They will no doubt try to use the European Patent Convention instead, create precedence for software patents, maybe try a new directive.
But it will not be easy for them. Many journalists, politicians, etc, actually have a clue by now. Many people have become active fighting software patents, and they are good at it. Every election brings in younger and more technology-savy members. Most countries have legislation against patents on software.
If they do kill this directive we might a few years of wild lawsuits and 10 years of fighting against new conventions and directives.
It doesn't matter where the company is located or where the product is produced. It only matters where the product is sold.
So both European and US companies can be sued (and lose) for patents violations when selling in the US but not the EU.
Which is why it is so stupid when european companies threaten to move to move the productions to US and India if they cannot get software patents in EU. They might move production to India anyway, but not because of Patents.
Yes, but why did he mention Lindows (Linspire) so prominently.
Maybe he just run Linspire on his desktop or maybe Sony made some kind of deal with Michael Robertson to put Lindows on the PS3. After the X-box challange MR would seem a good choice for that job:-)
I assume it was through sattellite. The plane was somewhere very north in Canada, coming from Europe. It was a VoIP (SIP) call, I received it on a D-Link SVG1402.
The calculation gives 240ms of latency (half RTT). That is a little more than is normally accepted in telephone systems, and there are more sources of latency. But it was not miserable.
>but the jitter as your packet is relayed via satellite back to firma >terra will be enough to discourage most calls (i have to imagine).
You imagine wrong. I received a VoIP (SIP) call from an airplane (a SAS flight). I wasn't a great line, but definitely useful and the latency was not too bad.
Yes, and it almost work. But go to any VoIP forum and you will se that many of the problems people are reporting are somehow related to NAT.
I know people who have VoIP adapters that they cannot use because of NAT. Yes, it is buggy routers (ie. Zyxel 650) supplied by the ISP, but without NAT, routers would be simpler and routers less buggy. Ie the problem with the Zyxel routers is that they try to be SIP-aware because of NAT.
SIP does not work well with symmetrical NAT.
SIP often does not work behind two or more levels of NAT. More than one level might not be necessary in most cases, but a lot of routers, WiFi accesspoints default to use NAT, and users do not know how to turn it off. NAT just complicates everything.
In short, STUN, portforwarding, tunneling, proxying etc all works. But from the point of view of a VoIP phone company it means that 15% (my guestimate) of its custumers cannot just plug in the phone and having it work. And almost every custumer that use his/hers phone on multiple networks will have problems with NAT. That is a big problem.
>Fine. Can I patent a device that includes software? "A device >consisting of a general-purpose CPU plus the following software..."?
No. Not if that line is drawn and the software can be uploaded to the device.
Of course we can discuss if the line is drawn at tre right place. I.e. the VLSI layout in a device could be patented even though some would consider it programming/software and FPGA software in a hardware description language could not be patented. But the line is there.
>More to the point, its useless; there's no "moral" difference between >software embedded into silicon and software burned to CDs.
I agree.
>The effort is better spent trying to distinguish "obvious" vs. >"non-obvious" patents, regardless of their form.
I do not want patents on everything non-obvious in iterature, movies, scientific results, software, sport, etc.
Besides it would be really hopeless to make the patent offices only grant non-obvious patents.
>First, it is simply impossible to draw the line between a pure software >patent and an embedded one,
>and between an embedded one and a pure hardware patent.
No, that line can be drawn: Software is something that can be stored on CD, emailed or downloaded from a webserver. Software can be distributed, sold etc without violating any patent. That is, if you can produce hardware and put the software (e.g, firmware) on a CD in the box you do not have to worry about patents for that software.
== Second, in this view, the RSA algorithm is not patentable - a brilliant piece of work by three top-rated minds solving a well-defined problem which has defied solution for several years. ==
A lot of brilliant mathematical work is not patented. Do you think Rivest, Shamir, and Adleman would not have worked on this if it was not patentable? They even published it the year before applying for the patent. And they did not invent it out of nothing. It is not that different from the earlier Pohlig-Hellman private key algorithm.
Yeah, just as free toilets, free paper towels, free sugar, free reading of newspapers, etc at cafes are classic tragedies of the commons. Yet most cafes do not charge for use of toilets. Everyone and their mom might come to read the newspaper at the cafe. But is not a tradegy as long as they buy a cup of coffee.
I.e. say a 1Mbit/s internet connections is $100/month. A VoIP call on average is 50Kbit/s, the store is open 300 hours/month. Then an hour of phoning cost the cafe 50*$100/300hours/1000 = 2 cents.
I probably buy a cup of coffee for every 10 minutes I talk. I think I already pay for what I use.
A lot cheaper in some cases. Eg. international calls.
I live in Denmark but use SIP phones when traveling. From the US phoning Denmark is $1.30 to $1.80 per minute (dependig on US roaming operator) with GSM and $0.02 with my SIP phone. Calling a US number is again $1.30 to $1.80 with GSM and $0.03 with SIP (A cent more because it is an internation call for a danish SIP-provider) That is unless I call another SIP number from my SIP phone in which case the call is free.
Receiving calls in the US cost me ~$1 with GSM and is free with SIP. I have en extra US number with ipkall.com, so calling me is a lot cheaper with SIP than GSM both from Denmark and the US.
And the US is not the worst place. I was in Vilnius in marts. It would have cost me $4/minute to call home.
> Much better suited for office work IMO.
Most of my friends and family have WiFi. Which means I can bring a SIP WiFi phone whenever I visit someone.
>I can easily see two or three of these things being brought into a >coffee shop and soaking up all the bandwidth.
The coffee shop would have a poor connection then. And most people do not phone all the time. E.g. the G729 codec use ca 30Kbit including RTP. A 1Mbit/1Mbit connection should handle 30 connections.
OO.org already have word->OASIS and OASIS->word filter. You would think it would be easy (compared to making OpenOffice) to develop MSOffice plugins that could read and write OASIS documents.
You would even think there would be custumers willing to pay for that. If someone, eg. the city of Munich, switch to OASIS they could put OASIS documents on webpages and in email (where HTML and cleartext is not approriate) along with a link to these plugins.
>Bittorrent has plenty of legit uses. But do you really think that's >what most people use it for? I'd say most are looking for porn, movies, >software, etc.
If you only control the DHCP server, and cannot even force people to use the DHCP server you do not have a chance.
Try to get an old Linux box and set it up as router. Then install PortSentry on the this router. Every virus will immidiately attack the router and portsentry will then cut it off.
Yes, I hear all these wild claims. I want someone to print out the compiled Linux kernel (ca 8mbits) as 0 and 1's, Hex or whatever they want. Put in on a wall, take a photo in normal sunlight, develop the photo, scan the photo, and boot kernel (a do a diff which is more precise but less fun).
Then do the same for OpenOffice using a color printer.
Slashdot Mirror
If we all start usings TOR, mixmaster, etc. only 60 pct or so will be spam, scam, etc, just like it is on the internet now.
> In you register, then it's worth going after people.
In the US, but we were disussing Europe.
You do not even have to put your name on it ot write "copyright" on it. /.
So _you_ owe quarter for any message on
An artist can already contract a single authority to distribute the music in the EU.
Except you do not have to deal with any copyright offices. You do not have to register copyright.
You can write a song, perform it, sell the recording all over Europe with registering anything.
> Why don't all these VOIP companies use some sort of encryption?
Some of them do (eg. musimi.dk which i use). They use S-RTP AES.
Not all clients support it though (Sipura, Zultrys do).
There are other SIP-clients for Linux. eg. KPhone
I have used SIP-phones behind many NAT routers, and I did not have to configure anything (they use STUN).
No they will not.
They will no doubt try to use the European Patent Convention instead, create precedence for software patents, maybe try a new directive.
But it will not be easy for them. Many journalists, politicians, etc, actually have a clue by now. Many people have become active fighting software patents, and they are good at it. Every election brings in younger and more technology-savy members. Most countries have legislation against patents on software.
If they do kill this directive we might a few years of wild lawsuits and 10 years of fighting against new conventions and directives.
It doesn't matter where the company is located or where the product is produced. It only matters where the product is sold.
So both European and US companies can be sued (and lose) for patents violations when selling in the US but not the EU.
Which is why it is so stupid when european companies threaten to move to move the productions to US and India if they cannot get software patents in EU. They might move production to India anyway, but not because of Patents.
Yes, but why did he mention Lindows (Linspire) so prominently.
:-)
Maybe he just run Linspire on his desktop or maybe Sony made some kind of deal with Michael Robertson to put Lindows on the PS3.
After the X-box challange MR would seem a good choice for that job
I assume it was through sattellite. The plane was somewhere very north in Canada, coming from Europe. It was a VoIP (SIP) call, I received it on a D-Link SVG1402.
The calculation gives 240ms of latency (half RTT). That is a little more than is normally accepted in telephone systems, and there are more sources of latency. But it was not miserable.
>but the jitter as your packet is relayed via satellite back to firma
>terra will be enough to discourage most calls (i have to imagine).
You imagine wrong.
I received a VoIP (SIP) call from an airplane (a SAS flight). I wasn't a great line, but definitely useful and the latency was not too bad.
> SIP can use an STUN to get around NAT.
Yes, and it almost work. But go to any VoIP forum and you will se that many of the problems people are reporting are somehow related to NAT.
I know people who have VoIP adapters that they cannot use because of NAT. Yes, it is buggy routers (ie. Zyxel 650) supplied by the ISP, but without NAT, routers would be simpler and routers less buggy. Ie the problem with the Zyxel routers is that they try to be SIP-aware because of NAT.
SIP does not work well with symmetrical NAT.
SIP often does not work behind two or more levels of NAT. More than one level might not be necessary in most cases, but a lot of routers, WiFi accesspoints default to use NAT, and users do not know how to turn it off. NAT just complicates everything.
In short, STUN, portforwarding, tunneling, proxying etc all works. But from the point of view of a VoIP phone company it means that 15% (my guestimate) of its custumers cannot just plug in the phone and having it work. And almost every custumer that use his/hers phone on multiple networks will have problems with NAT. That is a big problem.
>Yes, beacuse that will be SO much easier for his mother to use
Yes it will. Linux works for my mother. I am sure MacOs is easy too, but his mother would have to get a new computer.
Instead of complaing about MS, you should just use a different OS product. Linux, MacOS, Solaris,...
>Fine. Can I patent a device that includes software? "A device
>consisting of a general-purpose CPU plus the following software..."?
No. Not if that line is drawn and the software can be uploaded to the device.
Of course we can discuss if the line is drawn at tre right place. I.e. the VLSI layout in a device could be patented even though some would consider it programming/software and FPGA software in a hardware description language could not be patented. But the line is there.
>More to the point, its useless; there's no "moral" difference between
>software embedded into silicon and software burned to CDs.
I agree.
>The effort is better spent trying to distinguish "obvious" vs.
>"non-obvious" patents, regardless of their form.
I do not want patents on everything non-obvious in iterature, movies, scientific results, software, sport, etc.
Besides it would be really hopeless to make the patent offices only grant non-obvious patents.
>First, it is simply impossible to draw the line between a pure software
>patent and an embedded one,
>and between an embedded one and a pure hardware patent.
No, that line can be drawn: Software is something that can be stored on CD, emailed or downloaded from a webserver. Software can be distributed, sold etc without violating any patent. That is, if you can produce hardware and put the software (e.g, firmware) on a CD in the box you do not have to worry about patents for that software.
==
Second, in this view, the RSA algorithm is not patentable - a brilliant piece of work by three top-rated minds solving a well-defined problem which has defied solution for several years.
==
A lot of brilliant mathematical work is not patented. Do you think Rivest, Shamir, and Adleman would not have worked on this if it was not patentable? They even published it the year before applying for the patent. And they did not invent it out of nothing. It is not that different from the earlier Pohlig-Hellman private key algorithm.
Yeah, just as free toilets, free paper towels, free sugar, free reading of newspapers, etc at cafes are classic tragedies of the commons.
Yet most cafes do not charge for use of toilets.
Everyone and their mom might come to read the newspaper at the cafe. But is not a tradegy as long as they buy a cup of coffee.
I.e. say a 1Mbit/s internet connections is $100/month. A VoIP call on average is 50Kbit/s, the store is open 300 hours/month. Then an hour of phoning cost the cafe 50*$100/300hours/1000 = 2 cents.
I probably buy a cup of coffee for every 10 minutes I talk. I think I already pay for what I use.
> VOIP is cheaper sure
A lot cheaper in some cases. Eg. international calls.
I live in Denmark but use SIP phones when traveling. From the US phoning Denmark is $1.30 to $1.80 per minute (dependig on US roaming operator) with GSM and $0.02 with my SIP phone. Calling a US number is again $1.30 to $1.80 with GSM and $0.03 with SIP (A cent more because it is an internation call for a danish SIP-provider) That is unless I call another SIP number from my SIP phone in which case the call is free.
Receiving calls in the US cost me ~$1 with GSM and is free with SIP.
I have en extra US number with ipkall.com, so calling me is a lot cheaper with SIP than GSM both from Denmark and the US.
And the US is not the worst place. I was in Vilnius in marts. It would have cost me $4/minute to call home.
> Much better suited for office work IMO.
Most of my friends and family have WiFi. Which means I can bring a SIP WiFi phone whenever I visit someone.
>I can easily see two or three of these things being brought into a
>coffee shop and soaking up all the bandwidth.
The coffee shop would have a poor connection then.
And most people do not phone all the time. E.g. the G729 codec use ca 30Kbit including RTP. A 1Mbit/1Mbit connection should handle 30 connections.
>but Laser Disks are still used at the industrial level to store master digital recordings of movies.
Laser Disks does not store video digitally.
I does happen. From Copenhagen airport:
http://www.agol.dk/gallery/Ofterm/IMG_1005
So why haven't anyone done it?
OO.org already have word->OASIS and OASIS->word filter. You would think it would be easy (compared to making OpenOffice) to develop MSOffice plugins that could read and write OASIS documents.
You would even think there would be custumers willing to pay for that. If someone, eg. the city of Munich, switch to OASIS they could put OASIS documents on webpages and in email (where HTML and cleartext is not approriate) along with a link to these plugins.
>Bittorrent has plenty of legit uses. But do you really think that's
>what most people use it for? I'd say most are looking for porn, movies,
>software, etc.
I think you greatly underestimate the legal uses of Bittorrent.
Just the Knoppix torrent:
http://torrent.unix-ag.uni-kl.de/
Accounts for more than 80 TeraBytes.
If you only control the DHCP server, and cannot even force people to use the DHCP server you do not have a chance.
Try to get an old Linux box and set it up as router. Then install PortSentry on the this router. Every virus will immidiately attack the router and portsentry will then cut it off.
Yes, I hear all these wild claims. I want someone to print out the compiled Linux kernel (ca 8mbits) as 0 and 1's, Hex or whatever they want. Put in on a wall, take a photo in normal sunlight, develop the photo, scan the photo, and boot kernel (a do a diff which is more precise but less fun).
Then do the same for OpenOffice using a color printer.