Why can't people just download a codec pack of sorts if they don't already have a usable codec? We download Flash if we go to a site that needs it, so what makes downloading the codecs so impossible?
The point is to have it work out of the box, precisely to avoid the post-install "let this program that claims to be from Corporation X execute arbitrary code on my machine as root" clickthroughs. Why should you have to download codecs to view videos, when you don't have to to view images or use JavaScript or almost anything else? <video> should have been part of HTML long ago, just like <img>.
Also, Mozilla and Opera want to discourage the use of proprietary codecs as a matter of principle, for the health of the web. AFAIK, at least Firefox doesn't let you use anything other than Theora at all.
The funny thing is that for the last 50 years, I kept hearing that it was off by 10 years. Now, that the whole planet is involved, it will take 40 years. What do you think is the chance that it will not work?
My bet is that it will be some other means of doing it.
Whoosh. Grandparent is a reference to SimCity 2000, which a) let you build power plants like those of the OP, and b) unlocked fusion power in every game in the year 2050.
"Why the 'F does the page have to reload when I hit the back button"
Because a lot of web pages (including slashdot) are served with HTTP headers that tell the browser not to cache them, and the browser follows that directive. Yes, it's annoying. I use tabs for that reason too.
The HTTP specification explicitly says that caching headers are not to affect history traversal:
13.13 History Lists
User agents often have history mechanisms, such as "Back" buttons and history lists, which can be used to redisplay an entity retrieved earlier in a session.
History mechanisms and caches are different. In particular history mechanisms SHOULD NOT try to show a semantically transparent view of the current state of a resource. Rather, a history mechanism is meant to show exactly what the user saw at the time when the resource was retrieved.
By default, an expiration time does not apply to history mechanisms. If the entity is still in storage, a history mechanism SHOULD display it even if the entity has expired, unless the user has specifically configured the agent to refresh expired history documents.
This is not to be construed to prohibit the history mechanism from telling the user that a view might be stale.
Note: if history list mechanisms unnecessarily prevent users from
viewing stale resources, this will tend to force service authors
to avoid using HTTP expiration controls and cache controls when
they would otherwise like to. Service authors may consider it
important that users not be presented with error messages or
warning messages when they use navigation controls (such as BACK)
to view previously fetched resources. Even though sometimes such
resources ought not to cached, or ought to expire quickly, user
interface considerations may force service authors to resort to
other means of preventing caching (e.g. "once-only" URLs) in order
not to suffer the effects of improperly functioning history
mechanisms.
So no. The reason is because the browser doesn't want to cache all the necessary info, e.g., for memory usage reasons. Complain to your browser vendor if history navigation is too slow.
Me too. Best part is, you can get Glen's head back by running "google-chrome --glen". (At least on Linux.) I highly recommend you freak out all your fellow Chrome users by changing their browser shortcuts when they aren't looking!
For the tinfoil hat crowd, theres always SRWare Iron, which is Chrome, with updated webkit, with any google-related tracking removed. You lose site suggestion and auto-update tho, which personally i enjoy.
SRWare Iron has an older version of WebKit than Chrome developer builds. Their website says they use 532.0; about:version in my Chrome says 532.3. Chrome's developer builds track latest WebKit, AFAIK, or at least something like that. I doubt anything is going to be significantly more cutting-edge.
FWIW, I don't think there's anything in Chromium that transmits any info to Google and can't be easily disabled (e.g., change your search provider). On the other hand, I'd imagine a build by some tiny random software company would be less reliable than official Google builds, and getting support would be harder, so I see no reason to bother.
Who says SRWare is any more trustworthy than Google anyway? If you're paranoid, you're probably much better off compiling the (much more widely reviewed) official Chromium source yourself than relying on any company's official binaries.
It is pedantically true that there is no mention of wiretaps in the wording of the 4th Amendment - not surprising for a document written in 1789. But the first part of your "not factually wrong" conjunction (i.e. "Warrantless wiretapping infringes on our fundamental liberties") can only be true at all because of the 4th Amendment: there are other nations where it's not true at all.
Okay, we still aren't seeing eye-to-eye here. I'm not just pointing out that warrantless wiretapping is not specifically prohibited by the Fourth Amendment. I'm pointing out that it's not clear that it's even implicitly prohibited. All it says is "unreasonable search and seizure". It isn't at all clear what's "unreasonable". It's ambiguous; reasonable people can differ as to whether the Fourth Amendment is supposed to prohibit warrantless wiretapping. So saying without qualification that warrantless wiretapping is unconstitutional is not justifiable based on the text of the Constitution. It's a legitimate opinion, but not a fact either way.
As for your second point, I don't agree that fundamental liberties are determined by the Constitution or Bill of Rights. Those mandate certain liberties, but they're not an exhaustive list. The Ninth Amendment even says this: "The enumeration in the Constitution, of certain rights, shall not be construed to deny or disparage others retained by the people." The right to not be subject to warrantless wiretapping could be a fundamental liberty even if it's not in the Constitution anywhere.
You may request it - and I for one won't agree to your request. If we can't say what our rights are, and insist that only qualified professionals should be able to talk about them; if we have no sense that we can speak of that our rights have been violated, we will end up with none at all.
I never asked that anyone should refrain from saying what their rights are. I only said you shouldn't cite the Constitution when it actually doesn't support your case very well. Cite common sense, historical trends, natural law, the fundamental dignity of humankind, personal opinion, whatever you like, but the Constitution isn't clear either way. Something can be wrong and evil without being unconstitutional or illegal.
I am astonished that you think that a police officer is in a better position to make a judgement of your rights than you. You are uniquely qualified to stand up for your rights: you are willfully abdicating your sovereignty over your self to any authority figure with a certificate or a uniform and an officious manner.
That's not where liberty lies.
I never said I think warrantless wiretapping is okay or acceptable, or that anyone should think that warrantless wiretapping is okay or acceptable. I said that neither the literal text of the Constitution, nor legal precedent, clearly state that it's prohibited.
This sentence is not factually wrong: "Warrantless wiretapping infringes on our fundamental liberties, and anyone who tolerates it is legitimizing the trend toward a police state where the government has complete control over our lives."
This sentence is factually wrong: "Warrantless wiretapping is clearly against the Constitution."
It's written in Colonial English, so the meaning is less clear to us now. Here's how I interpret it:
1. Unreasonable searches and seizures shall not happen
2. Warrants shall not be issued unless there is probable cause and exactly what needs to be searched/seized is clearly outlined
The two statements, when taken together, imply that a warrant is indeed required in order to conduct a lawful search. Otherwise, why mention warrants in the same run-on sentence?
Warrants are obviously meant to be required sometimes, you're right, or else that clause is meaningless. But it doesn't say they're required all the time. If a police officer sees someone in the street openly carrying something that's illegal to possess, for instance, they don't need to get a warrant to confiscate it by any theory I've heard of.
So clearly warrants aren't required for all searches and seizures. Which ones they're required for is left open by the wording of the Constitution itself. It's clarified by court precedent, but again, in this particular case court precedent is inconclusive too (since no one has yet shown they have standing). So my conclusion still holds.
Okay, I know this post probably won't get modded up too high, given Slashdot libertarian groupthink, but: the Constitution doesn't say warrantless wiretapping is illegal. Let's take a look at the text of the Fourth Amendment:
The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.
This only says that searches must be "reasonable". It does not say "no search whatsoever can occur without a warrant". It mentions warrants, but doesn't say when exactly they're required. So, it's as legitimate an opinion as any to say that the government should have to get warrants for all domestic wiretapping, sure. But the Constitution doesn't say that.
Court precedent (based partly on the Constitution) might say that warrantless wiretapping is illegal, of course. Or it might not. There's no decision on the matter that hasn't been overruled, so it's an open question. I imagine, however, that most of the people calling warrantless wiretapping illegal and, e.g., advocating (+4 Insightful) assassination of the attorney general, are not lawyers and aren't really qualified to have an opinion on what the legal precedent implies.
So, might I request that we all make it clear what our personal opinions are, but don't claim support of the Constitution if it doesn't actually say anything clear on the issue?
Within the military community, you're absolutely correct, but politicians are rarely held to the same standard. If Joe Biden shot someone without provocation, Obama wouldn't face any problems but pressure to fire Biden and have him stand trial.
wait, so what happened with Bush & Cheney when Cheney did shoot someone?
. . . nothing, because it was obviously a hunting accident? Where, you know, the victim said it was an accident, and he was cleared of any wrongdoing after preliminary investigation? When you go around shooting things, it's possible for people to get hurt by mistake, and everyone involved understands and accepts this.
I don't know any details of the site's technical architecture beyond the obvious, but it's blazingly fast. My bet is that when you hit the site, you're pulling completed pages out of RAM on a customized and hardened Varnish, but that's just a guess. The HTTP headers identify the server technology as "White House."
I don't know where you came up with Varnish . . . there are lots of ways to get performance that's just as snappy. A CDN is a good start. And it's pretty easy to tell that that's exactly what's being used here:
They're using Akamai for most of their content, it seems. I get 35ms ping to www.whitehouse.gov from machines in New York, Denver, Holland, and Washington (the state). My Washington machine gets 2 ms ping, actually, so I'm guessing Akamai has a machine in the same data center. Varnish alone isn't going to get you anywhere close to that kind of performance – it can't beat light speed.
Does Ultimate come with a different kernel to Home? I was under the impression that the only differences between the versions were at the userland level. It's not like the older WinNT releases that actually did have slightly different kernels.
All Windows versions use the same kernel, but cheaper ones have some features disabled. For instance, in the very most expensive server version of Windows Vista, you can use 2 TB of RAM; in Home Basic it's something like 4 GB or 8 GB, IIRC. Home versions are also limited to only one or two physical CPUs, while the more expensive server editions get more. And so on.
But I don't think they deliberately reduce performance, beyond refusing to use extra hardware that you might have.
Possibly, but at least it can hot-swap monitors when you use it on a laptop. I've never been able to get Linux to pull that one off without a reboot.
I can get it to work reliably these days, as long as I don't mind typing xrandr commands. Complete with incomprehensible mode lines generated with cvt or gtf (they produce slightly different ones for some reason?) to set the resolution, after which I have to fiddle with the physical buttons on the monitor to get the image to be centered and the correct size. And I end up running a shell script on every boot to get my monitors actually working since I don't know how to get the settings remembered. Which is okay since I never reboot, except that sometimes I have to when I get erratic lock-ups that I suspect are kernel bugs of some kind. Which is annoying because my alarm is in cron and it doesn't go off, so I sometimes oversleep.
So, um. Not a ringing endorsement there, granted. But hey, still fun for me so far.
You don't really know what you're talking about here.
IE hasn't caught up to existing, published, finished standards- that's well before we even start talking about initial implementations of things from the in-progress HTML 5 standard. It's the worst browser in the bunch for CSS compatibility- with finished, published standards.
You don't really know what you're talking about here.
No relevant CSS standard beyond CSS1 (or CSS2 if you call that relevant) has reached Recommendation status (i.e., is "finished"). CSS2.1 (which IE8 has excellent support for) is a Candidate Recommendation. CSS3 is a collection of mostly unrelated standards, and the most advanced of those is also CR. Some of the most important, useful, widely-implemented CSS3 standards aren't even CR – Selectors is only Last Call, and AFAIK it has multiple full implementations.
HTML5, on the other hand, is a Working Draft as far as the W3C is concerned – just like most of CSS3. It's expected to reach Last Call by October. So there's really no serious difference here. (Leaving aside the fact that HTML5 is primarily published by the WHATWG, not the W3C, and the Working Draft version is an outdated and mostly useless snapshot that exists only for bureaucratic and political reasons, and which everyone ignores. That's beside the point.)
gee, and it really helps your case when the Microsoft rep on the HTML5 was one of the key people delaying the standard, isn't it?
Adrian Bateman gave a number of sound, reasonable objections to various aspects of the spec. Several echoed complaints that other commenters had made in the past. Some got immediate agreement from other vendors like Mozilla, resulting in speedy changes to the spec (like the removal of <bb>). There's ongoing discussion about other features, like <keygen>, which have resulted in productive changes.
It's fair to say that Microsoft is still only involved in the HTML5 community at arm's length. Adrian is the only MS rep to have commented so far, AFAIK, and his comments came years later than they might have. But MS is contributing constructively and is looking to implement HTML5. Quoting Mozilla developer Robert O'Callahan on his blog, referring to a talk by IE8 Technical Evangelist Giorgio Sardo,
His basic message was "we care about standards (including HTML5), we're doing it, our release cycles are slow because we have the most users but we will get there." What I thought was very interesting was that he made no attempt to distance themselves from HTML5 or even say that they'd be selective about which HTML5 features made sense to implement. His message tacitly assumed that HTML5 is simply something they will do.
Nothing that MS has done will slow down the standardization process. Major features like <video> are already set in stone in most respects, since they have multiple interoperable implementations. If MS has come late to the party, it just means they get less say. HTML5 is still set for Last Call in October last I heard. Its progress to further levels of the W3C echelon will be slowed based on the number of formal objections made by various parties; there's no indication yet that Microsoft will slow it down particularly much. As far as I know, Microsoft has not yet filed any formal objections.
The W3C standardization process is hardly relevant, though. Nobody cares what the W3C says. <video> works in the latest versions of all browsers except IE, for instance. It doesn't need to be part of a "finished" specification to be used in practice.
What possible use would having the boss of Verisign in their back pocket be?
Verisign fulfills a 'trust provider' function by signing people's website certificates. The only use for that would be to have a clean certificate for, say, a typosquatting site.
If you had control of a CA's key — and I think it should be treated as obvious that the NSA could get one — you could write fake certificates. So say someone goes to https://evil.com/ and the government wants to spy. They can order the ISP to secretly log all the traffic, but it's worthless: the traffic is encrypted. You could provide a fake certificate, but then scary warnings would go up about domain name mismatches or whatever. But if you have a recognized CA's key, you can make your own legitimate-looking certificate. Then just route all the traffic through your proxy, which uses the forged certificate's private key to act as an HTTPS server to the suspected miscreant, and the real certificate's public key to act as an HTTPS client to the real site.
Basically, SSL relies on the trustworthiness of certificate authorities. If you can forge a certificate, it's mostly worthless. All you need is one CA's key, and there are zillions of the buggers based in the US, so I really doubt the US government would have a hard time getting one if it cared.
Or some way to break the encryption, eg. they've got the boss of Verisign in their back pocket.
Um, VeriSign is a US company. All the governments needs is a warrant (and possibly not even that) to get them to hand over the keys. This is hardly a revelation. If you don't trust the US government, you can't trust any company based in the US either, because their executives are not likely to want to get hauled off to jail for not cooperating with law enforcement agencies.
Besides, SSL is based on stuff like RSA, AES, and so on. The NSA has approved those for use with confidential government data. It would have to be pretty stupid to do that if it knew that there was a break. If they can find it, so can the Chinese or whatever.
I'm all for not trusting the government too much, but you have to look at the practicalities here. There's not much room for sinister interpretations of the fact that the government allows distribution of SSL implementations.
Having had a few EXT3 filesystems go tits up because they've been quietly borking themselves on a 24/7/365 server being able to do a weekly "zpool scrub" in a 4TB array without the downtime is a beautiful thing. Kernel CIFS with proper ACLs and integration with ZFS snapshots is pretty great as well. When btrfs is released and gets a few miles on it I may switch back. But for now my file server stays OpenSolaris.
On the other hand, I'm somewhat involved in two different sites that use(d) ZFS. One, the Wikimedia toolserver, switched from ZFS to vxfs for/home after several serious issues where ZFS died and caused downtime (although no data loss). (more info)
The other site, Wikipedia, uses ZFS for image storage. Some time ago, it had to disable uploads for a while to move data off to other machines, because apparently ZFS (or some configuration of ZFS) slows down to a crawl when it gets too full. It's still using ZFS for the time being, AFAIK, just making sure it doesn't get too full.
So, I'm sure different people have different experiences, but just throwing my limited personal experience out there.
But among the worst things you can do is to hard-code low-level hardware handling and scatter that throughout the source code. Or some important code may be overly dependent on CPU specific feature (e.g. task gate for intel x86) which makes it essentially non-portable. I've never seen the source code of Windows, but I suspect this is the case with them. According to some Windows NT Internals book I read many years ago, it started out well-structured, with a nice hardware abstraction layer and all that. But since Windows dropped Alpha processor support, I think the abstraction started to suffer bit-rot and made things much worse than if they had no abstraction at all.
Windows still supports three architectures: x86, x64, and IA64. IA64 is probably different enough from the others to keep them at least somewhat honest, no? Windows Internals, Fifth Edition definitely says there's still a wall between the HAL and the rest of the OS.
So I imagine Windows wouldn't be too hard to port to ARM, given Microsoft's resources. But without any of the applications, it would be at an incredible disadvantage.
It's not so much that the driver doesn't know how, I think it's more of a protection decision. By default if a drive is marked "dirty" it will not auto mount.
Reinserting into a winbox is one method, however you can run into this when trying to mount a standard HDD from a windows machine that no longer boots. It's useful to know that you can also do a force mount the command is something like:
mount -t ntfs-3g/dev/sdb1/media/disk -o defaults,force,umask=0
Yeah, but does that replay the journal (thus ensuring the metadata is in a consistent state) or just throw out the journal, so you get the old FAT/ext2 "I hope it's consistent" behavior? Although now that I think about it, that can't really be called much of a disadvantage if the alternative is FAT . . .
Yes, that's why the people who care about that driver do the necessary work to get it merged in, instead of just throwing some code at the maintainers and expecting them to do it all.
Except when they don't, like the authors of the half-dozen other drivers GKH mentioned he'll cut for 2.6.32 if he doesn't see some activity. Microsoft is fairly typical here for a new Linux kernel contributor. Long-established companies like Intel that commit tons of high-quality code to support their hardware are the exception, not the rule. Heck, Google was one of the ones Greg panned in his status report, for being uninterested in getting Android drivers into shape, and they're a quite significant kernel contributor.
The ones turning this into an anti-Microsoft flamefest are really not being reasonable. No, Microsoft didn't do a lot more than it had to do, but it's hardly special in that regard. I certainly hate MS just as much as the next Slashdotter, but let's keep our heads on straight. We should be flaming Microsoft for its genuinely reprehensible behavior, like its anticompetitive practices, not for not being as enthusiastic as we'd like in cleaning up a driver for one of its largest competitors. Looking only for the worst in everything MS does just paints us as deranged extremists. (Which, sadly, some of us are.)
The Linux NTFS drivers are working well now. That's what I use on my shared partition.
Except if the portable drive is pulled out of the machine without being properly unmounted. Then the filesystem is unclean, and the Linux NTFS driver doesn't know how to replay the journal. So the thing becomes unusable until you stick it into a Windows machine and then remove it properly. Of course, you shouldn't be pulling it out without unmounting . . . that can cause serious data loss on some cheap USB drives, apparently, regardless of journaling.
Slashdot Mirror
Why can't people just download a codec pack of sorts if they don't already have a usable codec? We download Flash if we go to a site that needs it, so what makes downloading the codecs so impossible?
The point is to have it work out of the box, precisely to avoid the post-install "let this program that claims to be from Corporation X execute arbitrary code on my machine as root" clickthroughs. Why should you have to download codecs to view videos, when you don't have to to view images or use JavaScript or almost anything else? <video> should have been part of HTML long ago, just like <img>.
Also, Mozilla and Opera want to discourage the use of proprietary codecs as a matter of principle, for the health of the web. AFAIK, at least Firefox doesn't let you use anything other than Theora at all.
The funny thing is that for the last 50 years, I kept hearing that it was off by 10 years. Now, that the whole planet is involved, it will take 40 years. What do you think is the chance that it will not work?
My bet is that it will be some other means of doing it.
Whoosh. Grandparent is a reference to SimCity 2000, which a) let you build power plants like those of the OP, and b) unlocked fusion power in every game in the year 2050.
On the other hand, if we RTFS:
The latest versions of Firefox, Chrome, and Safari are supported
Note that IE is not on the list. Make an educated guess about the implications for the penetration of the video tag.
Microsoft has committed to supporting <video> and <audio>. When, heaven only knows, but <video> will win eventually. Hopefully sooner rather than later.
"Why the 'F does the page have to reload when I hit the back button"
Because a lot of web pages (including slashdot) are served with HTTP headers that tell the browser not to cache them, and the browser follows that directive. Yes, it's annoying. I use tabs for that reason too.
The HTTP specification explicitly says that caching headers are not to affect history traversal:
So no. The reason is because the browser doesn't want to cache all the necessary info, e.g., for memory usage reasons. Complain to your browser vendor if history navigation is too slow.
So run the dev channel. It has extensions today. Yes, including ad blockers. Dev channel is actually perfectly usable if you don't mind the occasional disembodied head taking the place of a button. Dev channel Chrome has been my primary browser for over a year now.
Me too. Best part is, you can get Glen's head back by running "google-chrome --glen". (At least on Linux.) I highly recommend you freak out all your fellow Chrome users by changing their browser shortcuts when they aren't looking!
For the tinfoil hat crowd, theres always SRWare Iron, which is Chrome, with updated webkit, with any google-related tracking removed. You lose site suggestion and auto-update tho, which personally i enjoy.
SRWare Iron has an older version of WebKit than Chrome developer builds. Their website says they use 532.0; about:version in my Chrome says 532.3. Chrome's developer builds track latest WebKit, AFAIK, or at least something like that. I doubt anything is going to be significantly more cutting-edge.
FWIW, I don't think there's anything in Chromium that transmits any info to Google and can't be easily disabled (e.g., change your search provider). On the other hand, I'd imagine a build by some tiny random software company would be less reliable than official Google builds, and getting support would be harder, so I see no reason to bother.
Who says SRWare is any more trustworthy than Google anyway? If you're paranoid, you're probably much better off compiling the (much more widely reviewed) official Chromium source yourself than relying on any company's official binaries.
It is pedantically true that there is no mention of wiretaps in the wording of the 4th Amendment - not surprising for a document written in 1789. But the first part of your "not factually wrong" conjunction (i.e. "Warrantless wiretapping infringes on our fundamental liberties") can only be true at all because of the 4th Amendment: there are other nations where it's not true at all.
Okay, we still aren't seeing eye-to-eye here. I'm not just pointing out that warrantless wiretapping is not specifically prohibited by the Fourth Amendment. I'm pointing out that it's not clear that it's even implicitly prohibited. All it says is "unreasonable search and seizure". It isn't at all clear what's "unreasonable". It's ambiguous; reasonable people can differ as to whether the Fourth Amendment is supposed to prohibit warrantless wiretapping. So saying without qualification that warrantless wiretapping is unconstitutional is not justifiable based on the text of the Constitution. It's a legitimate opinion, but not a fact either way.
As for your second point, I don't agree that fundamental liberties are determined by the Constitution or Bill of Rights. Those mandate certain liberties, but they're not an exhaustive list. The Ninth Amendment even says this: "The enumeration in the Constitution, of certain rights, shall not be construed to deny or disparage others retained by the people." The right to not be subject to warrantless wiretapping could be a fundamental liberty even if it's not in the Constitution anywhere.
You may request it - and I for one won't agree to your request. If we can't say what our rights are, and insist that only qualified professionals should be able to talk about them; if we have no sense that we can speak of that our rights have been violated, we will end up with none at all.
I never asked that anyone should refrain from saying what their rights are. I only said you shouldn't cite the Constitution when it actually doesn't support your case very well. Cite common sense, historical trends, natural law, the fundamental dignity of humankind, personal opinion, whatever you like, but the Constitution isn't clear either way. Something can be wrong and evil without being unconstitutional or illegal.
There you go again...
I am astonished that you think that a police officer is in a better position to make a judgement of your rights than you. You are uniquely qualified to stand up for your rights: you are willfully abdicating your sovereignty over your self to any authority figure with a certificate or a uniform and an officious manner.
That's not where liberty lies.
I never said I think warrantless wiretapping is okay or acceptable, or that anyone should think that warrantless wiretapping is okay or acceptable. I said that neither the literal text of the Constitution, nor legal precedent, clearly state that it's prohibited.
This sentence is not factually wrong: "Warrantless wiretapping infringes on our fundamental liberties, and anyone who tolerates it is legitimizing the trend toward a police state where the government has complete control over our lives."
This sentence is factually wrong: "Warrantless wiretapping is clearly against the Constitution."
Do you understand the point I'm making now?
It's written in Colonial English, so the meaning is less clear to us now. Here's how I interpret it:
1. Unreasonable searches and seizures shall not happen 2. Warrants shall not be issued unless there is probable cause and exactly what needs to be searched/seized is clearly outlined
The two statements, when taken together, imply that a warrant is indeed required in order to conduct a lawful search. Otherwise, why mention warrants in the same run-on sentence?
Warrants are obviously meant to be required sometimes, you're right, or else that clause is meaningless. But it doesn't say they're required all the time. If a police officer sees someone in the street openly carrying something that's illegal to possess, for instance, they don't need to get a warrant to confiscate it by any theory I've heard of.
So clearly warrants aren't required for all searches and seizures. Which ones they're required for is left open by the wording of the Constitution itself. It's clarified by court precedent, but again, in this particular case court precedent is inconclusive too (since no one has yet shown they have standing). So my conclusion still holds.
Okay, I know this post probably won't get modded up too high, given Slashdot libertarian groupthink, but: the Constitution doesn't say warrantless wiretapping is illegal. Let's take a look at the text of the Fourth Amendment:
This only says that searches must be "reasonable". It does not say "no search whatsoever can occur without a warrant". It mentions warrants, but doesn't say when exactly they're required. So, it's as legitimate an opinion as any to say that the government should have to get warrants for all domestic wiretapping, sure. But the Constitution doesn't say that.
Court precedent (based partly on the Constitution) might say that warrantless wiretapping is illegal, of course. Or it might not. There's no decision on the matter that hasn't been overruled, so it's an open question. I imagine, however, that most of the people calling warrantless wiretapping illegal and, e.g., advocating (+4 Insightful) assassination of the attorney general, are not lawyers and aren't really qualified to have an opinion on what the legal precedent implies.
So, might I request that we all make it clear what our personal opinions are, but don't claim support of the Constitution if it doesn't actually say anything clear on the issue?
Within the military community, you're absolutely correct, but politicians are rarely held to the same standard. If Joe Biden shot someone without provocation, Obama wouldn't face any problems but pressure to fire Biden and have him stand trial.
wait, so what happened with Bush & Cheney when Cheney did shoot someone?
. . . nothing, because it was obviously a hunting accident? Where, you know, the victim said it was an accident, and he was cleared of any wrongdoing after preliminary investigation? When you go around shooting things, it's possible for people to get hurt by mistake, and everyone involved understands and accepts this.
I don't know any details of the site's technical architecture beyond the obvious, but it's blazingly fast. My bet is that when you hit the site, you're pulling completed pages out of RAM on a customized and hardened Varnish, but that's just a guess. The HTTP headers identify the server technology as "White House."
I don't know where you came up with Varnish . . . there are lots of ways to get performance that's just as snappy. A CDN is a good start. And it's pretty easy to tell that that's exactly what's being used here:
$ dig +short www.whitehouse.gov
www.whitehouse.gov.edgekey.net.
e2561.g.akamaiedge.net.
96.16.18.135
They're using Akamai for most of their content, it seems. I get 35ms ping to www.whitehouse.gov from machines in New York, Denver, Holland, and Washington (the state). My Washington machine gets 2 ms ping, actually, so I'm guessing Akamai has a machine in the same data center. Varnish alone isn't going to get you anywhere close to that kind of performance – it can't beat light speed.
Does Ultimate come with a different kernel to Home? I was under the impression that the only differences between the versions were at the userland level. It's not like the older WinNT releases that actually did have slightly different kernels.
All Windows versions use the same kernel, but cheaper ones have some features disabled. For instance, in the very most expensive server version of Windows Vista, you can use 2 TB of RAM; in Home Basic it's something like 4 GB or 8 GB, IIRC. Home versions are also limited to only one or two physical CPUs, while the more expensive server editions get more. And so on.
But I don't think they deliberately reduce performance, beyond refusing to use extra hardware that you might have.
This is a new meaning of the word "reliably" with which I was not previously familiar.
How so? Reliability is orthogonal to ease of use. I do get it working every time.
Possibly, but at least it can hot-swap monitors when you use it on a laptop. I've never been able to get Linux to pull that one off without a reboot.
I can get it to work reliably these days, as long as I don't mind typing xrandr commands. Complete with incomprehensible mode lines generated with cvt or gtf (they produce slightly different ones for some reason?) to set the resolution, after which I have to fiddle with the physical buttons on the monitor to get the image to be centered and the correct size. And I end up running a shell script on every boot to get my monitors actually working since I don't know how to get the settings remembered. Which is okay since I never reboot, except that sometimes I have to when I get erratic lock-ups that I suspect are kernel bugs of some kind. Which is annoying because my alarm is in cron and it doesn't go off, so I sometimes oversleep.
So, um. Not a ringing endorsement there, granted. But hey, still fun for me so far.
You don't really know what you're talking about here.
IE hasn't caught up to existing, published, finished standards- that's well before we even start talking about initial implementations of things from the in-progress HTML 5 standard. It's the worst browser in the bunch for CSS compatibility- with finished, published standards.
You don't really know what you're talking about here.
No relevant CSS standard beyond CSS1 (or CSS2 if you call that relevant) has reached Recommendation status (i.e., is "finished"). CSS2.1 (which IE8 has excellent support for) is a Candidate Recommendation. CSS3 is a collection of mostly unrelated standards, and the most advanced of those is also CR. Some of the most important, useful, widely-implemented CSS3 standards aren't even CR – Selectors is only Last Call, and AFAIK it has multiple full implementations.
HTML5, on the other hand, is a Working Draft as far as the W3C is concerned – just like most of CSS3. It's expected to reach Last Call by October. So there's really no serious difference here. (Leaving aside the fact that HTML5 is primarily published by the WHATWG, not the W3C, and the Working Draft version is an outdated and mostly useless snapshot that exists only for bureaucratic and political reasons, and which everyone ignores. That's beside the point.)
gee, and it really helps your case when the Microsoft rep on the HTML5 was one of the key people delaying the standard, isn't it?
Adrian Bateman gave a number of sound, reasonable objections to various aspects of the spec. Several echoed complaints that other commenters had made in the past. Some got immediate agreement from other vendors like Mozilla, resulting in speedy changes to the spec (like the removal of <bb>). There's ongoing discussion about other features, like <keygen>, which have resulted in productive changes.
It's fair to say that Microsoft is still only involved in the HTML5 community at arm's length. Adrian is the only MS rep to have commented so far, AFAIK, and his comments came years later than they might have. But MS is contributing constructively and is looking to implement HTML5. Quoting Mozilla developer Robert O'Callahan on his blog, referring to a talk by IE8 Technical Evangelist Giorgio Sardo,
Nothing that MS has done will slow down the standardization process. Major features like <video> are already set in stone in most respects, since they have multiple interoperable implementations. If MS has come late to the party, it just means they get less say. HTML5 is still set for Last Call in October last I heard. Its progress to further levels of the W3C echelon will be slowed based on the number of formal objections made by various parties; there's no indication yet that Microsoft will slow it down particularly much. As far as I know, Microsoft has not yet filed any formal objections.
The W3C standardization process is hardly relevant, though. Nobody cares what the W3C says. <video> works in the latest versions of all browsers except IE, for instance. It doesn't need to be part of a "finished" specification to be used in practice.
What possible use would having the boss of Verisign in their back pocket be?
Verisign fulfills a 'trust provider' function by signing people's website certificates. The only use for that would be to have a clean certificate for, say, a typosquatting site.
If you had control of a CA's key — and I think it should be treated as obvious that the NSA could get one — you could write fake certificates. So say someone goes to https://evil.com/ and the government wants to spy. They can order the ISP to secretly log all the traffic, but it's worthless: the traffic is encrypted. You could provide a fake certificate, but then scary warnings would go up about domain name mismatches or whatever. But if you have a recognized CA's key, you can make your own legitimate-looking certificate. Then just route all the traffic through your proxy, which uses the forged certificate's private key to act as an HTTPS server to the suspected miscreant, and the real certificate's public key to act as an HTTPS client to the real site.
Basically, SSL relies on the trustworthiness of certificate authorities. If you can forge a certificate, it's mostly worthless. All you need is one CA's key, and there are zillions of the buggers based in the US, so I really doubt the US government would have a hard time getting one if it cared.
Or some way to break the encryption, eg. they've got the boss of Verisign in their back pocket.
Um, VeriSign is a US company. All the governments needs is a warrant (and possibly not even that) to get them to hand over the keys. This is hardly a revelation. If you don't trust the US government, you can't trust any company based in the US either, because their executives are not likely to want to get hauled off to jail for not cooperating with law enforcement agencies.
Besides, SSL is based on stuff like RSA, AES, and so on. The NSA has approved those for use with confidential government data. It would have to be pretty stupid to do that if it knew that there was a break. If they can find it, so can the Chinese or whatever.
I'm all for not trusting the government too much, but you have to look at the practicalities here. There's not much room for sinister interpretations of the fact that the government allows distribution of SSL implementations.
Having had a few EXT3 filesystems go tits up because they've been quietly borking themselves on a 24/7/365 server being able to do a weekly "zpool scrub" in a 4TB array without the downtime is a beautiful thing. Kernel CIFS with proper ACLs and integration with ZFS snapshots is pretty great as well. When btrfs is released and gets a few miles on it I may switch back. But for now my file server stays OpenSolaris.
On the other hand, I'm somewhat involved in two different sites that use(d) ZFS. One, the Wikimedia toolserver, switched from ZFS to vxfs for /home after several serious issues where ZFS died and caused downtime (although no data loss). (more info)
The other site, Wikipedia, uses ZFS for image storage. Some time ago, it had to disable uploads for a while to move data off to other machines, because apparently ZFS (or some configuration of ZFS) slows down to a crawl when it gets too full. It's still using ZFS for the time being, AFAIK, just making sure it doesn't get too full.
So, I'm sure different people have different experiences, but just throwing my limited personal experience out there.
But among the worst things you can do is to hard-code low-level hardware handling and scatter that throughout the source code. Or some important code may be overly dependent on CPU specific feature (e.g. task gate for intel x86) which makes it essentially non-portable. I've never seen the source code of Windows, but I suspect this is the case with them. According to some Windows NT Internals book I read many years ago, it started out well-structured, with a nice hardware abstraction layer and all that. But since Windows dropped Alpha processor support, I think the abstraction started to suffer bit-rot and made things much worse than if they had no abstraction at all.
Windows still supports three architectures: x86, x64, and IA64. IA64 is probably different enough from the others to keep them at least somewhat honest, no? Windows Internals, Fifth Edition definitely says there's still a wall between the HAL and the rest of the OS.
So I imagine Windows wouldn't be too hard to port to ARM, given Microsoft's resources. But without any of the applications, it would be at an incredible disadvantage.
It's not so much that the driver doesn't know how, I think it's more of a protection decision. By default if a drive is marked "dirty" it will not auto mount. Reinserting into a winbox is one method, however you can run into this when trying to mount a standard HDD from a windows machine that no longer boots. It's useful to know that you can also do a force mount the command is something like: mount -t ntfs-3g /dev/sdb1 /media/disk -o defaults,force,umask=0
Yeah, but does that replay the journal (thus ensuring the metadata is in a consistent state) or just throw out the journal, so you get the old FAT/ext2 "I hope it's consistent" behavior? Although now that I think about it, that can't really be called much of a disadvantage if the alternative is FAT . . .
Yes, that's why the people who care about that driver do the necessary work to get it merged in, instead of just throwing some code at the maintainers and expecting them to do it all.
Except when they don't, like the authors of the half-dozen other drivers GKH mentioned he'll cut for 2.6.32 if he doesn't see some activity. Microsoft is fairly typical here for a new Linux kernel contributor. Long-established companies like Intel that commit tons of high-quality code to support their hardware are the exception, not the rule. Heck, Google was one of the ones Greg panned in his status report, for being uninterested in getting Android drivers into shape, and they're a quite significant kernel contributor.
The ones turning this into an anti-Microsoft flamefest are really not being reasonable. No, Microsoft didn't do a lot more than it had to do, but it's hardly special in that regard. I certainly hate MS just as much as the next Slashdotter, but let's keep our heads on straight. We should be flaming Microsoft for its genuinely reprehensible behavior, like its anticompetitive practices, not for not being as enthusiastic as we'd like in cleaning up a driver for one of its largest competitors. Looking only for the worst in everything MS does just paints us as deranged extremists. (Which, sadly, some of us are.)
The Linux NTFS drivers are working well now. That's what I use on my shared partition.
Except if the portable drive is pulled out of the machine without being properly unmounted. Then the filesystem is unclean, and the Linux NTFS driver doesn't know how to replay the journal. So the thing becomes unusable until you stick it into a Windows machine and then remove it properly. Of course, you shouldn't be pulling it out without unmounting . . . that can cause serious data loss on some cheap USB drives, apparently, regardless of journaling.