When MS announced Hailstorm, much was made of potential future technologies that would have computers recognize you (with a camera), and keep an eye on what you are doing. This, say MS, means that the computer can decide how and when it should deliver pieces of information based upon its "understanding" of that information's importance. Now, if Hailstorm uses a centralized system - and it will have to for the various "medical records automatically sent to ambulances" and similar technologies - doesn't that mean that MS have access to a giant face database - similar to the technology employed here? What if the DOJ request (probably through the courts) access to this database? Does that mean that they can track anyone, and Windows PCs will inform the police of your location if a warrant is out for you?
While none of the MS employees I know want to create an Orwellian world, this would seem to be a very efficient way to go about it.
After reading this thread, and wondering what all the fuss was about, I went ahead and downloaded the current Internet Explorer 6 beta. You need to be running Win2k or the XP beta to use this release.
Anyway, the first thing to note is that by default Smart Tags are disabled - but they are really easy to enable. There is also an advanced setting to "always display smart tags" - giving users the option to override meta tags found on individual sites. It defaults to off, and as long as it stays that way I'm okay with it.
After switching on Smart Tags, the only smart tag I could find in this thread was the word Microsoft - and smart tags don't look like links. They are underlined with a dotted black and purple link. When you click on them, they don't take you anywhere - they pop up a tooltip with a little "information" icon in it and a pull-down list of pieces of information you might want to see. Very innocuous, very obviously not a link supplied by the document's author - and very unobtrusive.
The current IE6 beta also has some nice privacy features, particularly the option to block third-party cookies. Opera and Mozilla have had this for a while, but it's nice that IE has caught up!
I have to say, I think that this is a storm in a teacup (unless you are in the habit of making your links not act like links, and look different...)
Good point. On the other hand, have you ever seen someone try and get basic office work without MS Office skills? You can't even get work at a local temp agency here without at least MS Word or Powerpoint capability! So while I agree with you that things change, it is definitely worth making sure that kids who are about to graduate (and not all - or even very many in some areas - will go on to college) have skills that will be immediately useful in the workforce.
While in general I agree, there is one important reason to keep MS applications in public schools: you produce users familiar with MS office. Office remains the dominant business application, and if public education is to prepare students for future employment then it should provide a solid grounding in the applications that employers use.
and it doesn't matter that MS Office is bloated almost to the point of unusability. The only thing that matters is the perennial question : "Does it run Word?" and until this question can be answered "Yes!" (which presumably means a radically different Microsoft to the one we have now), the gloom will persist.
The ironic thing about MS Office's bloat is that nobody I know/consult for uses more than 80% of any Office program's featureset... but they all use a different 80% and become very upset if their personal subset isn't available.
The other thing lacking in all of the MS Office alternatives I've tried is that of a good VBA clone. Like it or loathe it, Visual Basic for Applications makes it really easy to knock together a customized office solution. (Don't try the "nobody does this" argument, because it simply isn't true - I make a good deal of my living producing customized office applets for clients!).
I've always had mixed feelings about OS/2, kind of a love-hate relationship. I remember first trying it (on a 386) at home, and being impressed with how snappy it seemed (relative to Windows 3.x) - and its relative stability. Running Windows apps was impressive - when it worked (lots of my windows apps *almost* worked perfectly, very few were perfect). I also remember the user interface making sense as a whole - and being incredibly frustrating because its default settings (such as which mouse button you use to drag) were preconfigured in a way that made NO sense. It also really annoyed me that I had to tune a silly number of internal variables for anything approaching decent performance - and it seemed that everyone I spoke to also had to tune these same variables. Almost a great product, but lacking polish.
Then there was the issue of applications... a few bits of shareware, some mediocre stuff written with gpp, some *NIX ports - and a lot of really expensive business apps I didn't need. Its emulation was decent, but Doom without sound? Never!
Anyway, I switched to NT3.51 and never looked back. Nowadays its Win2k Professional and FreeBSD for me!
Although this is getting off-topic, can anyone explain to me why a suitcase nuke scenario isn't more likely than an obvious ballistic attack, even from a rogue state?
In all honesty, a "suitcase nuke" (still kind of a big suitcase, but the word fits) attack is much more likely than a missile attack for a number of reasons:
Smaller blast; you tend to hit what you want to hit, not an indiscriminate area.
Less traceable; if you want to perform an action covertly, this is the way to do it. This is especially true as more and more states gain nuclear technologies. It's not inconceivable for a state-sponsored terror group to use a suitcase nuke, for example - and it would be really hard to decide which nation to retalliate against! Conversely, the iniator of a missile attack is readily identifiable.
More likely to hit what you want to hit. This doesn't apply to the US and Russia, whose missiles are nice and accurate - but for a rogue state with missiles of "hitting the right city is nice" accuracy (Scud-A, Frog, etc.) levels a suitcase bomb presents certain advantages.
Easier to get to target; no obvious warning signs, and nuclear packages are hard to spot.
On the other hand, in a large-scale (relative to the parties; the Gulf war was a HUGE war for Iraq, not so big for the allies) war scenario missiles have the advantage that you don't need to get your people into enemy territory. There is also the "missiles make you a real power" theory floating around, to which many smaller states seem to subscribe. Certainly, if your objective is denial (ie. "shoot the carrier battlegroup before it gets close enough to bomb you" and "make sure that nobody wants to visit airbases in the area" missiles with NBC [Nuclear-Biological-Chemical] warheads can still make a certain amount of sense.
That's a long winded way of saying that you are right - suitcase nukes are very probable.:-)
If the effects of damage through blast overpressure scaled linearly, then you'd be correct. Unfortunately, it doesn't quite work that way. (Again, refer to the Effects of Nuclear Weapons, or the large body of translated former-Sovet texts available in university libraries that cover Strategic Studies).
The effects of detonating a thermonuclear weapon (as opposed to your run of the mill nuclear weapon) high-above DC would be pretty devastating - depending upon the exact rating of the weapon, weather, topography, etc. you could reasonably expect much of DC to be destroyed. A high-altitude airburst would minimize residual radiation effects, maximize blast damage (greater line of sight area for blast overpressure damage - the primary kill factor for a thermonuclear weapon) - and you'd make quite a mess.
On the other hand, ground-bursting a thermonuclear weapon on a ship in harbour is going to be a LOT less destructive relative to a well-placed airburst. You'll have more secondary radiation issues to deal with (the fireball really HAS to reach ground level, by definition) but blast damage is likely to be relatively localized by the quantity of solid matter it has to pass through/over to knock more stuff over. Remember, blast reflects/wraps on solid objects that don't break - so while the first few buildings would crumble against pressure, the buildings behind them would receive MUCH weaker effects. Detonating a full-scale thermonuclear weapon on a ship really is a bit of a waste because of this!
All of this ignores the strategic lunacy of targeting cities in general, but that's another can of worms.:-)
A nuclear blast from a weapon inside a ship at port will destroy not only the port it's at but the entire city.
Depends upon the nuclear weapon.:-) Seriously, a man-portable nuclear weapon could detonate and probably not inflict massive damage beyond the first row of buildings/hard objects surrounding the port. Fallout would be an issue, although weapons designers can minimize this to an extent - certainly not as bad as the apocalyptic impression spread by phrases such as that one. Remember that people in shelter's around Hiroshima ground-zero were relatively unscathed - and train service for the city was functional again in a few hours. Nuclear != Armageddon, just a particular type of weapon.
Highly recommended reading:
S. Glasstone & P. Dolan, The Effects of Nuclear Weapons, GPO.
Also, US Navy sonar will most certainly detect something appraching underwater at over 250mph (and maybe they could do something about it, or maybe not...) but if a nuke is hidden on a ship then the Navy very well may allow them to dock.
This is a big issue, and one that the Pentagon is familiar with (although navy brass tend to reject the notion that nuclear weapons will ever be used). In a recent wargame in which I participated, red team was able to do a ridiculous amount of infrastructure damage to the US with this very tactic. I'm told that there are systems in R&D for over-the-horizon detection of nuclear weapons... although I have no idea how that would work. For all I know, it's military vapourware.
It should be possible to detect inbound weapons travelling at >250mpg, the question is what to do about it once you've detected it (the exact same problem as with ballistic missiles). Shooting it down requires the existence of a pretty impressive infrastructure designed to stop that particular type of attack - and as you say, probably won't help against the shipping boat packed with explosives.
If your website is a personal artistic expression, then great - specify whatever browser you want. You'll lose users, but for that may not be the point of the site.
On the other hand, an important lesson of the dot-com era seems to be that commercially oriented websites should encourage people to buy their products from their website. Sure, its easy for lazy webdesigners to "require" Internet Explorer, X version of Java/ECMA script, or similar - but that loses customers. If your page isn't readable with Lynx, you've lost a great many blind and visually impaired users, not to mention users who browse sans-graphics in the hope of speeding up their connections. If your page relies upon the existence of V4 browsers, you've lost a whole bunch of users on cellphones, PDAs and some primitive appliances. Even frames may be pushing it a bit in terms of global compatibility!
What this group are forgetting is that HTML is designed to convey content and not a specific look and feel. Ideally, a webpage should be useable on any device, irrespective of platform.
That said, it can be really hard to stop web developers from going nuts with Flash, sometimes.:-(
Great, now they'll try and sue Mr. Torvalds.;-p
Seriously, I share the sentiments of your letter. Is there a foundation dedicated to overturning ludicrous patents, yet? Sounds like a worthy cause to me.
Much of FreeBSD 5's kernel development is focussed on improving SMP performance, particularly with respect to the TCP/IP stack. I don't know how long it will take for it to be declared stable, but I've had no real problems with the version that shipped in the 6-CD BSD box I found at Staples.
If you are creating a website aimed at a large audience, you should consider offering more than one form of stream - and at least one of the offerings should be as mainstream as possible. The reason for this is simple: if you want people to use the service, you want to use a format they already support. (That said, if your site were targeting non-Windows users, this rule could work the other way around)
I recently had to evaluate some streaming video (over LAN) solutions, and we ended up with the Microsoft solution because its inexpensive (effectively bundled with 2k Server) and offers pretty good quality. A major factor in deciding, though, was platform: 99% of the likely userbase used Windows, and weren't techies - meaning that they didn't want to bother with difficult plugin installations!
So, in your case which you describe, it would be useful. How else would you be able to find the other 9 people in the city without having their DNA and name catalogued on file?
True, it does appear useful - but its only really useful if you can catalogue the whole world, or at least that portion of it that's likely to have been in a certain locale at a given time. For example, if you know that there are probably 9 other suspects living within a town (remember, the 1/10,000 figure is a statistical average - don't rely on it) how do you know that the suspect wasn't from a different area? If you have a system of voluntary DNA evidence donation, all you do is implicate the law abiding types who volunteer the evidence. If you build a database of the whole country, your sample becomes so large as to be relatively meaningless:
Conceivably, if you could check DNA evidence against a database of everyone in the UK, you could come up with a list of a few hundred thousand suspects for every crime. On the other hand, regular police work generally provides a much narrower suspect list!
DNA evidence is useful secondary evidence, but it should not be relied upon as primary (or worse, sole convictor) evidence.
I do not know the UK law system very well, but does the system have a fifth amendment type protection against self incrimination?
No, there is not. In fact, there have been several cases convicted purely on the basis of confessions that were later withdrawn! (Fortunately, that practice has since been outlawed in the aftermath of the Birmingham 6 and Guildford 4 inquiries).
When I was studying criminal law (in the UK), a section of one of our courses was dedicated to DNA evidence. Our professor cited several tracts showing that DNA identification is only accurate to around 1 in 10,000 people. That may sound pretty accurate, but that would yield 10 suspects for any piece of DNA evidence in a city of 100,000 people. I'd certainly hope that nobody would be convicted purely on DNA evidence unless the other 9 people had been traced!
That said, I find it pretty creepy that any body would have the legal (if not moral) right to compile databases of DNA information "just in case." So much for the presumption of innocence!
One of my clients uses a local ISP to provide their DSL connection. This in itself is something of a nightmare, since one phone company manages the internal wiring, another manages the connection from the building's outside to the exchange - and then routes it to the small ISP's network. The ISP doesn't have much of their own infrastructure, they largely resell another company's bandwidth - who in turn are reselling another company's!
This never really sounded sensible, and the absurdity of the situation became painfully obvious a few weeks ago when we discovered that nobody at my client could connect to a specific server on the other side of the country. After waiting long enough to be sure it wasn't the usual sort of Internet connectivity issue, we began trying to find out what was happening. We could ping the server, and could traceroute to it - but connections to its primary services always timed out. It took a week of phone tag to find out that the reason was that a router, several layers up the provider-tree, was misconfigured. Even then, it was only fixed because my boss happened to know someone at the appropriate company!
Its pretty frustrating, overall. Each ISP one calls tries very hard to filter you out with their regular support staff, presumably to ensure that their sysadmins remain sane. It shouldn't take a week of phone calls to get one configuration issue resolved (or explained) - but that's the downside of a big, distributed, barely regulated system!
There are a few problems with a completely flat DNS-replacement:
1. You would quickly have even more scuffling over "important" keywords. At best, it would force people to be a bit more creative in thinking up words for their businesses - but the current TLD system does this anyway. At worst, the obvious words ("music", "books", "computers", etc.) will be snapped up by big players and completely unavailable to the smaller fish. This is the primary problem I see with "keyword-based" internet navigation. You can see an example of this in action if you have a Windows box; MS are partnered with a keyword-based Internet company (the name of which escapes me right now). Type just a word into your IE5 Address box, and if there's an associated keyword - you get 1 site, and only 1 site.
2. Hierarchical DNS is really useful from an administrative point of view. Where I work, we have several subdomains (many pointing to the same physical location). If you combine that with a keyword system, and you start allowing mail.computercompany - guess what, you have DNS again! In larger companies, DNS hierarchies are useful for representing physical layout, department divisions, etc.
Arguably, what's needed is more subdivision, not less. From the point of view of the consumer, it would be nice if domain names were organized somehow - but it can be pretty difficult to categorize things, sometimes. From the point of view of forced organization through TLDs, I'd be inclined to say that it won't happen - look how meaningless the distinction between.net and.com has become, and how well the various.state.country TLDs fared!
Of course, we do have a pretty good system for organizing domains right now: search engines.
Slashdot Mirror
While none of the MS employees I know want to create an Orwellian world, this would seem to be a very efficient way to go about it.
Maybe I do need to switch entirely to FreeBSD...
Anyway, the first thing to note is that by default Smart Tags are disabled - but they are really easy to enable. There is also an advanced setting to "always display smart tags" - giving users the option to override meta tags found on individual sites. It defaults to off, and as long as it stays that way I'm okay with it.
After switching on Smart Tags, the only smart tag I could find in this thread was the word Microsoft - and smart tags don't look like links. They are underlined with a dotted black and purple link. When you click on them, they don't take you anywhere - they pop up a tooltip with a little "information" icon in it and a pull-down list of pieces of information you might want to see. Very innocuous, very obviously not a link supplied by the document's author - and very unobtrusive.
The current IE6 beta also has some nice privacy features, particularly the option to block third-party cookies. Opera and Mozilla have had this for a while, but it's nice that IE has caught up!
I have to say, I think that this is a storm in a teacup (unless you are in the habit of making your links not act like links, and look different...)
Good point. On the other hand, have you ever seen someone try and get basic office work without MS Office skills? You can't even get work at a local temp agency here without at least MS Word or Powerpoint capability! So while I agree with you that things change, it is definitely worth making sure that kids who are about to graduate (and not all - or even very many in some areas - will go on to college) have skills that will be immediately useful in the workforce.
While in general I agree, there is one important reason to keep MS applications in public schools: you produce users familiar with MS office. Office remains the dominant business application, and if public education is to prepare students for future employment then it should provide a solid grounding in the applications that employers use.
The ironic thing about MS Office's bloat is that nobody I know/consult for uses more than 80% of any Office program's featureset... but they all use a different 80% and become very upset if their personal subset isn't available.
The other thing lacking in all of the MS Office alternatives I've tried is that of a good VBA clone. Like it or loathe it, Visual Basic for Applications makes it really easy to knock together a customized office solution. (Don't try the "nobody does this" argument, because it simply isn't true - I make a good deal of my living producing customized office applets for clients!).
Then there was the issue of applications... a few bits of shareware, some mediocre stuff written with gpp, some *NIX ports - and a lot of really expensive business apps I didn't need. Its emulation was decent, but Doom without sound? Never!
Anyway, I switched to NT3.51 and never looked back. Nowadays its Win2k Professional and FreeBSD for me!
In all honesty, a "suitcase nuke" (still kind of a big suitcase, but the word fits) attack is much more likely than a missile attack for a number of reasons:
Smaller blast; you tend to hit what you want to hit, not an indiscriminate area.
Less traceable; if you want to perform an action covertly, this is the way to do it. This is especially true as more and more states gain nuclear technologies. It's not inconceivable for a state-sponsored terror group to use a suitcase nuke, for example - and it would be really hard to decide which nation to retalliate against! Conversely, the iniator of a missile attack is readily identifiable.
More likely to hit what you want to hit. This doesn't apply to the US and Russia, whose missiles are nice and accurate - but for a rogue state with missiles of "hitting the right city is nice" accuracy (Scud-A, Frog, etc.) levels a suitcase bomb presents certain advantages.
Easier to get to target; no obvious warning signs, and nuclear packages are hard to spot. On the other hand, in a large-scale (relative to the parties; the Gulf war was a HUGE war for Iraq, not so big for the allies) war scenario missiles have the advantage that you don't need to get your people into enemy territory. There is also the "missiles make you a real power" theory floating around, to which many smaller states seem to subscribe. Certainly, if your objective is denial (ie. "shoot the carrier battlegroup before it gets close enough to bomb you" and "make sure that nobody wants to visit airbases in the area" missiles with NBC [Nuclear-Biological-Chemical] warheads can still make a certain amount of sense.
That's a long winded way of saying that you are right - suitcase nukes are very probable. :-)
The effects of detonating a thermonuclear weapon (as opposed to your run of the mill nuclear weapon) high-above DC would be pretty devastating - depending upon the exact rating of the weapon, weather, topography, etc. you could reasonably expect much of DC to be destroyed. A high-altitude airburst would minimize residual radiation effects, maximize blast damage (greater line of sight area for blast overpressure damage - the primary kill factor for a thermonuclear weapon) - and you'd make quite a mess.
On the other hand, ground-bursting a thermonuclear weapon on a ship in harbour is going to be a LOT less destructive relative to a well-placed airburst. You'll have more secondary radiation issues to deal with (the fireball really HAS to reach ground level, by definition) but blast damage is likely to be relatively localized by the quantity of solid matter it has to pass through/over to knock more stuff over. Remember, blast reflects/wraps on solid objects that don't break - so while the first few buildings would crumble against pressure, the buildings behind them would receive MUCH weaker effects. Detonating a full-scale thermonuclear weapon on a ship really is a bit of a waste because of this!
All of this ignores the strategic lunacy of targeting cities in general, but that's another can of worms. :-)
Depends upon the nuclear weapon. :-) Seriously, a man-portable nuclear weapon could detonate and probably not inflict massive damage beyond the first row of buildings/hard objects surrounding the port. Fallout would be an issue, although weapons designers can minimize this to an extent - certainly not as bad as the apocalyptic impression spread by phrases such as that one. Remember that people in shelter's around Hiroshima ground-zero were relatively unscathed - and train service for the city was functional again in a few hours. Nuclear != Armageddon, just a particular type of weapon.
Highly recommended reading:
S. Glasstone & P. Dolan, The Effects of Nuclear Weapons, GPO.
Also, US Navy sonar will most certainly detect something appraching underwater at over 250mph (and maybe they could do something about it, or maybe not...) but if a nuke is hidden on a ship then the Navy very well may allow them to dock.
This is a big issue, and one that the Pentagon is familiar with (although navy brass tend to reject the notion that nuclear weapons will ever be used). In a recent wargame in which I participated, red team was able to do a ridiculous amount of infrastructure damage to the US with this very tactic. I'm told that there are systems in R&D for over-the-horizon detection of nuclear weapons... although I have no idea how that would work. For all I know, it's military vapourware.
It should be possible to detect inbound weapons travelling at >250mpg, the question is what to do about it once you've detected it (the exact same problem as with ballistic missiles). Shooting it down requires the existence of a pretty impressive infrastructure designed to stop that particular type of attack - and as you say, probably won't help against the shipping boat packed with explosives.
"News for Geeks" does not mean "National Inquirer level humour."
(start javascript)
If Javascript isn't supported then redirect to "you can't view this page"
(end javascript)
On the other hand, an important lesson of the dot-com era seems to be that commercially oriented websites should encourage people to buy their products from their website. Sure, its easy for lazy webdesigners to "require" Internet Explorer, X version of Java/ECMA script, or similar - but that loses customers. If your page isn't readable with Lynx, you've lost a great many blind and visually impaired users, not to mention users who browse sans-graphics in the hope of speeding up their connections. If your page relies upon the existence of V4 browsers, you've lost a whole bunch of users on cellphones, PDAs and some primitive appliances. Even frames may be pushing it a bit in terms of global compatibility!
What this group are forgetting is that HTML is designed to convey content and not a specific look and feel. Ideally, a webpage should be useable on any device, irrespective of platform.
That said, it can be really hard to stop web developers from going nuts with Flash, sometimes. :-(
Much of FreeBSD 5's kernel development is focussed on improving SMP performance, particularly with respect to the TCP/IP stack. I don't know how long it will take for it to be declared stable, but I've had no real problems with the version that shipped in the 6-CD BSD box I found at Staples.
I recently had to evaluate some streaming video (over LAN) solutions, and we ended up with the Microsoft solution because its inexpensive (effectively bundled with 2k Server) and offers pretty good quality. A major factor in deciding, though, was platform: 99% of the likely userbase used Windows, and weren't techies - meaning that they didn't want to bother with difficult plugin installations!
True, it does appear useful - but its only really useful if you can catalogue the whole world, or at least that portion of it that's likely to have been in a certain locale at a given time. For example, if you know that there are probably 9 other suspects living within a town (remember, the 1/10,000 figure is a statistical average - don't rely on it) how do you know that the suspect wasn't from a different area? If you have a system of voluntary DNA evidence donation, all you do is implicate the law abiding types who volunteer the evidence. If you build a database of the whole country, your sample becomes so large as to be relatively meaningless:
Conceivably, if you could check DNA evidence against a database of everyone in the UK, you could come up with a list of a few hundred thousand suspects for every crime. On the other hand, regular police work generally provides a much narrower suspect list!
DNA evidence is useful secondary evidence, but it should not be relied upon as primary (or worse, sole convictor) evidence.
No, there is not. In fact, there have been several cases convicted purely on the basis of confessions that were later withdrawn! (Fortunately, that practice has since been outlawed in the aftermath of the Birmingham 6 and Guildford 4 inquiries).
That said, I find it pretty creepy that any body would have the legal (if not moral) right to compile databases of DNA information "just in case." So much for the presumption of innocence!
This never really sounded sensible, and the absurdity of the situation became painfully obvious a few weeks ago when we discovered that nobody at my client could connect to a specific server on the other side of the country. After waiting long enough to be sure it wasn't the usual sort of Internet connectivity issue, we began trying to find out what was happening. We could ping the server, and could traceroute to it - but connections to its primary services always timed out. It took a week of phone tag to find out that the reason was that a router, several layers up the provider-tree, was misconfigured. Even then, it was only fixed because my boss happened to know someone at the appropriate company!
Its pretty frustrating, overall. Each ISP one calls tries very hard to filter you out with their regular support staff, presumably to ensure that their sysadmins remain sane. It shouldn't take a week of phone calls to get one configuration issue resolved (or explained) - but that's the downside of a big, distributed, barely regulated system!
There are a few problems with a completely flat DNS-replacement: 1. You would quickly have even more scuffling over "important" keywords. At best, it would force people to be a bit more creative in thinking up words for their businesses - but the current TLD system does this anyway. At worst, the obvious words ("music", "books", "computers", etc.) will be snapped up by big players and completely unavailable to the smaller fish. This is the primary problem I see with "keyword-based" internet navigation. You can see an example of this in action if you have a Windows box; MS are partnered with a keyword-based Internet company (the name of which escapes me right now). Type just a word into your IE5 Address box, and if there's an associated keyword - you get 1 site, and only 1 site. 2. Hierarchical DNS is really useful from an administrative point of view. Where I work, we have several subdomains (many pointing to the same physical location). If you combine that with a keyword system, and you start allowing mail.computercompany - guess what, you have DNS again! In larger companies, DNS hierarchies are useful for representing physical layout, department divisions, etc. Arguably, what's needed is more subdivision, not less. From the point of view of the consumer, it would be nice if domain names were organized somehow - but it can be pretty difficult to categorize things, sometimes. From the point of view of forced organization through TLDs, I'd be inclined to say that it won't happen - look how meaningless the distinction between .net and .com has become, and how well the various .state.country TLDs fared!
Of course, we do have a pretty good system for organizing domains right now: search engines.