated approach. It also looked like it wouldn't get in the way, the way it was implemented.
I also liked the translucent file selector. That's the first time I've seen translucency done in a relevant, useful, manner. Yes, I do want to see the window underneath, damn it! Combined with Apple's "attaching selectors to the window they came from" philosophy, you could have quite a massive improvement in usability.
It's nice to see some of the techniques developed largely as eye-candy actually find uses where they have functional, not just subjectively aesthetic, justification.
Yeah, Wal*Mart workers are free to give money to unions. They just better not join them, or if they do, take part in any union activity.
BTW, the laziness you cite is common throughout the industry, it's just anti-union kooks notice it more with unionized companies. One of the most absurdly anti-union companies in the US at the moment is Wal*Mart.
Tell us what you think about the average Wal*Mart employee's attitude to work.
- Are written in collaborative languages like COBOL, C#, and Java
- Require little or optimization
- Require no imagination or development of new ideas. Games are usually simple black boxes, easily spec'd, with no unusual programming techniques required
- Require no specialized art beyond simple UI icons and existing corporate logos
- Whoever heard of a game with audio? Only programmers need to be involved in the development of new games
- Require similar skillsets amongst all developers. A programmer working on one aspect of a game's design is almost certainly suited to developing every other part. So different programmers can be swapped in and out of development as needed, and to increase development speed, all you have to do is throw in more programmers
I've used a wide variety of games in my time. Most, of course, were bespoke, developed for specific giant corporations, to manage their payrolls for example, or online ordering systems. Just because greedy people like John Carmack have made their millions through automated bank statement printers (thanks Carmack's "Quake 3 Arena" for our outrageous banking fees!) and supermarket inventory control systems, doesn't mean we need people like them developing the next generationSo don't tell me that you can't out-source games development! It's just a matter of firing everyone and sending the specs of any new games you need developed to an outside agency.
(Yes, it's sarcasm)
(Yes, I'm aware there's probably a lot of talented people all over the world. But that's not what outsourcing is. Outsourcing is about making use of shared pools of programmers operating according to specs that have travelled half way around the world, who can program more cheaply than the people they're replacing. If a job requires talent, that's not possible. You can, obviously, open an EA office in India and headhunt the best hardware hackers, artists, etc, but that's not exactly something you can do overnight.)
Most certainly. Indeed, the more arbitrary monopolies you allow people to make for themselves, the easier it is to create opportunities for that investment.
An example, methinks, of how the Free Market isn't always the same thing as Capitalism.
Whether it actually removes good.site from the index has to do with, apparently, the PageRank of both sites.
It really wasn't until I read a full explanation and they covered that bit that the whole thing clicked for me.
The best you can have is to ask ISPs to support some form of protocol that would take an IP address and return some kind of geographic information related to the user of that IP address at that very instant. Presumably dialup users would take that from the CLI of the phone calling the modem bank and assigned to that IP address. DSL and Cable would have to be based on the account user's information. It's relatively hairy, there are a lot of ways in which it can fail to provide reliable information.
As a general rule, I'm a strong believer that any argument that consumers must be responsible has to be balanced by responsibility from the producers of the products they consume. That means clear warnings about how a product or service might differ from expectations, critical safety issues, etc. If an informed consumer in a genuinely competitive market then chooses a product unsuited to them, then that's their problem.
Looking at Vonage's product page, it's clear 911 is advertised as a benefit, not an optional extra that isn't a real 911 service. While the information that it's actually an optional (albeit free) feature is a click away, there's no actual reason in the way 911 service is presented to a potential customer to encourage such a person to assume there would be more to it, and that clicking on that link actually provides important information.
The link, in some ways, is discredited by what it links to. "911 Dialing" isn't a "great benefit" of Vonage, it's optional, it's not a real 911 service ("Your Call Will Go To A General Access Line at the Public Safety Answering Point (PSAP). This is different from the 911 Emergency Response Center where traditional 911 calls go."), and in some circumstances will not work anyway.
I think the more something becomes a critical safety issue, the more important it is for a provider to be upfront and straightforward and tell the truth - and not just do so by avoiding lying or using turns of phrase it knows will be misunderstood - but to actually say "If you get this product, you should know that XYZ will have the following limitations".
Vonage isn't doing this. I think they should. And I think, be it through lawsuits or FCC actions, they should be forced to.
I personally wouldn't have followed the link had I not read this article. It all looks a little more flaky than I thought was the case.
Like it or not, piracy - in the English language - has more than one meaning.
(I sense a potential new meme)
Wikipedia has an interesting article about the sudden crash of many manufacturers in 1983. One of the few survivors was Atari who cancelled the 7800 (bringing it back later in the eighties but still to no avail.) I didn't really see people buying consoles, or see them taken seriously, until the early nineties, when Nintendo started to make serious headway in the US and Sega in Europe, the latter after several false starts. Both only really succeeded because of the strength of the Japanese market during the eighties.
Maybe that's it...
It's a li[tt]le like the ocean. You have your sharks and dolphins (big businesses and little businesses, with specific business needs), and you have your regular fish - clownfish, for example, for those who liked "Finding Nemo", and cod. While they all may swim in the same ocean and have similar needs, the fact these needs aren't identical means they end up eating different things. Sharks, for example, will happily eat seals, not so cod. What you end up with is a different style, sharks will not even hunt for their food in the same way as smaller fish. An algae-eater, for example, will constantly be feeding on the walls of coral and other areas where algae may hang out.
In the same way, centralised computer systems may make sense for businesses. But for individuals, families, and other households, they're just inappropriate. A large business can eat a seal and not have to feed again for a while, but an algae-eating games player needs localised power at their fingertips to provide them with the game playing environment they crave. Grandma, wanting to surf the net or write email, will want the computing equivalent of plankton, power available when she needs it, localized to her.
Personal computing will simply never die. It will go through periods of being more or less application specific, but I suspect if you were to draw an image of the average household in 2015, then, like it was in 1995, will you see a PC in every home. Just as you do more with your PC today - manage MP3 archives, view remote web pages, etc - than you did in '95, so will the PC of 2015 be a more sophisticated, more important, engine.
It's like a restaurant. You're stuck with the menu the restaurant has. Now, it's not that you can't necessarily get the kitchen to do a ham and cheese, but you have to do it in terms it understands (for example, you can order a burger that has ham and cheese, and order it without the beefburger, salad, etc), kind of like h[tt]p:, which runs on port 80. You can do it via the firewall, but it has to look like an HTTP request, which means running it on port 80. You can then say "Ohh, it's not really a burger, it's a ham and cheese sandwich" but as far as the kitchen's concerned, it's just one of their regular burgers. You might look at port 8090 as the ham - they're likely to have cheese burgers, but a ham, cheese, and beef burger? Not likely. So you can't have your ham and cheese because you haven't come up with a sandwich that really works within the framework you're given.
The only option is to leave the restaurant, and cook your own sandwich, but that's not always an option, especially if you actually work at the restaurant so can't leave until 5pm, but you're a waiter or you work at the bar or you greet people or wash up or something so you can't actually make the sandwich yourself (well, not in a unionized restaurant anyway. A union-free restaurant might allow it, but you don't want to upset the staff, and it's probably going against company policy.)
Port 8090 isn't supported by most corporate firewalls, so making all URLs point at it would just prevent Slashdot's working readers (the vast majority) from "eating their ham and cheese sandwich" - or, in other words, accessing the website. This would damage Slashdot long term as people would just stop reading it except for a few people at Universities and in Cybercafes, neither of which are appealing to Slashdot's advertisers.
I don't know if Earthlink, my erstwhile ISP (don't have DSL available in my new home), does it, but they do block outgoing port 25 anyway, so I might as well have been on the DUL.
Damn it!
I think virtual servers are becoming increasingly the only way to get certain things done that ought to be perfectly fine but aren't because ISPs put dumb things in their T&Cs, thanks to the antics of the extremists amongst the anti-spammers.
Great. Fantastic. *throws up*
(Sorry, see my most recent JE for the explanation of that last bit. Nothing to do with this. I think, anyway.)