I suggest you re-read Genesis 11:1-9, preferably using a Bible with good notes, since you seem to be having a hard time finding anything other than literal meaning there. First off, the Tower of Babel is a reference to the chief ziggurat of Babylon, the Esagila. we've already built buildings much, much taller than that, and God's not struck them down. the story of babel isn't about tall buildings, it's about building an urban culture in which God has no place. it's about pride and presumption. if you're looking for scriptural backing of this sort of endavor (mars exploration, and indeed exploratory science in general), i'd point you at 2nd Timothy 1:7 - "For God hath not given us the spirit of fear; but of power, and of love, and of a sound mind." i'd point further to the several parables about how not using your gifts (including the intelect and reason provided us) is sinful, and displeasing to God (sorry, can't find verse numbers right now). i'd further point out that, in anything other than the strictest literal reading, one which ignores everything we know about astronamy, meterology, and general science, "heaven" isn't contained in the sky. in the Babel story, it's a metaphor; the same thing ticked God off there as it did in the garden of Eden, and in Sodom (no, that story isn't about homosexuality): pride and arrogance - thinking oneself equal to God. leaving the planet is no more offensive to God (from reading the Bible, anyway - i have no inside information here) than leaving your town; riding a space shuttle no worse than riding a horse.
i hate political discussions, but this is a bit over-the-top to not comment on. you're kidding, right? have you just not been paying atention for the last several months? in our recent election, which party brought things to the federal courts, taking the power away from the states, first? GOP. which was the first to look to have a fedreal court overturn the decisions of a state court? GOP. which advocated having the federal gvt. take over running national elections? GOP. none of this is to say the Dems are huge states right advocates: they're not. but while the GOP was a firm advocate of state's rights 50-100 years ago, in the past 20 or so, this has fallen to the whims of whatever's most politicaly expedient.
what's worse, you seem to be confused about keeping this at a state level being good for UCITA oposition. i'd rather have it national, where the more technologicly aware centers are able to impact the decision for the entire nation.
and what's worse still, as an earlier poster has pointed out, you're mis-representing the effects of "state's rights" on individual rights. there isn't guaranteed to be any corolation there.
your CPU works better than you think. your brain does indeed rely on stereo vision (for which you obviously need both eyes) for detph perception. that's not the brain's only method - it also takes clues from context, such as relative size of known objects. but the stereo vision is its primary method. the problems you describe - trouble with velocity, or in quickly-changing situations - are exactly the result of this sort of limited depth perception. even if one eye has degraded vision for whatever reason, your brain still makes use of the information that eye provides, particularly for good depth perception.
alright, there's a number of people in this thread using the fact that many ISPs have prohibitions against NAT (regardless of whether they enforce them or not) as evedince that we should move to IPv6. i mean, that would make IP addresses no longer a commodity, right? but that's not the least bit relavant. the whole point of NAT is that you're only using one IP. the commodity status of IPs isn't the least bit relavant. what's being restricted (or what they're trying to restrict) is the number of boxes you can run off their link. their first shot at this is to only give out one IP per line (works with DSL, usually with cable modem). this eliminates most people who'd want to use multiple boxes. so some of them get NAT. that's harder (if not impossible) to prohibit technically, so the build the restriction into the license. they're not interested in restricting IPs, per se. the switch to IPv6 wouldn't help anything. restricting IPs is a means to an end, which is restricting the number of boxes.
oh, and they do have prices reflect the "no servers" rule - if you want to run a server, you're supposed to get a business grade connection. same with multiple IPs. they're already doing exactly what you said, and you don't like it.
a few corrections. first, the DreamCast does indeed have a storage medium: two, in fact. it's got a fairly large read-only one in the form of a CD-ROM drive, and a fairly small read-write one in the form of NVRAM.
also, even if it had no persistant storage, that doesn't detract it from being an excelent network computer. ever used an NCD X-terminal or a diskless workstation? typically, these things even get their kernel over a network (some NCDs use PCMCIA memory cards instead), and those that need more get their file system over the network. this is an established practice, and it is actualy a plus for network computers, reducing maintenance costs, time spent on a specific peice of hardware, and total cost of ownership. it also makes admining the boxes a lot more fun.
and it is neither fair nor acurate to say Inferno is "far less advanced" then Linux. it doesn't have as much stuff, i'll grant that. and some of that "stuff" is even good: wide applications support, diverse hardware (specifically peripherials) support. but plenty isn't (X, for starters, or NFS... i know they do useful things, and are better than having nothing, but they're not the best answers to the questions). Inferno has much better solutions to a specific set of problems (see Inferno's draw interface, or the file protocol, Styx), and is much smaller, more consistantly designed, and easier to understand to boot.
and to clarify the price issue, Inferno can be had for free, with the exception of some core OS code. that means i can ship DreamCast CDs around for free just fine, thank you. and i can ship out source for most of the system (including every app) to go with it.
hey, maybe it's not for you. but to say that makes it a weak story is a little narrow minded, wouldn't you say?
you're kidding, right? ah, i'm continually amazed by the incredibly insular nature of the tech community, and even more by the tech-elite community. my mother doesn't have a clue what linux is. neither does my sister. nor anyone my mother works with (a bunch of elementry school teachers). neither has most of the world, or even most of america. that's a far cry from "just about every one". the average person doesn't know anything about operating systems, doesn't care, and shouldn't need to. if people are ever going to buy set top boxes, it won't be "because it runs linux". people want it to do certain things, in a way they can understand. the technical details are beyond most people, and that's fine: i don't know the details of how my car works, but i know i want it to do certain things. if Inferno can provide a good interface to functions people want, "the average Joe" isn't going to care whether it's Inferno or linux.
i doubt Vita Nuova plans on making much money with Inferno on Dreamcast, no. the SH[34] chip is used in loads of handheld devices (last i checked, the majority of WinCE devices were SH[34] devices, with a large chunk being MIPS). the DreamCast was suggested (by a/. reader, not by Vita Nuova) as a reference platform for doing the port, which would have the nice side effect of getting us a really cool OS for our DreamCasts.
hey, I use it. i'm osting the message from Charon, the Inferno web browser, running hosted on Plan 9. Inferno and Plan 9 are the only OSs i use (aside from a remote Solaris box once in a while). more to the point, i think the focus of Vita Nuova's original statement was getting Inferno running on the SH[34], not the DreamCast specifically. the SH[34] chips are used in a whole bunch of WinCE and other handheld devices. there's also a NetBSD port (no, really?) to those chips, and various things that use them.
um, Plan 9 isn't based around WORM storage, really. the file server itself is designed to run with an optical jukebox, and does some really neat things with it, but Plan 9 (even the above-mentioned file server) can be run off regular magnetic disk (or out of nvram, or off CD, or whatever) just fine. Plan 9 (and Inferno) also doesn't care whether that storage is local or not, making it much nicer for a set-top box i don't want to store thigns on, and want to be able to swap out on demand.
Re:Been done here for ages, and it works.
on
The Unblinking Eye
·
· Score: 1
Its just a question of trusting the authoriteies.
exactly the issue. "the authorities" have something of a history of abusing power (almost regardless of who "the authorities" happen to be at the time). and it's worse than, say, not trusting congress, or our state legislature, who're more-or-less beholden to us for their jobs. we're talking about trusting the police. i don't know of anywhere on the planet where you can "vote out" cops. maybe in a direct participatory democracy this would work. not in a representative democracy. and hell, we don't even have that!
There's no benefit to just showing [your source] around, if you don't allow people to make improvements or derivatives.
well, that's not quite true. showing your code around certainly doesn't make you Open Source, and you don't get a lot of the benefits that model gives. but you do get some benefits (or rather, the community does). we can examine code for security issues; we can't fix them, but we can let people (including the author) know they're there. it might also help write better apps - M$ has long been said to have hooks in their OS that they don't tell other people about. it'd be nice to know about those. that being said, you are correct that this is a far cry from Open Source, and still reflects the "trust us, we know best" attitude we all know so well.
I have yet to see an x86 server where the fan is actively controlled by chassis temperature.
while i agree it's rare, i have seen at least one: an ALR Revolution Quad 6 we bought a few years back. probably the best x86 server system i've ever seen. and it had a fan on each P6 CPU controlled by that CPUs temp, four case fans controlled by ambiant temperature, and a power supply fan (which was always on full, yes). all configurable, while running, via a LCD touch-sensitive pannel on the front. sweet. i don't know what ALR's doing these days, but if they've got a current generation of this box, i'd recomend it for a high-end x86 server. we ran Solaris 2.6 on it with no problems.
oh, it gets even better. Inferno is the OS, Limbo is the programing language, which compiles into Dis executables. the web browser is Charon, the communications protocol is Styx. it all comes from Dante's Divine Comedy.
virtual machine's arn't new concepts, no. Java didn't invent it, either; it's older than that. but whereas the JavaVM provides just the virtual machine, the Inferno plugin (and Inferno in general) provides an entire virtual OS. you get networking capabilities built in, for example, and a consistant model for accessing devices, of all types, remote or local. and, arguably, it's the "everything is a file" model that makes it all worth it. that's what enables the consistant, simple model for device and network access, which in turn results in simpler, smaller code.
the fact that IPv6 is IPv4-compatable is a given. it's also not relavant. nor is the fact that it provides for site-local addresses. so does IPv4, sort of: 10.0.0.0, for example. there's a whole lot more to e2e than address space. let's say i have a network (a satalite network transmitting video, for example) for which IP is a lousy protocol. i should be able to connect that to my terrestrial IP network, and the Internet at large. the Internet's infrastructure shouldn't have to know how to deal with every protocol, no. but it should provide for interoperability with "external" protocols without a break of its underlying model. the very concept of "external" violates e2e, so that clearly can't work.
When routers and switches are moving packets they are dumb
again, true but useless. wires move packets, too. they're dumb. you willing to replace all your routers with wires? or with routers that don't do route lookup? 'cource not. you want that inteligence there. in the network. i'm not arguing e2e is impossible. never have been. just that it's not a good idea. DNS is problematic enough, given the glitchieness of the 'net; i can't believe you want to have to go to a route server to look up routing information, and then have your host either program routes or controll the routers. yeah, that'd scale to, oh, a few dozen hosts.
Just follow the standards and no one gets hurt.
which standards are those? the e2e standards? oh, right. wanna gimme an IETF number? hell, i'd settle for an RFC. oh, and standards don't change?
End-to-End requirements are not some holy thing that ip geeks love because some standards body says to. People argue for it because it works.
here i have to say i think you're just wrong. see the above routing arguement, for example. pure e2e doesn't work, never has, never will, for anything larger than a bus network (token ring, for example). and even were it true, it's a useless statement. "it works," huh? how long? under what circumstances? 'till what size? pure central controll "works" too, untill you get to a certain size. the largest network in the world is still central-controll based, with modifications, not e2e with modifications. PSTN, baby. more endpoints than the Internet. way more.
i'm not interested in gratuitious incompatability, either. but i think e2e is too strick a requirement, given the growing size and diversity of the internet.
your arguement might hold up for DNS, but not routing. when people talk about dumb vs. inteligent networks, do you think they're talking about the wires? it's routers, switches and other network equipment that have that "inteligence" or not. and they clearly have some - like the routing info. the e2e model doesn't truely exist unless you've got either direct controll over your switch or a direct connection to every endpoint. note i'm advocating neither; i'm just pointing out that's what's required for a truely e2e network.
2^32 = plenty. More than enough.
2^128 = loads. Way more than enough.
ah, good old "proof by lack of imagination". think about it. if every device (where device is some set of my toaster, TV, socks, refrigerator, etc.) has an IP address, and is connected to the network, that leads to way more infrastructure, too. more nodes for management. it's a worse-than-linear increase. but given that, the numbers are still pretty generous. bringing us to your assertion:
The problems with ipv4 do not involve the number of nodes, it is a lack of management and arguably managability of the address space an org is allocated.
to which i agree entirely. another reason e2e doesn't make sense. i should be able to design my network any way i like, you should be able to design yours any way you like. it should then be up to us to provide for inter-connectivity. it's ironic that, in the internet culture dominated by talk about de-centrilization of power and no single "root", ICANN and IANA still define what pretty much everyone can and can't do. to say nothing of the DNS mess. those I* folks have way to much to say about how i run my network.
I'm sure you're experiences with NAT all involve you home or campus network.
you are mistaken. unless you take a really, really broad definition of "campus". try dealing with such things for multinational corporations. ones that used to be the telephone company. i'm well aware of the issues surrounding NAT today. but i believe that's because they're trying to work within an e2e network, rather than being able to admit e2e isn't always the way to go.
that's a remarkably short-sighted view for the networked age. people who don't live in cities are all anti-social? rubbish. ever heard of telecommuting? i like being able to go anywhere and hook up a modem (or just make a phone call). and i can certainly see the apeal for living in neither a city nor a suburb. those more sparsly populated areas are the ones it's more expensive, per person, to get service too. in addition to providing for different people living in different places, it paves the way for growth into new areas by enabling the construction of an infrastructure before market preasures would otherwise require it. also, keep in mind that the real issue is "hard to service areas" not "people i think are useless". it's far more difficult to run lines into the appalachian mountains than down the block in Manhattan. yet interesting people or organizations can be found in both places. the diversity ubiquitous telephone access provides society is far and away worth the small cost.
the article posted so far, and from what i can tell from it, the conference itself, assumes that e2e is a good idea. the discussions all revolve around "how do we save e2e" or "what do we do about things that violate e2e" before asking "is e2e a good idea, given the current state of the internet, and its likely evolution?".
i know i'm risking flaming here, as e2e is something of a holy cow on the internet. but bear with me. there's two ways of looking at it: internet-centric or network design in general. let's look at it first from an internet-centric point of view first. the basis of e2e is that all the "inteligence" is pushed out to the edges. well, upon examination, this doesn't hold up in the practice of the internet. please raise your hand if you've got the entire internet routing table on your host. or the full hostname->ip mapping for the internet. no? that's right, that info (inteligence) comes from the network - switches, routers, DNS servers, and the like. a common counter is "DNS isn't an application" but to the network, it is. it's just transferring bits around. so's routing. but the inteligence that makes it work lives in the network, not on the edge (my host). now look at it from a general network design, non-internet-centric point of view. from the facilities point of view, the only application the "network" knows about is setting up and tearing down data paths. whether it's a TCP stream, a UDP datagram, ATM connection, or a phone call - the "network" delivers data in defined ways. while QoS can be described as a (potentially problematic) upgrade to this service, NAT is not inherently anything more than a breakdown of the DNS system (although, as currently implemented, it introduces more problems). it's compensating for a failure of IP: the failure to allow real heirarchical address management. IPv6 doesn't solve this, it just pushes the problem farther out into the future. if, as many people have suggested, my toaster, car, telephone, TV, and socks will all have IP addresses one day, even the new addresses provided by IPv6 won't last very long. so what to do? on the internet today, DNS first resolves a name into a number. it's the numbers that get things done; it's the numbers that're important. imagine a network where the names were what was important; where i can define, for example, myco/us/ny/ny/10/107 as the 107th terminal on the 10th floor of the NY, NY office of my company. that's a crappy naming system, but the point is that in a world where the names are the key, i can have much better control over the structure of my network and addresses. it also simplifies routing and such, as the area of responsability becomes much clearer.it's a direct violation of the e2e model, but so what. oh, and before you say "it can't be done, you need numbers" or whatever, it's already been done. several times. read up on Datakit, a network architecture developed by Bell Labs way back when. it had other problems, like static routing, but it got a lot of things right that the Internet's still fumbling with. again from the network design point of view, answer this question. which is easier to upgrade: 1 server or 100,000 clients? imagine if every time AT&T wanted to change your billing rate or add the capacity for three-way calling they had to upgrade your telephone. and everyone else's telephone. by concentrating the inteligence (and, therefore, the complexity) the telephone network has become the most succesfull network in the world, with more endpoints than the Internet, and way more users. concentrating the complexity means less overall work, less chance for something to fail, and lower barriers to innovation.
that being said, the business model of the Bell system has introduced other barriers to third-party innovation and service, but those are business issues, not technical ones. e2e results in increased overall complexity and additional problems in the name of distributing it away from any central point. sounds like a bad trade to me.
Plan 9's been around for a while, but it's catching on very well now. this is largely due to the recent new release and the open source license it's released under. the new release also includes substantial reworking of the internals and numerous other technical improvements. i highly recomend giving it a shot. i've been using Plan 9 for a few years. my progression (outside of work) basically went Micro$oft, Linux, FreeBSD, Plan 9. i've tried various Linux distros, *BSD, Solaris, and a few other smaller projects. Plan 9 blows them all away, hands down, in terms of everything but hardware support and application base.
Linux has a good goal - make a free (or open, or whatever) Unix kernel. Unix is good, free is good. but it's about cloning Unix. it's not about advancing the state of the art; it's not about innovation. and adding new support for a really fast networking card or getting a file system to run %15 faster isn't innovation. Linux doesn't advance the art of the OS any. if you're interested in that, look to Plan 9 and Inferno.
two points of criticism. first, the article itself is pretty minimal in the context it provides. it doesn't talk about what went on before or after, nor what was going on at the same time. still some good information, but nothing to get worked up over. second, the actual work described wasn't as revolutionary as people here are saying. very good, to be sure, but the really revolutionary stuff came when the idea made its way into other people's hands, and other systems. try to imagine a shell without pipes or even IO redirection. the shell, as it first existed, was mostly an evolution of batch processing. a big improvement, to be sure, but just evolution. the really interesting stuff showed up with the Unix shell, at the hands of Thompson, Ritchie, and a few others. see this paper by Dennis Ritchie, particularly the sections on process control, IO redirection, and (most especially) pipes for some really revolutionary developments in the history of the shell.
not that it impacts your point any, but the statement "...equipped with (usually) Intel's SA-110 (StrongARM)..." isn't quite true. Hitachi's SH-3 and SH-4 line has good market share, especially in WinCE (er, whatever it's called today) devices. MIPS chips have some embeded market share, too. but the winner, by virtue of its inclusion in the Palm series devices, are the Motorola 68K series chips. these are relatively low power consuming chips, but are slow (the same arch powered the pre-PPC macs and per-SPARC Suns). but IIRC, all these are smaller than the Crusoe chips i've seen described. incidently, i really like the StrongARM chips, and would like to see them gain lots of ground. but they're not there yet.
this is really interesting. you've managed to take the same observations about problems with existing sub-notebooks and sub-subs other/.ers have noted, duplicate them, and reach some bad conclusions. you're missing three points, as i see it: sure, tiny screens with low resolution suck. but that doesn't mean small(er) devices are a bad idea. check out what Microvision has to offer. it'll be a while before their stuff is cost effective, but they can (and do, for the military) already suply the equivalent of a 19" monitor, in really tiny space. input might be harder - i don't know of anything but chording keyboards (which i can't comment on, never having used one) that give long-term usability. dictation systems arn't nearly good enough, graffiti-type entry is useless for entering documents or terminal stuff. any other good ideas? any one last big issue with the above post is this bit:
...manufacturers have discovered that people don't want laptops much below 10"x7"x1" or below about 3.5 pounds.
huh? i can see the issues with shrinking size below about what you said, but weight? what's the down side to sub-1lb devices? just because the current state of the art makes really tiny devices uncomfortable doesn't mean it'll always be that way.
sorry, but you're mistaken. there are other (even older) open source, free OS projects out there, that developed independantly of the FSF. (Net|Open|Free)?BSDi? are all examples of such projects. while all the BSD distributions i know of currently ship with GNU code, none were originally dependant on it. i even built a Linux distribution using BSD tools and utilities rather than the standard GNU gunk. don't get me wrong: i think RMS has had a very strong positive influence on the political free software and open source (like it or not, RMS) movements. but they're not beholden to him, he did not create them, and neither was his idea. further, i think you're right about non-Linux projects being very worthwhile. but attributing all the distaste for HURD to anti-RMS/FSF sentiment is off base. the fact that HURD is still in alpha, seven years or so after being started, is a big minus. but the bigest problem is the fact that your statement:
it is admirable that they are trying to advance the state of free operating systems
is simply false. HURD has nothing new to offer. most of its ideas are lifted wholesale from other places, and the organization of those ideas is entirely non-innovative. there are plenty of non-Linux projects that're very worthy of your statement above. i'm not even a Linux user any more, in favor of them. but HURD doesn't qualify. if you're looking for good OSs, QNX is a good microkernel, Mac OS X has some very good potential (although the free parts are basically just another Unix), and i personally prefer Inferno and Plan 9 (which HURD took a few ideas from).
this is a common issue. several major people and organizations (on both sides of the War on Some Drugs) have come out criticizing D.A.R.E. for equating use with abuse. its generally viewed as the single biggest cause for their questionable (at best) performance. some reasoning and explanation:
D.A.R.E. was started by one police officer (i forget his name), and is staffed and tought in schools by other police officers. as you might imagine, this leads to a very legalistic view of drugs. use and abuse are, for the most part, the same to the law, so they're the same to D.A.R.E.
the study cited above is not the first to call into question the effectiveness of D.A.R.E. i've seen about a half dozen that provide statistical evidence that, beyond just not having an effect on urban drug use, it incureases drug use in suburban schools where it's tought.
the problem, in terms of fixing the problems with it, is that it's exactly this legalistic attitude that makes it so popular. in the current political environment, it makes the police, school board, and local government look good to have D.A.R.E. in schools. also, neither the police departments nor the schools want the general public to know that they're wasting taxpayer money on a program that simply doesn't work (leaving aside the questionable value of its aims). but D.A.R.E. has lots of money, and a big name to go with it, so it's a high profile "win" for people who implement it, showing their "commitment" to protecting our children.
even other drug abuse prevention groups, like FCD, have said D.A.R.E. is a silly approach to the problem. that may just be because they're addressing the wrong problem.
Slashdot Mirror
I suggest you re-read Genesis 11:1-9, preferably using a Bible with good notes, since you seem to be having a hard time finding anything other than literal meaning there. First off, the Tower of Babel is a reference to the chief ziggurat of Babylon, the Esagila. we've already built buildings much, much taller than that, and God's not struck them down. the story of babel isn't about tall buildings, it's about building an urban culture in which God has no place. it's about pride and presumption.
if you're looking for scriptural backing of this sort of endavor (mars exploration, and indeed exploratory science in general), i'd point you at 2nd Timothy 1:7 - "For God hath not given us the spirit of fear; but of power, and of love, and of a sound mind." i'd point further to the several parables about how not using your gifts (including the intelect and reason provided us) is sinful, and displeasing to God (sorry, can't find verse numbers right now).
i'd further point out that, in anything other than the strictest literal reading, one which ignores everything we know about astronamy, meterology, and general science, "heaven" isn't contained in the sky. in the Babel story, it's a metaphor; the same thing ticked God off there as it did in the garden of Eden, and in Sodom (no, that story isn't about homosexuality): pride and arrogance - thinking oneself equal to God. leaving the planet is no more offensive to God (from reading the Bible, anyway - i have no inside information here) than leaving your town; riding a space shuttle no worse than riding a horse.
i hate political discussions, but this is a bit over-the-top to not comment on. you're kidding, right? have you just not been paying atention for the last several months? in our recent election, which party brought things to the federal courts, taking the power away from the states, first? GOP. which was the first to look to have a fedreal court overturn the decisions of a state court? GOP. which advocated having the federal gvt. take over running national elections? GOP. none of this is to say the Dems are huge states right advocates: they're not. but while the GOP was a firm advocate of state's rights 50-100 years ago, in the past 20 or so, this has fallen to the whims of whatever's most politicaly expedient.
what's worse, you seem to be confused about keeping this at a state level being good for UCITA oposition. i'd rather have it national, where the more technologicly aware centers are able to impact the decision for the entire nation.
and what's worse still, as an earlier poster has pointed out, you're mis-representing the effects of "state's rights" on individual rights. there isn't guaranteed to be any corolation there.
your CPU works better than you think.
your brain does indeed rely on stereo vision (for which you obviously need both eyes) for detph perception. that's not the brain's only method - it also takes clues from context, such as relative size of known objects. but the stereo vision is its primary method. the problems you describe - trouble with velocity, or in quickly-changing situations - are exactly the result of this sort of limited depth perception.
even if one eye has degraded vision for whatever reason, your brain still makes use of the information that eye provides, particularly for good depth perception.
alright, there's a number of people in this thread using the fact that many ISPs have prohibitions against NAT (regardless of whether they enforce them or not) as evedince that we should move to IPv6. i mean, that would make IP addresses no longer a commodity, right? but that's not the least bit relavant.
the whole point of NAT is that you're only using one IP. the commodity status of IPs isn't the least bit relavant. what's being restricted (or what they're trying to restrict) is the number of boxes you can run off their link. their first shot at this is to only give out one IP per line (works with DSL, usually with cable modem). this eliminates most people who'd want to use multiple boxes. so some of them get NAT. that's harder (if not impossible) to prohibit technically, so the build the restriction into the license.
they're not interested in restricting IPs, per se. the switch to IPv6 wouldn't help anything. restricting IPs is a means to an end, which is restricting the number of boxes.
oh, and they do have prices reflect the "no servers" rule - if you want to run a server, you're supposed to get a business grade connection. same with multiple IPs. they're already doing exactly what you said, and you don't like it.
a few corrections.
first, the DreamCast does indeed have a storage medium: two, in fact. it's got a fairly large read-only one in the form of a CD-ROM drive, and a fairly small read-write one in the form of NVRAM.
also, even if it had no persistant storage, that doesn't detract it from being an excelent network computer. ever used an NCD X-terminal or a diskless workstation? typically, these things even get their kernel over a network (some NCDs use PCMCIA memory cards instead), and those that need more get their file system over the network. this is an established practice, and it is actualy a plus for network computers, reducing maintenance costs, time spent on a specific peice of hardware, and total cost of ownership. it also makes admining the boxes a lot more fun.
and it is neither fair nor acurate to say Inferno is "far less advanced" then Linux. it doesn't have as much stuff, i'll grant that. and some of that "stuff" is even good: wide applications support, diverse hardware (specifically peripherials) support. but plenty isn't (X, for starters, or NFS... i know they do useful things, and are better than having nothing, but they're not the best answers to the questions). Inferno has much better solutions to a specific set of problems (see Inferno's draw interface, or the file protocol, Styx), and is much smaller, more consistantly designed, and easier to understand to boot.
and to clarify the price issue, Inferno can be had for free, with the exception of some core OS code. that means i can ship DreamCast CDs around for free just fine, thank you. and i can ship out source for most of the system (including every app) to go with it.
hey, maybe it's not for you. but to say that makes it a weak story is a little narrow minded, wouldn't you say?
you're kidding, right? ah, i'm continually amazed by the incredibly insular nature of the tech community, and even more by the tech-elite community.
my mother doesn't have a clue what linux is. neither does my sister. nor anyone my mother works with (a bunch of elementry school teachers). neither has most of the world, or even most of america. that's a far cry from "just about every one".
the average person doesn't know anything about operating systems, doesn't care, and shouldn't need to. if people are ever going to buy set top boxes, it won't be "because it runs linux". people want it to do certain things, in a way they can understand. the technical details are beyond most people, and that's fine: i don't know the details of how my car works, but i know i want it to do certain things. if Inferno can provide a good interface to functions people want, "the average Joe" isn't going to care whether it's Inferno or linux.
i doubt Vita Nuova plans on making much money with Inferno on Dreamcast, no. the SH[34] chip is used in loads of handheld devices (last i checked, the majority of WinCE devices were SH[34] devices, with a large chunk being MIPS). the DreamCast was suggested (by a /. reader, not by Vita Nuova) as a reference platform for doing the port, which would have the nice side effect of getting us a really cool OS for our DreamCasts.
hey, I use it. i'm osting the message from Charon, the Inferno web browser, running hosted on Plan 9. Inferno and Plan 9 are the only OSs i use (aside from a remote Solaris box once in a while).
more to the point, i think the focus of Vita Nuova's original statement was getting Inferno running on the SH[34], not the DreamCast specifically. the SH[34] chips are used in a whole bunch of WinCE and other handheld devices. there's also a NetBSD port (no, really?) to those chips, and various things that use them.
um, Plan 9 isn't based around WORM storage, really. the file server itself is designed to run with an optical jukebox, and does some really neat things with it, but Plan 9 (even the above-mentioned file server) can be run off regular magnetic disk (or out of nvram, or off CD, or whatever) just fine. Plan 9 (and Inferno) also doesn't care whether that storage is local or not, making it much nicer for a set-top box i don't want to store thigns on, and want to be able to swap out on demand.
that being said, you are correct that this is a far cry from Open Source, and still reflects the "trust us, we know best" attitude we all know so well.
oh, it gets even better. Inferno is the OS, Limbo is the programing language, which compiles into Dis executables. the web browser is Charon, the communications protocol is Styx. it all comes from Dante's Divine Comedy.
virtual machine's arn't new concepts, no. Java didn't invent it, either; it's older than that. but whereas the JavaVM provides just the virtual machine, the Inferno plugin (and Inferno in general) provides an entire virtual OS. you get networking capabilities built in, for example, and a consistant model for accessing devices, of all types, remote or local.
and, arguably, it's the "everything is a file" model that makes it all worth it. that's what enables the consistant, simple model for device and network access, which in turn results in simpler, smaller code.
i'm not interested in gratuitious incompatability, either. but i think e2e is too strick a requirement, given the growing size and diversity of the internet.
ah, good old "proof by lack of imagination". think about it. if every device (where device is some set of my toaster, TV, socks, refrigerator, etc.) has an IP address, and is connected to the network, that leads to way more infrastructure, too. more nodes for management. it's a worse-than-linear increase. but given that, the numbers are still pretty generous. bringing us to your assertion: to which i agree entirely. another reason e2e doesn't make sense. i should be able to design my network any way i like, you should be able to design yours any way you like. it should then be up to us to provide for inter-connectivity. it's ironic that, in the internet culture dominated by talk about de-centrilization of power and no single "root", ICANN and IANA still define what pretty much everyone can and can't do. to say nothing of the DNS mess. those I* folks have way to much to say about how i run my network.
you are mistaken. unless you take a really, really broad definition of "campus". try dealing with such things for multinational corporations. ones that used to be the telephone company. i'm well aware of the issues surrounding NAT today. but i believe that's because they're trying to work within an e2e network, rather than being able to admit e2e isn't always the way to go.
that's a remarkably short-sighted view for the networked age. people who don't live in cities are all anti-social? rubbish. ever heard of telecommuting? i like being able to go anywhere and hook up a modem (or just make a phone call). and i can certainly see the apeal for living in neither a city nor a suburb. those more sparsly populated areas are the ones it's more expensive, per person, to get service too. in addition to providing for different people living in different places, it paves the way for growth into new areas by enabling the construction of an infrastructure before market preasures would otherwise require it.
also, keep in mind that the real issue is "hard to service areas" not "people i think are useless". it's far more difficult to run lines into the appalachian mountains than down the block in Manhattan. yet interesting people or organizations can be found in both places. the diversity ubiquitous telephone access provides society is far and away worth the small cost.
the article posted so far, and from what i can tell from it, the conference itself, assumes that e2e is a good idea. the discussions all revolve around "how do we save e2e" or "what do we do about things that violate e2e" before asking "is e2e a good idea, given the current state of the internet, and its likely evolution?".
i know i'm risking flaming here, as e2e is something of a holy cow on the internet. but bear with me. there's two ways of looking at it: internet-centric or network design in general. let's look at it first from an internet-centric point of view first.
the basis of e2e is that all the "inteligence" is pushed out to the edges. well, upon examination, this doesn't hold up in the practice of the internet. please raise your hand if you've got the entire internet routing table on your host. or the full hostname->ip mapping for the internet. no? that's right, that info (inteligence) comes from the network - switches, routers, DNS servers, and the like. a common counter is "DNS isn't an application" but to the network, it is. it's just transferring bits around. so's routing. but the inteligence that makes it work lives in the network, not on the edge (my host).
now look at it from a general network design, non-internet-centric point of view. from the facilities point of view, the only application the "network" knows about is setting up and tearing down data paths. whether it's a TCP stream, a UDP datagram, ATM connection, or a phone call - the "network" delivers data in defined ways. while QoS can be described as a (potentially problematic) upgrade to this service, NAT is not inherently anything more than a breakdown of the DNS system (although, as currently implemented, it introduces more problems). it's compensating for a failure of IP: the failure to allow real heirarchical address management. IPv6 doesn't solve this, it just pushes the problem farther out into the future. if, as many people have suggested, my toaster, car, telephone, TV, and socks will all have IP addresses one day, even the new addresses provided by IPv6 won't last very long.
so what to do? on the internet today, DNS first resolves a name into a number. it's the numbers that get things done; it's the numbers that're important. imagine a network where the names were what was important; where i can define, for example, myco/us/ny/ny/10/107 as the 107th terminal on the 10th floor of the NY, NY office of my company. that's a crappy naming system, but the point is that in a world where the names are the key, i can have much better control over the structure of my network and addresses. it also simplifies routing and such, as the area of responsability becomes much clearer.it's a direct violation of the e2e model, but so what.
oh, and before you say "it can't be done, you need numbers" or whatever, it's already been done. several times. read up on Datakit, a network architecture developed by Bell Labs way back when. it had other problems, like static routing, but it got a lot of things right that the Internet's still fumbling with.
again from the network design point of view, answer this question. which is easier to upgrade: 1 server or 100,000 clients? imagine if every time AT&T wanted to change your billing rate or add the capacity for three-way calling they had to upgrade your telephone. and everyone else's telephone. by concentrating the inteligence (and, therefore, the complexity) the telephone network has become the most succesfull network in the world, with more endpoints than the Internet, and way more users. concentrating the complexity means less overall work, less chance for something to fail, and lower barriers to innovation.
that being said, the business model of the Bell system has introduced other barriers to third-party innovation and service, but those are business issues, not technical ones. e2e results in increased overall complexity and additional problems in the name of distributing it away from any central point. sounds like a bad trade to me.
Plan 9's been around for a while, but it's catching on very well now. this is largely due to the recent new release and the open source license it's released under. the new release also includes substantial reworking of the internals and numerous other technical improvements. i highly recomend giving it a shot.
i've been using Plan 9 for a few years. my progression (outside of work) basically went Micro$oft, Linux, FreeBSD, Plan 9. i've tried various Linux distros, *BSD, Solaris, and a few other smaller projects. Plan 9 blows them all away, hands down, in terms of everything but hardware support and application base.
Linux has a good goal - make a free (or open, or whatever) Unix kernel. Unix is good, free is good. but it's about cloning Unix. it's not about advancing the state of the art; it's not about innovation. and adding new support for a really fast networking card or getting a file system to run %15 faster isn't innovation. Linux doesn't advance the art of the OS any. if you're interested in that, look to Plan 9 and Inferno.
two points of criticism. first, the article itself is pretty minimal in the context it provides. it doesn't talk about what went on before or after, nor what was going on at the same time. still some good information, but nothing to get worked up over.
second, the actual work described wasn't as revolutionary as people here are saying. very good, to be sure, but the really revolutionary stuff came when the idea made its way into other people's hands, and other systems. try to imagine a shell without pipes or even IO redirection. the shell, as it first existed, was mostly an evolution of batch processing. a big improvement, to be sure, but just evolution. the really interesting stuff showed up with the Unix shell, at the hands of Thompson, Ritchie, and a few others.
see this paper by Dennis Ritchie, particularly the sections on process control, IO redirection, and (most especially) pipes for some really revolutionary developments in the history of the shell.
not that it impacts your point any, but the statement "...equipped with (usually) Intel's SA-110 (StrongARM)..." isn't quite true. Hitachi's SH-3 and SH-4 line has good market share, especially in WinCE (er, whatever it's called today) devices. MIPS chips have some embeded market share, too. but the winner, by virtue of its inclusion in the Palm series devices, are the Motorola 68K series chips. these are relatively low power consuming chips, but are slow (the same arch powered the pre-PPC macs and per-SPARC Suns). but IIRC, all these are smaller than the Crusoe chips i've seen described.
incidently, i really like the StrongARM chips, and would like to see them gain lots of ground. but they're not there yet.
sure, tiny screens with low resolution suck. but that doesn't mean small(er) devices are a bad idea. check out what Microvision has to offer. it'll be a while before their stuff is cost effective, but they can (and do, for the military) already suply the equivalent of a 19" monitor, in really tiny space.
input might be harder - i don't know of anything but chording keyboards (which i can't comment on, never having used one) that give long-term usability. dictation systems arn't nearly good enough, graffiti-type entry is useless for entering documents or terminal stuff. any other good ideas?
any one last big issue with the above post is this bit: huh? i can see the issues with shrinking size below about what you said, but weight? what's the down side to sub-1lb devices?
just because the current state of the art makes really tiny devices uncomfortable doesn't mean it'll always be that way.
don't get me wrong: i think RMS has had a very strong positive influence on the political free software and open source (like it or not, RMS) movements. but they're not beholden to him, he did not create them, and neither was his idea.
further, i think you're right about non-Linux projects being very worthwhile. but attributing all the distaste for HURD to anti-RMS/FSF sentiment is off base. the fact that HURD is still in alpha, seven years or so after being started, is a big minus. but the bigest problem is the fact that your statement: is simply false. HURD has nothing new to offer. most of its ideas are lifted wholesale from other places, and the organization of those ideas is entirely non-innovative. there are plenty of non-Linux projects that're very worthy of your statement above. i'm not even a Linux user any more, in favor of them. but HURD doesn't qualify.
if you're looking for good OSs, QNX is a good microkernel, Mac OS X has some very good potential (although the free parts are basically just another Unix), and i personally prefer Inferno and Plan 9 (which HURD took a few ideas from).
this is a common issue. several major people and organizations (on both sides of the War on Some Drugs) have come out criticizing D.A.R.E. for equating use with abuse. its generally viewed as the single biggest cause for their questionable (at best) performance. some reasoning and explanation:
D.A.R.E. was started by one police officer (i forget his name), and is staffed and tought in schools by other police officers. as you might imagine, this leads to a very legalistic view of drugs. use and abuse are, for the most part, the same to the law, so they're the same to D.A.R.E.
the study cited above is not the first to call into question the effectiveness of D.A.R.E. i've seen about a half dozen that provide statistical evidence that, beyond just not having an effect on urban drug use, it incureases drug use in suburban schools where it's tought.
the problem, in terms of fixing the problems with it, is that it's exactly this legalistic attitude that makes it so popular. in the current political environment, it makes the police, school board, and local government look good to have D.A.R.E. in schools. also, neither the police departments nor the schools want the general public to know that they're wasting taxpayer money on a program that simply doesn't work (leaving aside the questionable value of its aims). but D.A.R.E. has lots of money, and a big name to go with it, so it's a high profile "win" for people who implement it, showing their "commitment" to protecting our children.
even other drug abuse prevention groups, like FCD, have said D.A.R.E. is a silly approach to the problem. that may just be because they're addressing the wrong problem.