Along with fix-a-flat there will be ten-mile fuel cell quick jump kits for your car to get you to your next plug in.
What would be an even better (and just as useful) measure would be to run inductive chargers under the pavement on long stretches of the interstate system where you charge as you ride, or largely ride for free. Or (if we _must_) you have a Charge Pass(tm) meter to get billed for charge as you go.
Make fun of the charge-for-free if you like, but it is likely cheaper to give the passers-through the free ride than to clean up the mess of the fuel economy (plus billing etc). Imagine how well the power systems would be regulated if the fed were footing the bill for the electricity delivered to the interstate highway system.
Regardless of that sky-pie...
Since there is no inherent danger to making electricity available, as opposed to the inherent danger of making petrochemical distillates available, getting a charge as part of parking at your rest stop or attraction (or while parked at work because of your employers clean-air program) could delete the entire concept of the dedicated fueling station.
Go to Denies, or MacDonalds, or Taco Time, and get a (full or partial) charge while you eat.
Sound unlikely? Well in colder climates they already rig their parking lots with plugs to keep your engine block from freezing while you shop or eat.
Hell, having reduced-price or even free charging at the local highway rest-stop would cut way down on tired long-distance driving accidents. Pull over, take a rest, get a charge on us 'cause it's cheaper than having to dispatch Life Flight etc.
Once the "fuel" is electricity, which is literally everywhere in our society, we lose the need for the "filling station" to be a super-defined super-regulated contentious resource. Anybody can sell "their" electricity if they have the amperage to spare.
Once motive force stops being "rare and dangerous" gasoline, the paradigm totally goes to distributed recharging as the most sensible practice.
The democrats (a class of which I am nominally a member) can not seem to figure out that "running to the middle" is a loser strategy. It just turns you into a nobody.
Since nobody completely agrees with anybody, everybody will vote for the somebody before they vote for the nobody.
For every supporter you pander to, you lose two supporters you already have. Doesn't matter what side you are coming from.
In short, at the center of the pool, you will find a drain
- BSD ensures that the *next* recipient of the code may do what they want. - GPL ensures that *every* recipient of the code may do what they want....conversely...
- BSD ensures that the *publisher* of the code can take received contributions to the code and make them part of a private domain if he chooses. - GPL ensures that each *contributor* to the code can ensure that the publisher cannot incorporate his contribution into a private domain.
I would say that BSD code is far less encumbered than GPL code. The BSD license is effectively a release to the public domain. Given that you have to leave the copyright in place but you don't ever have to give out the code, makes the provisions of the BSD as toothless as a declaration of abandonment for downstream use. The BSD does however require symmetric abandonment by contributors moving patches upstream.
In contrast the GPL requires democratic fairness of all parties. Fair isn't Free, just as "freedom isn't without cost".
Consider:
1) guy makes skeleton of project under license X 2) guy collects contributions to project under license X 3) guy takes skeleton, contributions, and additional code F and produces a binary for sale.
if X == BSD guy is capable of charging contributors to use their own contributions to his product (3) unless they can deduce all contributions from 1+Sigma(2)+F [framework plus all contributors from step 2 plus code F]. Similarly all derivations of the new (3) state can be kept secret from contributors(2) etc.
if X == GPL then the distribution of project(3) is now accompanied by the source. The guy _could_ charge at least one contributor for their work, but that contributor, having protection under the Fairness doctrine is now able to re-level the system at state (3), possibly making guy irrelevant if he was dealing unfairly.
So...
BSD -> more free, because anybody can grab from the pot with no duty to fair exchange or fair dealing no matter what the subsequent circumstances.
GPL -> less free, because anybody can grab from the pot but if they want return to the audience with the results, they have to go _public_ with them.
so really
GPL => I care about the future of this no matter where it goes.
BSD => I don't care about what happens to this once it's outside of my immediate control.
or even more snarky...
GPL => I care about this work as its own, permanently public thing, and I like to have my name on it, permanently and publicly.
BSD => I care about this work as a public thing, but only when I can see the light shining on it and back up onto me, and I can live with only imagining my work hidden in all sorts of places.
Really the people who are passionate about one license and who think the other has no role, are generally one or the other sort of egoist.
Funny thing is, lots of people put something out BSD, then discover it's in some very valuable product, and only _then_ go "hay! where is my cash, or at least my applause? They _stole_ my work!" and then suffer in their misery. The GPL guys can pry their applause back out of the "offending" company.
It's just a happy accident, or perhaps a diabolical plot, that when the GPL guys get to pry into the company for "not playing fair" a whole bunch of other stuff comes out in the wash as well. Like bad DRM and cheap tricks used to rip people off by forcing them to pay $20 for $2 worth of stuff. etc.
Would you like to play a nice game of chess Professor Falcon?
(Seriously, companies in every industry are constantly exchanging employees. Every industry ends up terribly inbred. So the war is eternal and nobody can "win" it. Saying it's a Web Talent war is like "the war on [concept]". It's a kind of Professional Jingoism.)
The only real social problems I have with smoking are:
1) Second-hand smoke. Why can't somokers understand that their right to smoke stops at my lungs? i have the right to keep and bare arms, you have the right not to get gunned down at the Circle-K. Rights are arranged in a hierarchy where your right to do is naturally supposed to be limited by other's right not to do, because that is your right not to do as well.
2) WTF is the thing where smokers think its okay to toss butts on the ground or out the window of a car. Back before filters that might have been okay because the tiny scrap of tissue and bits of leaf would basically go away. Why is it that someone who wouldn't throw a candy wrapper on the ground will litter hundreds of little polyester stubs wherever they go?
I think most of the smoking thing would self resolve if there was a $5 deposit on each butt. A responsible smoker would only have to pay the $100 for the fist pack and would keep returning his butts for the next pack. The idiots would starve.
It wouldn't stop the "I have a right to blow my somke in your face no matter what because you have the right to leave" idiots though. They don't even get the I have a right to shoot at your face and you have the right to flee as a congruent falsehood.
Smoking, it seems, makes you willfully ignorant. 8-)
I'll give you all of that and more, but where do you draw the line? your porch sure. your yard, okay? the street, maybe. My yard? not without invite. the bus? You want to spend a crowded rush hour on the subway sitting there in a standing-room only trip, while some guy you don't know is bumping his tumescence in your ear every time the train lurches?
Now making some guy a sex offender because he "went streaking" for his eighteenth birthday and some eleven year old happened to see him is ridiculous. (this is happening in the news just now.)
So yea, I don't think there is a damn thing wrong with people being naked where they can be seen, if it is the casual (or even the funny) kind of naked you see as an ongoing proposition at nude beach etc.
As a social convention it becomes essentially unworkable on a daily occurrence. Do we thing have a three-tap-rule for whether a guy is scratching his nuts or fondling himself? And who is responsible for keeping count?
In simple point of fact, outside of special considerations, hygiene alone provides enough incentive to require clothing in public. Small tribes get away with being essentially naked because they are all going to be sick or well together anyway. In the scale of even a small town population it becomes immediately unworkable.
And clothes keep us from having to process every suppurating sore, hairy mole, and personal abrupt ion of ever other person we meet every moment of the day. Not just sitting on towels but body scratch then elevator button or doorknob transaction, every office wall leaned against, every bus seat and service counter.
You say a polite person brings their own towel to sit on? Sure. Might as well attach them securely to the body so you dont' have to hold them. heck, make them specially crafted and tailored towels expressly for that purpose. Then give them a special name... like "pants".
If you can honestly say that you wouldn't mind seeing every person you see each and any day, butt-naked, well more luck too you.
Clothing advances civilization because of hygiene and because we don't need to spend half our day, every day, looking for the eye-bleach and the memory wipes... 8-)
I figured people would get that part. Since the complaint is that BitTorrent (which does most of its stuff over TCP) is stepping on the VoIP, my point was that fixing the TCP and letting/making the UDP skip to the head of the line are the ways to fix the problem.
Notoriously false. Take the scripts and do the experiment by setting INTERFACE_SPEED (which is uplink) and tuning the numerator of CEILING to effect 100% or more and test against a near-by speed test site.
You can step right off the edge of the crappy upstream buffer almost instantly.
Since changing the ceiling only (effectively) selects between whether the backlog is in the router or the modem, it becomes a direct demonstration of overloading the upstream buffer.
Yes, I did the experiment extensively.
(Note that this may vary by provider, I have Comcast)
Because the guy was complaining about using BitTorrent (which uses TCP) and it's impact on his VoIP (which doesn't).
But bandwidth it bandwidth.
Plus, since the BitTorrent transfers are going to be filling your outgoing packet backlog, if you check out the scripts referenced, you will find that the VoIP packets can "pass on the curve" in the scheduler inside the linux box.
Remember that VoIP is all about _timly_ dilevery, a "late" is indestinguishable from a "drop", so all the elements have to be considered including the protocols that are raising the interference bar (BitTorrent) instead of just the protocall being used.
What most people don't understand about TCP (and therefore bittorrent etc) and Cable Modems could fill a book.
The thing most people don't understand about cable modem is that it has virtually no buffer for outbound traffic (e.g. the traffic you do control) so subsequently it is almost a given that you will overrun the transmit buffer on your cable modem doing the simplest of things. This in turn will destroy all your throughput because...
The thing most people don't understand about TCP is that it accelerates linearly and falls back exponentially. So whenever you drop an acknowledgment frame (outgoing) then your incoming data session tends to stumble to a near halt. (that is each successful frame you send increases the transmit window by one frame, but each failure cuts the transmit window in half, and most failures cause at least two drops.)
This can be seen when you use a "near by" internet speed test (a la speakeasy) and you see the speedometer surge and fall like someone revving their engine. But each "fall" is actually a bunch of trash hitting your system and then getting discarded as the stream colapses back on itself.
Now for a cable modem provider, they have no interest in throttling data coming to you to your downstream cap. That would be expensive and would just clog up the memory in their routers. your downstream limit is really implemented as an aggregate of your upstream capabilities and how their time division multiplexer is configured to cascade into statistical multiplexing. (See Comcast's "speed boost" as an example of free-wheeling and only cutting back if it must.)
So anyway, I have posted my firewall and traffic shaper scripts to my slashdot journal. They are drop-in ready for Ubuntu and Slackware, minimal editing may be needed for RedHat or others.
Try them out. Be particular to tune the top of the shaper file for the upstream speed to match your _advertised_ cable modem rate (INTERFACE_SPEED=768 in the file) and then you can fine tune the numerator part of the fraction in CEILING=... (98 worked best for me).
I get my full 8M down PLUS whatever speed boost is doing (24M down often) and my VoIP works great, even during peak usage etc, while people are gaming and web surfing in the house (my house is a high usage environment with multiple housemates skyping and gaming etc).
On top of that, the script is actually _BENEFICIAL_ to my ISP. By shaping my outgoing traffic I waste virtually zero bandwidth on retransmits so I am a great net citizen.
(If you set your firefox to enable pipelining and set max pipeline requests to a value like eighty (yes 80) you will find that you are a most efficent and therefore quite spunky web citizen.)
Not to self reply...
I put my firewall and traffic shaper scripts in my slashdot user journal so that people may share and enjoy.
Note that you will want to set some of the variables (like INTERRFACE_SPEED and the numerator in CEILING to fine-tune for best performance on your link).
Weight can also play a nice trick or two.
Note that you need both scripts, or at least all of shaper and the CLASSIFY bits of firewall to make this work right.
The firewall script is something I found on the internet and its a better companion/primer example of IPTABLES than any other script or article I have ever found. It is somewhat customized but should work "damn well" as posted.
You might want to junk the various LOG elements from the firewall if you use your console as root much.
Another super feature is that I added a ssh throttle so that if an ip address starts probing your site with random SSH attempts (more than three an hour) it will get blocked till it goes away for at least a day. Enough access for you but not a gaping hole for bots! 8-)
Share and enjoy.
I use a reasonably cheap PC setup for my boarder router (used to be just a 486) and have flawless Vonage VoIP service.
The thing you are doing wrong is that you are not _THROTTLING_ the link from your router to your cable modem.
In point of fact, and sadly, there is virtually no buffer for outgoing data on a cable modem. If you are configured for 768kbps upstream then sending data any faster than that will lead to all sorts of misery.
So in a properly configured firewall you want to throttle your _outgoing_ data to about 99% of your rated upstream bandwidth and _then_ use packet shaping to make sure that the right kind of packets get to "go first" in the QOS stack.
This turns your router into the buffer that your boundary modem lacks and will both make your VoIP flawless _AND_ _VASTLY_ improve your TCP/IP (web etc) throughput.
I have six ranks in my QOS gateway: 1) TCP ACKs (actually tcp packets less than 80 chars in length) 2) SSH (for emergencies) 3) VoIP (udp from my vonage device) 4) special occasions (none of your business 8-) 5) Games (udp in general) 6) Everything else.
Doing both of these things together will speed up everything in your house (including bittorrent) and leave you with outstanding voice quality even when gaming and running bittorrent while watching video on demand.
If found the basic rules files searching aroud the net, and then tweaked them with dynamic math and weightings.
Anybody who has ever worked in a bar, and had to clean up after a person (typically a woman) in a skirt with no underwear on has sat on a bar stool knows why public nudity is a bad thing.
The euphemism is "snail trail" and it is grotesque...
That said, I have enjoyed being naked in public at places like a nude beach where there is no common accommodation and everybody is sitting on their own private towel, and that is fine.
The reason that we don't want people screwing in the streets can be easily reduced to manners. No need for higher moral cause handed down from some all-father. _EVERYBODY_ has a list of "sexual acts" they never want to participate in, or see, or even have to acknowledge. I've never seen "two girls one cup" but the fact that I even know it exists has diminished me in some way. If I had to run the risk of seeing that happen at any street corner bistro would keep me from leaving my house.
Society, like any good fiction, knows that the more explicit sex you put in the plain text, the more you diminish your saleability and limit your audience to a niche.
Now I have no interest in barring people from going to a "sex club" (where, hopefully, they have pressure washers for daily use 8-) because in that circumstance the attendees know before hand what they can expect to see.
But in real life, the idea of all those people at the supermarket being naked, scratching their asses and then checking the tomatoes for freshness, or getting santorum all over the pop-tarts is repugnant. Not that cross-contamination never happens, but making it the expected default does not advance civilization.
So, I vote no to public nudity of the pedestrian sort, while voting yes to the public nudity of may specific sorts.
I am pretty sure you would have to be one of the protected class of people, which include a provider (...) of an interactive computer service [or someone who has been directed to act "on behalf of" same}.
So your logic only works if you also have a web page with at least one button or link on it, or if you know someone who has such a web page who told you to do stuff.
Not exactly high bar.
The real win is where you set up the web page and then direct "everyone on the internet" to remove Microsoft Word from every computer they can identify. That way you don't have to do the work and all the hackers enjoy safe harbor because they were under your lawful direction.
consider provisions of this bill "do not apply to any monitoring of, or interaction with, a subscriber's Internet or other network connection or service, or a protected computer, by or at the direction of a telecommunications carrier, cable operator, computer hardware or software provider, financial institution or provider of information services or interactive computer service..."
and "(10) detection or prevention of the unauthorized use of software fraudulent or other illegal activities."
Well clearly, as per the article they are slipping in "any enforcement we choose" actions regarding the ability of the BSA (etc) to pry into your computer with spyware like tools...
But worse, the spyware perpetrators themselves gain free immunity to all their spyware actions if they can proved they are "a provider of an information service" which, in fact, they are. They provide my information to their paying customers.
Now not only is spyware made penalty free (by accident) but Auditing Trojans that "accidentally" destroy all your data while "trying to detect" whether you have stolen Barbie's Big Adventure
The corporations, both legal and illegal, now own your computer in every way that matters.
I should have said "soviet socialism". I hoist myself on my on petard there a little, but since the discussion of "soviet socialism" is distinct from genera socialism in many ways, I fell victim to the shorthand of calling it just "soviet" which makes sense in my circles but may not make sense to a more general audience.
For instance "communist china" is a socialist state, but while it is chock full of committees, they are not arranged in the soviet system of committees. But now I mince already over-fine hairs.
It was _neither_ fascism nor dictatorship. When they called it the United Soviet Socialist Republic (USSR) they named what they got with perfect accuracy.
You are correct that they never had Comunisim, in fact no country on earth has _ever_ had comunisim in government as, among other things, there is no identifiable government, indeed no "state" at all, when you are in a communist state.
It's like that magical point where Libertarianism and Anarchy intersect, a point defined only in terms of irrational numbers.
But back on point, the soviet state was the planned precursor to communism. It was the eternal temporary precursor to that time when people, being pure of thought and absent of motive, would automatically do the shit work because it was the right thing to do, freeing their peers to pursue the higher pursuits of man.
Unfortunately, since there were people involved, and everybody wants to be doing the "higher pursuits" stuff "the peoples revolution" in all its forms, always stalls before the soviets (e.g. the organizing transitional committees) can achieves the perfection which would allow them to declare themselves irrelevant and so disband.
Fascism is instead a dictum of compelled uniform adherence. Named for the bundle of meager twigs that cannot be broken (e.g. the arrows on the great seal of the united states is "a fasc" as there is quite a lot of fascism inherent in our system of representative democracy.
The fact that there was a polit-buro alone disproves Dictatorship, which is by definition the self-appointed-kingship, where all authority and relief comes from the decisions of one individual who rules by strength.
Further, since there is no money or means of exchange, in a communist system there is literally no revenue from any industry to flow anywhere. Each pursuit is considered industry of immeasurable value, and the results of that industry are freely self-distributed as yet more industry in a magical and undepleteable pool of individuals invariably doing the most good for the most people individually and collectively.
Communism has never been implemented anywhere as it can not exist where there is even the slightest whiff of self interest or vice. In short it cannot exist, by definition, where people are involved. Hell, it turns out it doesn't exist in beehives and anthills either.
So your first sentence is correct, but the second and subsequent fail utterly as misinformation or dangerous misinterpretation.
As a completely non-aggressive response, our servers should engage in an "internet shunning" of MediaDefender.
Every Linux admin on the planet should put TARPIT (or at least DROP) rules in their firewalls for any address range that comes from MediaDefender.
I'm surprised that most carriers (Sprint Net, AT&T, Comcast etc) don't do this just to protect themselves from those 9Gbps DOS attacks that come down that link.
TARPIT rules are the ultimate "you are not welcome here". They don't have to come to my site, and if they do my site will put them on hold.
It wouldn't stop a SYN flood, but it would stop them from being able to poison your trackers in the first place.
DROPS are second best, of course, but better to put someone on indefinite-hold than hang up on them, especially when both have the same cost for you.
Does anybody have the MediaDefender IP address range available to post?
It is just about time for the entire internet, or at least loving server operators everywhere, to add TARPIT or at least DROP rules to their firewalls for every address range that can be traced to Media Defender.
Remember, in shunning an internet wrongdoer, the TARPIT is the better option. You can configure your server to feel almost no impact, and their server gets actual resources pinned down and "used up" for nontrivial periods of time.
I would think that any ISP who doesn't want to cary the _burden_ of forwarding 9Gbps of DOS for this company would want to do the same.
Does anybody know the IP addresses that Media Defender uses to do these DOS attacks?
And if you are shunning, you should include the corporate addresses and false torrent seeder addresses as well.
The fact that the comercial grade tools you are used to cost you a lot of money doesn't speak to the base cost of the minimum necessary tools to hack your system.
A single PCI/PCIe/etc "hobbiest" grade board with an FPGA on can be trivially programmed to snapshot running memory to an external serial interface (where another computer will record it) triggered by an entirely real-world event (watch the birdie while I push this button that shorts these two pins).
The fact of the matter is that "back in the day" these sorts of debuggers were a dime a dozen because they were needed. Now days they are rarely needed, and when they are needed they need to be subtle, so they are pricey.
As soon as this TPM buffoonery, or anything like it, becomes common place super obvious and cheap versions of this sort of thing will be back.
Hell, on top of all that there is the recent work that tells us that volatile RAM turns out not to be that volatile after all and you can practically "just reboot" to get a snapshot of memory if you prevent the POST. Hardware doesn't get much cheaper than "what you already have".
Get over your delusions. Whatever your software is, it isn't secure from the guy with the screwdriver standing over the case, and it _NEVER_ will be.
You best _PRAY_ that your companies busineess model doesn't _DEPEND_ on TPM, or at least you should keep your Resume up to date.
What this sort of thing _can_ do for comcast (etc) is keep the bandwidth usage from crossing one of their (toll road) borders.
Back when "the internet was monitized" and companies started charging for everything by "actual us" as opposed to "bandwidth promised" these super-smart companies discovered that they had footbulleted themselves. Now they pay X to get an OC12 but then then pay Y to actually transmit data over that link. E.g. live by the meeter, die by the meeter.
So if comcast can get you to use _their_ P2P (or VOIP etc) then they can make sure that they keep their traffic on their segments. Your tracker will suddenly only connect you to other peers on the same provider, and Comcast saves bunches.
Of course by balkinizing the network, the number of peers is diminished and the likely lifetime of a swarm declines faster. So they can also terminate offerings once they are no longer popular "enough" in the mind of the sysadmin in question.
That and since it is in-network you will start to see it looking like a win once net neutrality is killed. You know "see here, on the improved internet, local is better! by our sponsors product! be happy and consume!" etc.
So yea, in terms of green folding overhead, there is a lot less to be paid if you get your P2P from your self-approved value-added supplier.
Slashdot Mirror
Its not a matter of the joke being beaten to death its more a matter of the implementation strangling performance.
Along with fix-a-flat there will be ten-mile fuel cell quick jump kits for your car to get you to your next plug in.
What would be an even better (and just as useful) measure would be to run inductive chargers under the pavement on long stretches of the interstate system where you charge as you ride, or largely ride for free. Or (if we _must_) you have a Charge Pass(tm) meter to get billed for charge as you go.
Make fun of the charge-for-free if you like, but it is likely cheaper to give the passers-through the free ride than to clean up the mess of the fuel economy (plus billing etc). Imagine how well the power systems would be regulated if the fed were footing the bill for the electricity delivered to the interstate highway system.
Regardless of that sky-pie...
Since there is no inherent danger to making electricity available, as opposed to the inherent danger of making petrochemical distillates available, getting a charge as part of parking at your rest stop or attraction (or while parked at work because of your employers clean-air program) could delete the entire concept of the dedicated fueling station.
Go to Denies, or MacDonalds, or Taco Time, and get a (full or partial) charge while you eat.
Sound unlikely? Well in colder climates they already rig their parking lots with plugs to keep your engine block from freezing while you shop or eat.
Hell, having reduced-price or even free charging at the local highway rest-stop would cut way down on tired long-distance driving accidents. Pull over, take a rest, get a charge on us 'cause it's cheaper than having to dispatch Life Flight etc.
Once the "fuel" is electricity, which is literally everywhere in our society, we lose the need for the "filling station" to be a super-defined super-regulated contentious resource. Anybody can sell "their" electricity if they have the amperage to spare.
Once motive force stops being "rare and dangerous" gasoline, the paradigm totally goes to distributed recharging as the most sensible practice.
The democrats (a class of which I am nominally a member) can not seem to figure out that "running to the middle" is a loser strategy. It just turns you into a nobody.
Since nobody completely agrees with anybody, everybody will vote for the somebody before they vote for the nobody.
For every supporter you pander to, you lose two supporters you already have. Doesn't matter what side you are coming from.
In short, at the center of the pool, you will find a drain
I would agree, but I would refocus just a touch.
- BSD ensures that the *next* recipient of the code may do what they want. ...conversely...
- GPL ensures that *every* recipient of the code may do what they want.
- BSD ensures that the *publisher* of the code can take received contributions to the code and make them part of a private domain if he chooses.
- GPL ensures that each *contributor* to the code can ensure that the publisher cannot incorporate his contribution into a private domain.
I would say that BSD code is far less encumbered than GPL code. The BSD license is effectively a release to the public domain. Given that you have to leave the copyright in place but you don't ever have to give out the code, makes the provisions of the BSD as toothless as a declaration of abandonment for downstream use. The BSD does however require symmetric abandonment by contributors moving patches upstream.
In contrast the GPL requires democratic fairness of all parties. Fair isn't Free, just as "freedom isn't without cost".
Consider:
1) guy makes skeleton of project under license X
2) guy collects contributions to project under license X
3) guy takes skeleton, contributions, and additional code F and produces a binary for sale.
if X == BSD guy is capable of charging contributors to use their own contributions to his product (3) unless they can deduce all contributions from 1+Sigma(2)+F [framework plus all contributors from step 2 plus code F]. Similarly all derivations of the new (3) state can be kept secret from contributors(2) etc.
if X == GPL then the distribution of project(3) is now accompanied by the source. The guy _could_ charge at least one contributor for their work, but that contributor, having protection under the Fairness doctrine is now able to re-level the system at state (3), possibly making guy irrelevant if he was dealing unfairly.
So...
BSD -> more free, because anybody can grab from the pot with no duty to fair exchange or fair dealing no matter what the subsequent circumstances.
GPL -> less free, because anybody can grab from the pot but if they want return to the audience with the results, they have to go _public_ with them.
so really
GPL => I care about the future of this no matter where it goes.
BSD => I don't care about what happens to this once it's outside of my immediate control.
or even more snarky...
GPL => I care about this work as its own, permanently public thing, and I like to have my name on it, permanently and publicly.
BSD => I care about this work as a public thing, but only when I can see the light shining on it and back up onto me, and I can live with only imagining my work hidden in all sorts of places.
Really the people who are passionate about one license and who think the other has no role, are generally one or the other sort of egoist.
Funny thing is, lots of people put something out BSD, then discover it's in some very valuable product, and only _then_ go "hay! where is my cash, or at least my applause? They _stole_ my work!" and then suffer in their misery. The GPL guys can pry their applause back out of the "offending" company.
It's just a happy accident, or perhaps a diabolical plot, that when the GPL guys get to pry into the company for "not playing fair" a whole bunch of other stuff comes out in the wash as well. Like bad DRM and cheap tricks used to rip people off by forcing them to pay $20 for $2 worth of stuff. etc.
It's just a minor screed about how everything is "a war" even when it isn't anything like a war at all.
The original article depicts people moving between similar companies as a "Web 2.0 Talent War".
The Parent tries to "infiltrate" a third company.
I make a "War Games" movie reference and then follow it with a serious observation that not all conflicting interests constitute a war.
You have to consider it as a whole...
Ah, more screed on the death of context... 8-)
... is not to play.
Would you like to play a nice game of chess Professor Falcon?
(Seriously, companies in every industry are constantly exchanging employees. Every industry ends up terribly inbred. So the war is eternal and nobody can "win" it. Saying it's a Web Talent war is like "the war on [concept]". It's a kind of Professional Jingoism.)
The only real social problems I have with smoking are:
1) Second-hand smoke. Why can't somokers understand that their right to smoke stops at my lungs? i have the right to keep and bare arms, you have the right not to get gunned down at the Circle-K. Rights are arranged in a hierarchy where your right to do is naturally supposed to be limited by other's right not to do, because that is your right not to do as well.
2) WTF is the thing where smokers think its okay to toss butts on the ground or out the window of a car. Back before filters that might have been okay because the tiny scrap of tissue and bits of leaf would basically go away. Why is it that someone who wouldn't throw a candy wrapper on the ground will litter hundreds of little polyester stubs wherever they go?
I think most of the smoking thing would self resolve if there was a $5 deposit on each butt. A responsible smoker would only have to pay the $100 for the fist pack and would keep returning his butts for the next pack. The idiots would starve.
It wouldn't stop the "I have a right to blow my somke in your face no matter what because you have the right to leave" idiots though. They don't even get the I have a right to shoot at your face and you have the right to flee as a congruent falsehood.
Smoking, it seems, makes you willfully ignorant. 8-)
I'll give you all of that and more, but where do you draw the line? your porch sure. your yard, okay? the street, maybe. My yard? not without invite. the bus? You want to spend a crowded rush hour on the subway sitting there in a standing-room only trip, while some guy you don't know is bumping his tumescence in your ear every time the train lurches?
Now making some guy a sex offender because he "went streaking" for his eighteenth birthday and some eleven year old happened to see him is ridiculous. (this is happening in the news just now.)
So yea, I don't think there is a damn thing wrong with people being naked where they can be seen, if it is the casual (or even the funny) kind of naked you see as an ongoing proposition at nude beach etc.
As a social convention it becomes essentially unworkable on a daily occurrence. Do we thing have a three-tap-rule for whether a guy is scratching his nuts or fondling himself? And who is responsible for keeping count?
In simple point of fact, outside of special considerations, hygiene alone provides enough incentive to require clothing in public. Small tribes get away with being essentially naked because they are all going to be sick or well together anyway. In the scale of even a small town population it becomes immediately unworkable.
And clothes keep us from having to process every suppurating sore, hairy mole, and personal abrupt ion of ever other person we meet every moment of the day. Not just sitting on towels but body scratch then elevator button or doorknob transaction, every office wall leaned against, every bus seat and service counter.
You say a polite person brings their own towel to sit on? Sure. Might as well attach them securely to the body so you dont' have to hold them. heck, make them specially crafted and tailored towels expressly for that purpose. Then give them a special name... like "pants".
If you can honestly say that you wouldn't mind seeing every person you see each and any day, butt-naked, well more luck too you.
Clothing advances civilization because of hygiene and because we don't need to spend half our day, every day, looking for the eye-bleach and the memory wipes... 8-)
I figured people would get that part. Since the complaint is that BitTorrent (which does most of its stuff over TCP) is stepping on the VoIP, my point was that fixing the TCP and letting/making the UDP skip to the head of the line are the ways to fix the problem.
Notoriously false. Take the scripts and do the experiment by setting INTERFACE_SPEED (which is uplink) and tuning the numerator of CEILING to effect 100% or more and test against a near-by speed test site.
You can step right off the edge of the crappy upstream buffer almost instantly.
Since changing the ceiling only (effectively) selects between whether the backlog is in the router or the modem, it becomes a direct demonstration of overloading the upstream buffer.
Yes, I did the experiment extensively.
(Note that this may vary by provider, I have Comcast)
Interesting... It didn't ever "reject" the number I used. I never looked to see if it was just capping or rounding down the value. /doh! 8-)
(e.g. you put in 80 and it seems fine, now I'll have to find out if its being clipped or ignored or what... 8-)
Because the guy was complaining about using BitTorrent (which uses TCP) and it's impact on his VoIP (which doesn't).
But bandwidth it bandwidth.
Plus, since the BitTorrent transfers are going to be filling your outgoing packet backlog, if you check out the scripts referenced, you will find that the VoIP packets can "pass on the curve" in the scheduler inside the linux box.
Remember that VoIP is all about _timly_ dilevery, a "late" is indestinguishable from a "drop", so all the elements have to be considered including the protocols that are raising the interference bar (BitTorrent) instead of just the protocall being used.
It's called "Systems Theory"... 8-)
What most people don't understand about TCP (and therefore bittorrent etc) and Cable Modems could fill a book.
The thing most people don't understand about cable modem is that it has virtually no buffer for outbound traffic (e.g. the traffic you do control) so subsequently it is almost a given that you will overrun the transmit buffer on your cable modem doing the simplest of things. This in turn will destroy all your throughput because...
The thing most people don't understand about TCP is that it accelerates linearly and falls back exponentially. So whenever you drop an acknowledgment frame (outgoing) then your incoming data session tends to stumble to a near halt. (that is each successful frame you send increases the transmit window by one frame, but each failure cuts the transmit window in half, and most failures cause at least two drops.)
This can be seen when you use a "near by" internet speed test (a la speakeasy) and you see the speedometer surge and fall like someone revving their engine. But each "fall" is actually a bunch of trash hitting your system and then getting discarded as the stream colapses back on itself.
Now for a cable modem provider, they have no interest in throttling data coming to you to your downstream cap. That would be expensive and would just clog up the memory in their routers. your downstream limit is really implemented as an aggregate of your upstream capabilities and how their time division multiplexer is configured to cascade into statistical multiplexing. (See Comcast's "speed boost" as an example of free-wheeling and only cutting back if it must.)
So anyway, I have posted my firewall and traffic shaper scripts to my slashdot journal. They are drop-in ready for Ubuntu and Slackware, minimal editing may be needed for RedHat or others.
Try them out. Be particular to tune the top of the shaper file for the upstream speed to match your _advertised_ cable modem rate (INTERFACE_SPEED=768 in the file) and then you can fine tune the numerator part of the fraction in CEILING=... (98 worked best for me).
I get my full 8M down PLUS whatever speed boost is doing (24M down often) and my VoIP works great, even during peak usage etc, while people are gaming and web surfing in the house (my house is a high usage environment with multiple housemates skyping and gaming etc).
On top of that, the script is actually _BENEFICIAL_ to my ISP. By shaping my outgoing traffic I waste virtually zero bandwidth on retransmits so I am a great net citizen.
(If you set your firefox to enable pipelining and set max pipeline requests to a value like eighty (yes 80) you will find that you are a most efficent and therefore quite spunky web citizen.)
Share and Enjoy...
Not to self reply... I put my firewall and traffic shaper scripts in my slashdot user journal so that people may share and enjoy. Note that you will want to set some of the variables (like INTERRFACE_SPEED and the numerator in CEILING to fine-tune for best performance on your link). Weight can also play a nice trick or two. Note that you need both scripts, or at least all of shaper and the CLASSIFY bits of firewall to make this work right. The firewall script is something I found on the internet and its a better companion/primer example of IPTABLES than any other script or article I have ever found. It is somewhat customized but should work "damn well" as posted. You might want to junk the various LOG elements from the firewall if you use your console as root much. Another super feature is that I added a ssh throttle so that if an ip address starts probing your site with random SSH attempts (more than three an hour) it will get blocked till it goes away for at least a day. Enough access for you but not a gaping hole for bots! 8-) Share and enjoy.
I use a reasonably cheap PC setup for my boarder router (used to be just a 486) and have flawless Vonage VoIP service.
The thing you are doing wrong is that you are not _THROTTLING_ the link from your router to your cable modem.
In point of fact, and sadly, there is virtually no buffer for outgoing data on a cable modem. If you are configured for 768kbps upstream then sending data any faster than that will lead to all sorts of misery.
So in a properly configured firewall you want to throttle your _outgoing_ data to about 99% of your rated upstream bandwidth and _then_ use packet shaping to make sure that the right kind of packets get to "go first" in the QOS stack.
This turns your router into the buffer that your boundary modem lacks and will both make your VoIP flawless _AND_ _VASTLY_ improve your TCP/IP (web etc) throughput.
I have six ranks in my QOS gateway:
1) TCP ACKs (actually tcp packets less than 80 chars in length)
2) SSH (for emergencies)
3) VoIP (udp from my vonage device)
4) special occasions (none of your business 8-)
5) Games (udp in general)
6) Everything else.
Doing both of these things together will speed up everything in your house (including bittorrent) and leave you with outstanding voice quality even when gaming and running bittorrent while watching video on demand.
If found the basic rules files searching aroud the net, and then tweaked them with dynamic math and weightings.
Flawless.
Anybody who has ever worked in a bar, and had to clean up after a person (typically a woman) in a skirt with no underwear on has sat on a bar stool knows why public nudity is a bad thing.
The euphemism is "snail trail" and it is grotesque...
That said, I have enjoyed being naked in public at places like a nude beach where there is no common accommodation and everybody is sitting on their own private towel, and that is fine.
The reason that we don't want people screwing in the streets can be easily reduced to manners. No need for higher moral cause handed down from some all-father. _EVERYBODY_ has a list of "sexual acts" they never want to participate in, or see, or even have to acknowledge. I've never seen "two girls one cup" but the fact that I even know it exists has diminished me in some way. If I had to run the risk of seeing that happen at any street corner bistro would keep me from leaving my house.
Society, like any good fiction, knows that the more explicit sex you put in the plain text, the more you diminish your saleability and limit your audience to a niche.
Now I have no interest in barring people from going to a "sex club" (where, hopefully, they have pressure washers for daily use 8-) because in that circumstance the attendees know before hand what they can expect to see.
But in real life, the idea of all those people at the supermarket being naked, scratching their asses and then checking the tomatoes for freshness, or getting santorum all over the pop-tarts is repugnant. Not that cross-contamination never happens, but making it the expected default does not advance civilization.
So, I vote no to public nudity of the pedestrian sort, while voting yes to the public nudity of may specific sorts.
I am pretty sure you would have to be one of the protected class of people, which include a provider (...) of an interactive computer service [or someone who has been directed to act "on behalf of" same}.
So your logic only works if you also have a web page with at least one button or link on it, or if you know someone who has such a web page who told you to do stuff.
Not exactly high bar.
The real win is where you set up the web page and then direct "everyone on the internet" to remove Microsoft Word from every computer they can identify. That way you don't have to do the work and all the hackers enjoy safe harbor because they were under your lawful direction.
IANAL, but it says that very thing in plain text.
consider provisions of this bill "do not apply to any monitoring of, or interaction with, a subscriber's Internet or other network connection or service, or a protected computer, by or at the direction of a telecommunications carrier, cable operator, computer hardware or software provider, financial institution or provider of information services or interactive computer service..."
and "(10) detection or prevention of the unauthorized use of software fraudulent or other illegal activities."
Well clearly, as per the article they are slipping in "any enforcement we choose" actions regarding the ability of the BSA (etc) to pry into your computer with spyware like tools...
But worse, the spyware perpetrators themselves gain free immunity to all their spyware actions if they can proved they are "a provider of an information service" which, in fact, they are. They provide my information to their paying customers.
Now not only is spyware made penalty free (by accident) but Auditing Trojans that "accidentally" destroy all your data while "trying to detect" whether you have stolen Barbie's Big Adventure
The corporations, both legal and illegal, now own your computer in every way that matters.
Ta Da!
Why should they have to comment their code, let alone comment it well? HUGE amounts of open source code is atrociously commented if at all...
There is no need to hold Nokia to a higher standard... that is just mean.
I should have said "soviet socialism". I hoist myself on my on petard there a little, but since the discussion of "soviet socialism" is distinct from genera socialism in many ways, I fell victim to the shorthand of calling it just "soviet" which makes sense in my circles but may not make sense to a more general audience.
For instance "communist china" is a socialist state, but while it is chock full of committees, they are not arranged in the soviet system of committees. But now I mince already over-fine hairs.
It was _neither_ fascism nor dictatorship. When they called it the United Soviet Socialist Republic (USSR) they named what they got with perfect accuracy.
You are correct that they never had Comunisim, in fact no country on earth has _ever_ had comunisim in government as, among other things, there is no identifiable government, indeed no "state" at all, when you are in a communist state.
It's like that magical point where Libertarianism and Anarchy intersect, a point defined only in terms of irrational numbers.
But back on point, the soviet state was the planned precursor to communism. It was the eternal temporary precursor to that time when people, being pure of thought and absent of motive, would automatically do the shit work because it was the right thing to do, freeing their peers to pursue the higher pursuits of man.
Unfortunately, since there were people involved, and everybody wants to be doing the "higher pursuits" stuff "the peoples revolution" in all its forms, always stalls before the soviets (e.g. the organizing transitional committees) can achieves the perfection which would allow them to declare themselves irrelevant and so disband.
Fascism is instead a dictum of compelled uniform adherence. Named for the bundle of meager twigs that cannot be broken (e.g. the arrows on the great seal of the united states is "a fasc" as there is quite a lot of fascism inherent in our system of representative democracy.
The fact that there was a polit-buro alone disproves Dictatorship, which is by definition the self-appointed-kingship, where all authority and relief comes from the decisions of one individual who rules by strength.
Further, since there is no money or means of exchange, in a communist system there is literally no revenue from any industry to flow anywhere. Each pursuit is considered industry of immeasurable value, and the results of that industry are freely self-distributed as yet more industry in a magical and undepleteable pool of individuals invariably doing the most good for the most people individually and collectively.
Communism has never been implemented anywhere as it can not exist where there is even the slightest whiff of self interest or vice. In short it cannot exist, by definition, where people are involved. Hell, it turns out it doesn't exist in beehives and anthills either.
So your first sentence is correct, but the second and subsequent fail utterly as misinformation or dangerous misinterpretation.
As a completely non-aggressive response, our servers should engage in an "internet shunning" of MediaDefender.
Every Linux admin on the planet should put TARPIT (or at least DROP) rules in their firewalls for any address range that comes from MediaDefender.
I'm surprised that most carriers (Sprint Net, AT&T, Comcast etc) don't do this just to protect themselves from those 9Gbps DOS attacks that come down that link.
TARPIT rules are the ultimate "you are not welcome here". They don't have to come to my site, and if they do my site will put them on hold.
It wouldn't stop a SYN flood, but it would stop them from being able to poison your trackers in the first place.
DROPS are second best, of course, but better to put someone on indefinite-hold than hang up on them, especially when both have the same cost for you.
Does anybody have the MediaDefender IP address range available to post?
It is just about time for the entire internet, or at least loving server operators everywhere, to add TARPIT or at least DROP rules to their firewalls for every address range that can be traced to Media Defender.
Remember, in shunning an internet wrongdoer, the TARPIT is the better option. You can configure your server to feel almost no impact, and their server gets actual resources pinned down and "used up" for nontrivial periods of time.
I would think that any ISP who doesn't want to cary the _burden_ of forwarding 9Gbps of DOS for this company would want to do the same.
Does anybody know the IP addresses that Media Defender uses to do these DOS attacks?
And if you are shunning, you should include the corporate addresses and false torrent seeder addresses as well.
The fact that the comercial grade tools you are used to cost you a lot of money doesn't speak to the base cost of the minimum necessary tools to hack your system.
A single PCI/PCIe/etc "hobbiest" grade board with an FPGA on can be trivially programmed to snapshot running memory to an external serial interface (where another computer will record it) triggered by an entirely real-world event (watch the birdie while I push this button that shorts these two pins).
The fact of the matter is that "back in the day" these sorts of debuggers were a dime a dozen because they were needed. Now days they are rarely needed, and when they are needed they need to be subtle, so they are pricey.
As soon as this TPM buffoonery, or anything like it, becomes common place super obvious and cheap versions of this sort of thing will be back.
Hell, on top of all that there is the recent work that tells us that volatile RAM turns out not to be that volatile after all and you can practically "just reboot" to get a snapshot of memory if you prevent the POST. Hardware doesn't get much cheaper than "what you already have".
Get over your delusions. Whatever your software is, it isn't secure from the guy with the screwdriver standing over the case, and it _NEVER_ will be.
You best _PRAY_ that your companies busineess model doesn't _DEPEND_ on TPM, or at least you should keep your Resume up to date.
What this sort of thing _can_ do for comcast (etc) is keep the bandwidth usage from crossing one of their (toll road) borders.
Back when "the internet was monitized" and companies started charging for everything by "actual us" as opposed to "bandwidth promised" these super-smart companies discovered that they had footbulleted themselves. Now they pay X to get an OC12 but then then pay Y to actually transmit data over that link. E.g. live by the meeter, die by the meeter.
So if comcast can get you to use _their_ P2P (or VOIP etc) then they can make sure that they keep their traffic on their segments. Your tracker will suddenly only connect you to other peers on the same provider, and Comcast saves bunches.
Of course by balkinizing the network, the number of peers is diminished and the likely lifetime of a swarm declines faster. So they can also terminate offerings once they are no longer popular "enough" in the mind of the sysadmin in question.
That and since it is in-network you will start to see it looking like a win once net neutrality is killed. You know "see here, on the improved internet, local is better! by our sponsors product! be happy and consume!" etc.
So yea, in terms of green folding overhead, there is a lot less to be paid if you get your P2P from your self-approved value-added supplier.