If you are doing large datasets, you can end up trippling your filesize easily. Think about this, if you have 90 datapoints recorded 100 times a second, that's 9K+ sets of tags that have to be decoded every second with at least 7 extra characters per tag set. That doesn't include the performance hit for encoding & decoding the datastructure. <data> <time stamp=[timestamp]> <sensor number=1>5.00</sensor> <sensor number=2>5.01</sensor> ... <sensor number=90>5.00</sensor> </time> <time stamp=[timestamp+1]> ... </data>
For large, structured datadumps like that, CSV can actually be a faster way to work. A hybrid works even better to make a more flexible system with some of the best features of both.
<format> <interval> <unit>Second</unit> <value>0.01</value> </interval> <field name='Sensor1' position=1/> ... <field name='Sensor90' position=90/> </format> <CSV_Data> 1,2,3,4,5,.. .,90 1,2,3,5,4,...,90 ... </CSV_Data> It uses XML to define the data format, but at the same time, it doesn't require a lot of processing to decode the actual dataset. Also the file size is minimally impacted.
If there is a formal request made to MS to present the specific IP to the community so that it can be removed, and MS fails to do so, then they can loose the right to persue the claims in court. In short, once they are legally asked to put up or shut up, failure to put up makes all their - 'we'll sue' FUD worthless.
The court and jury disagree with you. They found M$ guilty and imposed a monster fine which reflected their outrage and the magnitude of the offense.
And a jury found a substitute-teacher guilty of 'endangering minors' because of a porn storm on her PC. Juries base thier judgment on the information presented to them, and their opinion of the people involved --- "He's a good looking man & he would never have to rape someone" was the actual reasoning given by a juror on why she wouldn't vote guilty.
MS, Apple, Creative, and RCA all licence MP3 technology through a different company. Alcatel recently merged with another company and acquired certain patents, one of which is this MP3 patent jointly held between Alcatel & another (IIRC) French company. Everyone pays the other company, nobody was paying the company that Alcatel just bought out. So this isn't about MS not paying for the patent, this is another company popping up after the fact & saying 'gimmee too'. The reality of this is that either everyone is violating Alcatel's patents, the patents are invalid, or nobody who is paying this other company is violating the patent. Up until this ruling, everyone selling MP3 players/software was firmly in the 3rd camp.
Now, as much as I dispise MS, I really don't think they were violating this companies IP. They purchased the rights to the software through the company recognized by the industry as the proper licensor.
Updating the wikipedia entry with information you know to be false is vandalism. It's not a question of wikipedia not liking it. It's an act by an individual(s) to forcibly alter the content of the page in a manner not consistant with reality.
I do consider things like government officials having their staff daily editing their entries to post their 'stand' on an issue, and deleting all the evidence to the contrary, to be vandalism. The same as I consider someone adding these false claims to Fuzzy's entry to be vandalism.
These are not the same as people posting incorrect dates or honestly confusing 2 the 2 'Howard Sterns' in the news right now (One a radio DJ & the other a lawyer). This particular instance appears to be a malicious addition to his entry for whatever reason. That makes it both libel, and vandalism.
Actually, you can make decisions to not squeaze every penny out of a product through persuing lawsuits etc.
Lawsuits are not cheap on several fronts. First there is the actual cost - from additional document retention right up through the actual cost of the lawyers themselves. The other cost is goodwill. It's a lot less tangible, but ask SCO how it feels to sue your customers. Or perhaps ask people their opinion of the MPAA & RIAA.
There should be no board of directors in the world that will look at a case of an individual trading a few copies of their songs & going after them for copyright infringment. The loss of goodwill and the legal fees would far outway any possible positive result from the suit. The RIAA & MPAA however insulate the actual board of directors for their respective cartels. Because they do not have to cope with the loss of goodwill, the total cost of suing the consumers of their product is reduced to a level that they find acceptable.
The key factor in showing it to be a scare campain is that they are not likewise trumpeting their successes in catching pirate copies of pressed CDs.
Do we really want to let people loose on Linux who can't [be bothered to] install it themselves?
Do we really want to let people loose on Windows who can't be bothered to secure it themselves?
1 in 4 Windows boxes is currently part of a botnet. Hand out Linux boxes instead of Windows ones and at least we can kill off the bots for the "Email & Web browsing" crowd (about 1/3 of current residential users).
Question from someone who's clueless about the intricacies of the internet: Is it possible to get your data better treatment under such a protocol by falsely labeling it? Or would that be pointless since no priority ruleset would be strictly better than another?
It would really depend on what type of data you are discussing, what type of protocols are involved, and the type of examination going on. The answers to the first 2 questions also determine whether or not it's going to actually improve your perceived performance.
[internet 101]
The TCP/IP stack takes a stream of data and breakes it down into 1 or more individual packets. Each packet under the current IPv4 is wrapped in a header that contains a minimum of:
a sequence number
a destination
a source
a protocol
a port
There are some other items in there also, but those are the major ones.
When a packet is sent using TCP [Traffic Control Protocol], the IP [Internet Protocol] stack at the receiving end takes care of re-assembling the packets into the correct order & if one is missing, requesting that it be re-transmitted. Some protocols - ftp, http, bittorrent, certain buffered streamed media - which assemble larger files from these small packets are not significantly effected by a high latency transmission. If packet 2 arrives after packet 3, the stack puts them in the right order and the end user is none the wiser. These protocols can be routed to sub-optimal paths and as long as the packet loss/retransmission rate is kept under about 10% the end user doesn't notice any change.
Other types of protocols are sensative to sequence, latency, and lost packets. These include unbuffered streaming media (audio and video), VOIP, and real time games. These normally use a packet type called UDP [User Datagram Protocol]. This service does not allow for retransmission, nor does it correct for improper sequencing. If packet 2 is missing, packet 3 is decoded & implimented without reguard for whatever was in 2. If packet 6 comes in after packet 7, it's thrown away.
As you can see, this type of transfer is not acceptable for file transfers since a single missing packet will corrupt the file. However, this type of data transfer is required for real-time applications. If you are talking to someone using VOIP, a missing packet results in a short moment of silence - usually not long enough to actually be heard, just to degrade the overall quality. However, requesting that that missing packet be re-transmitted, would actually create a longer pause.
[/internet 101]
[QoS]
During the actual transmission of the data packets, they are handed from one router to another - often from one carrier to another. If one of the routers becomes busy, it begins queueing the packets. When it's buffer becomes full, it starts dropping packets into the bit bucket. Currently the first type of packet to be dropped is the ICMP [Internet Connection Maintenance Protocol] set - these include the trace & ping packets. After that, packets are dropped on a first come first served basis. If packet R1[receive 1] hits the router while the buffer is full, it's dropped. If packet R2 [receive 2] hits a milisecond later - just after packet T1[transmission 1] was transmitted & cleared from the queue, it's accepted. Packet C would then be rejected as the buffer is again full.
That is net neutrality at it's most basic. All of the data packets from all of the protocols are handled identically. Recognizing that this is not nescessarily the best method, QoS was introduced as a means of handling the different latency requirements of the different protocols. In a basic system, a QoS aware router examines the protocol flag & buffers the data in 2 seperate queues - one time sensative, one less sensative. Data is then transmitted according to a priority scheme - 2:1, 11:9, etc - which allows the time sensative data to be transmitted through faster & with le
I really don't want to keep finding framegrabs of Prince playing the guitar at the Superbowl when I go searching for my daily dose of PRON. (Yes, people were complaining that Prince, was [looks both ways before whispering] sexual during his performance & his shadow looked like a big penis)
Ohh, I bet we can work up a log to post to the internet showing some congressional connections to that child porn bust. Few MAC addresses, the block of IP addresses for their ISP, a disclaimer showing that it's only an example of the possible abuse of this type of legislation. Oh, a nice bit of IP address spoofing/trumping to show it happening live - that would be good.
Last time this came up, it was estimated to cost over $400M/year. The estimated number of arrests it would help generate? 700. The FBI said just give them the $400M for agents & they could do a hell of a lot better. The truth is that the 60-90 day cycle that most of these companies already have is enough to cover the vast majority of the requests by the police - this is asking the industry to absorb $400M in costs for an infitesimal gain.
Funny the AG didn't want to do that... guess it didn't sound as good politically.
You lease politicians. You can never really buy them. You lease them for a few votes then return them back to their parties - just like a leased car going back to it's dealership.
The only reason the service might not be sufficient is if the ISP advertised more bandwidth than they really have.
Charter did this in some places in Western Ma. Parts of Chicopee & Holyoke had problems with up to 80% packet loss every night between 6-9PM as everyone got home & tried to get on the internet at the same time. The UBR's didn't have the bandwidth to cover the collections of modems attached to them.
The free market only works if there is a 'market' to work with. When it's a single provider, or providers working as a cartel, the free market cannot produce market corrections.
I had an AT&T fiber line run past my house close enough to spit on from my porch. It was 3 years before I could get broadband from them. It was 4 years before Verizon even offered DSL to me (768 is still the fastest they offer). Oh, and Direct Way satallite service is out too, I have a hill between me & the sat. That means that I had a 3 year wait between Broadband being offered in town center (5 blocks away) and getting it at my house, and another year before I had an option - 1/4 mile down the road there's still only 1 provider, Comcast.
So for the vast majority of the people in CN & USA, there is no significant alternative provider to convert to. Thus - there is no free market.
Net neutrality is the opposite of that. It dictates that all traffic must be treated equally.
Actually it's a bit more refined than that. Net neutrality dictates that all similar traffic must be treated equally.
All HTTP requests have to be delivered equivalently reguardless of source or destination
All VOIP requests have to be....
All torrent requests have to be....
....
However, VOIP & HTTP requests can be routed with different priorities - VOIP is sensative to lag, HTTP isn't.
The concept of traffic shaping is to provide a QoS [Quality of Service]flagged route for packets which maximizes the use of the fastest, cleanest routes for lag/packet loss sensative protocols, while relegating less sensative packets to routes which may not be as responsive. The Telco extention[perversion] of packet shaping is to convert the selection criteria from protocol needs to accounting balance. Thus some of the Canadian telcos have already started to throttle Vonage service to the point of compromising service quality - remarkably just before they roll out their own service which doesn't seem to suffer the same problems.
it knows that this runs counter to recommended policy, and it doesn't care because it plans to the leave the issue to the dominant telecommunications providers.
Otherwise known as we pay experts to tell us things, but we get paid by telcos to ignore them. Different government - same corruption.
A pattern? Say, 10,000 Videos they don't have rights to? That's a pattern.
The only pattern 10,00 incorrectly identified videos represents is one of gross negligence. A pattern would be 40% of the videos they do not have rights to being commentaries claiming that one of their shows is a steaming pile of bullshit. That shows a pattern of deliberately targetting videos which negatively comment on their products.
At this stage, I know of only 2 videos - this one & one from Harvard. That's a long way from 10,000.
Only the author or those deriving their rights through the author can rightfully claim copyright.
An admitedly quick search doesn't pull up any cases where someone was sued for infringement based on a false claim of title. Also given that these notices were filed under the DMCA & the DMCA itself has penalties for false filings(1), trying to push this through would probably be doomed to failure at the dismissal request.
Mistakenly claiming ownership will result in damages. Sorry, we meant video 12659 not 12650.
Falsely claiming ownership to create a competitive advantage will result in a tort suit for interfierance and purjury.
Falsely claiming ownership with malicious intent gives you slander of title and purjury. Title 17 Chapter 5 (501c)
(c) Fraudulent Copyright Notice. - Any person who, with fraudulent intent, places on any article a notice of copyright or words of the same purport that such person knows to be false, or who, with fraudulent intent, publicly distributes or imports for public distribution any article bearing such notice or words that such person knows to be false, shall be fined not more than $2,500.
Is the only other section I find in the actual text of title 17 reguarding falsely claiming copyright ownership. The section you quote is from a copyright primer - not the code itself. Even 501c may not apply because they are not placing notices on the works, just claiming to own them - a fine distinction that might be enought to avoid the fine.
Title 17 Chapter 1 (106a) covers rights of atribution for visual works, but again doesn't invoke infrignement for falsely claiming that atribution.
(1)DMCA 512(c)(3)(A)(vi) makes knowingly making a false claim an act of Purjury. Proving knowingly and not error is hard enough to prevent these from showing up in court unless the person requesting the takedown gets picked to be made an example of.
It really depends on what level of accuracy was involved. If there were 40 videos out of 100,000 that were incorrectly identified, and the correct videos are a transposed digit away/a number off/etc then it's forgiveable. If 30% of the videos are incorrectly identified then it's an example of gross negligence & needs to be smacked down hard enough that they feel the sting every time they try this.
As for damages, that is going to depend on the exact video involved. If this were video 5 of a 6 part video collection that I was basing sales on, it might hurt me badly - giving my company an unprofessional appearance at the least.
Slashdot Mirror
If you are doing large datasets, you can end up trippling your filesize easily.
...
. .,90
Think about this, if you have 90 datapoints recorded 100 times a second, that's 9K+ sets of tags that have to be decoded every second with at least 7 extra characters per tag set. That doesn't include the performance hit for encoding & decoding the datastructure.
<data>
<time stamp=[timestamp]>
<sensor number=1>5.00</sensor>
<sensor number=2>5.01</sensor>
<sensor number=90>5.00</sensor>
</time>
<time stamp=[timestamp+1]>
...
</data>
For large, structured datadumps like that, CSV can actually be a faster way to work. A hybrid works even better to make a more flexible system with some of the best features of both.
<format>
<interval>
<unit>Second</unit>
<value>0.01</value>
</interval>
<field name='Sensor1' position=1/>
...
<field name='Sensor90' position=90/>
</format>
<CSV_Data>
1,2,3,4,5,.
1,2,3,5,4,...,90
...
</CSV_Data>
It uses XML to define the data format, but at the same time, it doesn't require a lot of processing to decode the actual dataset. Also the file size is minimally impacted.
... isn't that close enough?
I don't see how HID is planning on getting around the education & research exemption in the patent process.
If there is a formal request made to MS to present the specific IP to the community so that it can be removed, and MS fails to do so, then they can loose the right to persue the claims in court. In short, once they are legally asked to put up or shut up, failure to put up makes all their - 'we'll sue' FUD worthless.
And a jury found a substitute-teacher guilty of 'endangering minors' because of a porn storm on her PC. Juries base thier judgment on the information presented to them, and their opinion of the people involved --- "He's a good looking man & he would never have to rape someone" was the actual reasoning given by a juror on why she wouldn't vote guilty.
MS, Apple, Creative, and RCA all licence MP3 technology through a different company. Alcatel recently merged with another company and acquired certain patents, one of which is this MP3 patent jointly held between Alcatel & another (IIRC) French company. Everyone pays the other company, nobody was paying the company that Alcatel just bought out. So this isn't about MS not paying for the patent, this is another company popping up after the fact & saying 'gimmee too'. The reality of this is that either everyone is violating Alcatel's patents, the patents are invalid, or nobody who is paying this other company is violating the patent. Up until this ruling, everyone selling MP3 players/software was firmly in the 3rd camp.
Now, as much as I dispise MS, I really don't think they were violating this companies IP. They purchased the rights to the software through the company recognized by the industry as the proper licensor.
Back in the 70s IBM used the FUD campaign to help maintain it's monopoly. It's part of what started the anti-trust suit against them.
Updating the wikipedia entry with information you know to be false is vandalism. It's not a question of wikipedia not liking it. It's an act by an individual(s) to forcibly alter the content of the page in a manner not consistant with reality.
I do consider things like government officials having their staff daily editing their entries to post their 'stand' on an issue, and deleting all the evidence to the contrary, to be vandalism. The same as I consider someone adding these false claims to Fuzzy's entry to be vandalism.
These are not the same as people posting incorrect dates or honestly confusing 2 the 2 'Howard Sterns' in the news right now (One a radio DJ & the other a lawyer). This particular instance appears to be a malicious addition to his entry for whatever reason. That makes it both libel, and vandalism.
Actually, you can make decisions to not squeaze every penny out of a product through persuing lawsuits etc.
Lawsuits are not cheap on several fronts. First there is the actual cost - from additional document retention right up through the actual cost of the lawyers themselves. The other cost is goodwill. It's a lot less tangible, but ask SCO how it feels to sue your customers. Or perhaps ask people their opinion of the MPAA & RIAA. There should be no board of directors in the world that will look at a case of an individual trading a few copies of their songs & going after them for copyright infringment. The loss of goodwill and the legal fees would far outway any possible positive result from the suit. The RIAA & MPAA however insulate the actual board of directors for their respective cartels. Because they do not have to cope with the loss of goodwill, the total cost of suing the consumers of their product is reduced to a level that they find acceptable. The key factor in showing it to be a scare campain is that they are not likewise trumpeting their successes in catching pirate copies of pressed CDs.
1 in 4 Windows boxes is currently part of a botnet. Hand out Linux boxes instead of Windows ones and at least we can kill off the bots for the "Email & Web browsing" crowd (about 1/3 of current residential users).
I might trade my fringe benefits for the French benefits if she's cute & willing.
It would really depend on what type of data you are discussing, what type of protocols are involved, and the type of examination going on. The answers to the first 2 questions also determine whether or not it's going to actually improve your perceived performance.
[internet 101]
The TCP/IP stack takes a stream of data and breakes it down into 1 or more individual packets. Each packet under the current IPv4 is wrapped in a header that contains a minimum of:
There are some other items in there also, but those are the major ones.
When a packet is sent using TCP [Traffic Control Protocol], the IP [Internet Protocol] stack at the receiving end takes care of re-assembling the packets into the correct order & if one is missing, requesting that it be re-transmitted. Some protocols - ftp, http, bittorrent, certain buffered streamed media - which assemble larger files from these small packets are not significantly effected by a high latency transmission. If packet 2 arrives after packet 3, the stack puts them in the right order and the end user is none the wiser. These protocols can be routed to sub-optimal paths and as long as the packet loss/retransmission rate is kept under about 10% the end user doesn't notice any change.
Other types of protocols are sensative to sequence, latency, and lost packets. These include unbuffered streaming media (audio and video), VOIP, and real time games. These normally use a packet type called UDP [User Datagram Protocol]. This service does not allow for retransmission, nor does it correct for improper sequencing. If packet 2 is missing, packet 3 is decoded & implimented without reguard for whatever was in 2. If packet 6 comes in after packet 7, it's thrown away.
As you can see, this type of transfer is not acceptable for file transfers since a single missing packet will corrupt the file. However, this type of data transfer is required for real-time applications. If you are talking to someone using VOIP, a missing packet results in a short moment of silence - usually not long enough to actually be heard, just to degrade the overall quality. However, requesting that that missing packet be re-transmitted, would actually create a longer pause.
[/internet 101]
[QoS]
During the actual transmission of the data packets, they are handed from one router to another - often from one carrier to another. If one of the routers becomes busy, it begins queueing the packets. When it's buffer becomes full, it starts dropping packets into the bit bucket. Currently the first type of packet to be dropped is the ICMP [Internet Connection Maintenance Protocol] set - these include the trace & ping packets. After that, packets are dropped on a first come first served basis. If packet R1[receive 1] hits the router while the buffer is full, it's dropped. If packet R2 [receive 2] hits a milisecond later - just after packet T1[transmission 1] was transmitted & cleared from the queue, it's accepted. Packet C would then be rejected as the buffer is again full.
That is net neutrality at it's most basic. All of the data packets from all of the protocols are handled identically. Recognizing that this is not nescessarily the best method, QoS was introduced as a means of handling the different latency requirements of the different protocols. In a basic system, a QoS aware router examines the protocol flag & buffers the data in 2 seperate queues - one time sensative, one less sensative. Data is then transmitted according to a priority scheme - 2:1, 11:9, etc - which allows the time sensative data to be transmitted through faster & with le
I really don't want to keep finding framegrabs of Prince playing the guitar at the Superbowl when I go searching for my daily dose of PRON. (Yes, people were complaining that Prince, was [looks both ways before whispering] sexual during his performance & his shadow looked like a big penis)
Which will be hugely amusing since they spent millions developing secure ones to get around the 'great firewall'.
Ohh, I bet we can work up a log to post to the internet showing some congressional connections to that child porn bust. Few MAC addresses, the block of IP addresses for their ISP, a disclaimer showing that it's only an example of the possible abuse of this type of legislation. Oh, a nice bit of IP address spoofing/trumping to show it happening live - that would be good.
Last time this came up, it was estimated to cost over $400M/year. The estimated number of arrests it would help generate? 700. The FBI said just give them the $400M for agents & they could do a hell of a lot better. The truth is that the 60-90 day cycle that most of these companies already have is enough to cover the vast majority of the requests by the police - this is asking the industry to absorb $400M in costs for an infitesimal gain.
Funny the AG didn't want to do that... guess it didn't sound as good politically.
You lease politicians. You can never really buy them. You lease them for a few votes then return them back to their parties - just like a leased car going back to it's dealership.
Charter did this in some places in Western Ma. Parts of Chicopee & Holyoke had problems with up to 80% packet loss every night between 6-9PM as everyone got home & tried to get on the internet at the same time. The UBR's didn't have the bandwidth to cover the collections of modems attached to them.
The free market only works if there is a 'market' to work with. When it's a single provider, or providers working as a cartel, the free market cannot produce market corrections.
I had an AT&T fiber line run past my house close enough to spit on from my porch. It was 3 years before I could get broadband from them. It was 4 years before Verizon even offered DSL to me (768 is still the fastest they offer). Oh, and Direct Way satallite service is out too, I have a hill between me & the sat. That means that I had a 3 year wait between Broadband being offered in town center (5 blocks away) and getting it at my house, and another year before I had an option - 1/4 mile down the road there's still only 1 provider, Comcast.
So for the vast majority of the people in CN & USA, there is no significant alternative provider to convert to. Thus - there is no free market.
However, VOIP & HTTP requests can be routed with different priorities - VOIP is sensative to lag, HTTP isn't.
The concept of traffic shaping is to provide a QoS [Quality of Service]flagged route for packets which maximizes the use of the fastest, cleanest routes for lag/packet loss sensative protocols, while relegating less sensative packets to routes which may not be as responsive. The Telco extention[perversion] of packet shaping is to convert the selection criteria from protocol needs to accounting balance. Thus some of the Canadian telcos have already started to throttle Vonage service to the point of compromising service quality - remarkably just before they roll out their own service which doesn't seem to suffer the same problems.
no text, yep, I can't spell worth crap - we should all know that by now :)
The only pattern 10,00 incorrectly identified videos represents is one of gross negligence. A pattern would be 40% of the videos they do not have rights to being commentaries claiming that one of their shows is a steaming pile of bullshit. That shows a pattern of deliberately targetting videos which negatively comment on their products.
At this stage, I know of only 2 videos - this one & one from Harvard. That's a long way from 10,000.
An admitedly quick search doesn't pull up any cases where someone was sued for infringement based on a false claim of title. Also given that these notices were filed under the DMCA & the DMCA itself has penalties for false filings(1), trying to push this through would probably be doomed to failure at the dismissal request.
Mistakenly claiming ownership will result in damages. Sorry, we meant video 12659 not 12650.
Falsely claiming ownership to create a competitive advantage will result in a tort suit for interfierance and purjury.
Falsely claiming ownership with malicious intent gives you slander of title and purjury.
Title 17 Chapter 5 (501c)
Is the only other section I find in the actual text of title 17 reguarding falsely claiming copyright ownership. The section you quote is from a copyright primer - not the code itself. Even 501c may not apply because they are not placing notices on the works, just claiming to own them - a fine distinction that might be enought to avoid the fine.
Title 17 Chapter 1 (106a) covers rights of atribution for visual works, but again doesn't invoke infrignement for falsely claiming that atribution.
(1)DMCA 512(c)(3)(A)(vi) makes knowingly making a false claim an act of Purjury. Proving knowingly and not error is hard enough to prevent these from showing up in court unless the person requesting the takedown gets picked to be made an example of.
It really depends on what level of accuracy was involved. If there were 40 videos out of 100,000 that were incorrectly identified, and the correct videos are a transposed digit away/a number off/etc then it's forgiveable. If 30% of the videos are incorrectly identified then it's an example of gross negligence & needs to be smacked down hard enough that they feel the sting every time they try this.
As for damages, that is going to depend on the exact video involved. If this were video 5 of a 6 part video collection that I was basing sales on, it might hurt me badly - giving my company an unprofessional appearance at the least.
I said no text