Credits are stored only on the machine giving you credit; they are not shared, and any claims to credit are ignored.
Thus, you only get credits from people you've uploaded to; this rewards you most if you're usually downloading the same type of stuff, as you'll probably find your credits put you in a group of people who boost you up.
This is why the judge gets discretion to cap costs, and to refuse to award costs; if I sued IBM, and lost on a technicality, IBM would probably still pay costs here in the UK.
Oddly enough, there were forms available for people who needed access to potentially offensive research material; they allowed you to specify what sort of material (and bestiality was one option), and how long you'd need access for.
Assuming your universities have similar usage policies to ours, then there are two reasons:
Campus firewall. If it blocks what you're after (e.g. incoming SMTP), you're SOL.
Usage policy. My alma mater has a usage policy that requires that your use of the network is predominately for academic reasons (we couldn't get private access). If private Internet access was available, I imagine that they'd ban non-academic use entirely.
Who said anything about stealing it? I simply do without new music (other than radio) because I can't use any of the download services (IRIX, Solaris and Linux are the available OSes here), and I feel that both CDs and legal downloads are overpriced for what you get.
Not all of us use illegal services to obtain stuff that we're not happy to buy. And some of us even support the idea behind the current crackdown on illegal P2P filesharers, even if the implementation goes awry.
The PCI bus is 32-bit transfers, at 33MHz. Assuming no loss to bus overhead, or to other users, a PCI bus can handle 1375 16bit/48KHz audio streams. Allowing a decent chunk for bus overhead, you've still got over 1000 streams flowing. Assume that they all come off an interface streaming in over the PCI bus, and you're still looking at a minimum of 400 streams.
So, PCI has more than enough bandwidth for silly numbers of audio streams; it's just that no-one bothers with serious audio.
I don't find any legit mail getting that flag. It has nothing to do with the EHLO or HELO command, merely with the IP address that's connected to the server.
Put briefly, when your MTA connects to my mail server, my mail server knows the IP that's connected to it. I then do a PTR lookup on that IP to get a name for the mailserver; if this is a name that's in a blacklist, I reject the mail. I then do a forward lookup on the name I've acquired, to obtain all A, AAAA, or A6 records (as appropriate). If none of these records match (after follwing CNAME chains and the like), then I flag the server as suspicious.
Most of the servers I do catch are Chinese IP address, or ATTBI addresses, claiming to be hotmail.com or yahoo.co.uk servers. To follow your point about RFC2821; DNS behaviour is not specified in RFC2821, or elsewhere in the RFCs. However, my empirical experience is that I don't catch wanted mail this way, but I do catch unwanted mail; YMMV, but it works for me.
As a simple test, aimed at making spoofing a little more awkward, my mail server does a reverse lookup on the IP of the connecting mail server, then does a forward lookup on that name; if the connecting IP does not match the IP looked up, a special header is inserted, which SpamAssassin uses to score the mail a little more strongly. No mail lost, but you cannot easily spoof someone else's domain (set your reverse lookup to mail.hotmail.com, and my server will increase your spam score).
Sendmail's licence is a hybrid between the GPL and the BSD licences. I think it lets you get away from the patent issue though (ask a lawyer to be certain).
Does it really cause people nasty problems that there are over ten different manufacturers of NTSC decoding circuits, and over 100 different combinations of NTSC decoder and display device? Do people really find it difficult to produce content that conforms to all the combinations of NTSC receivers that exist?
The problem in software is simply that we're not very good at producing software standards, or at keeping to them. If we had good standards properly implemented for the WWW, the user's choice of browser would be a non-issue. It doesn't bother people that there's more than one different DVD player, or TV set design. Each still does the same thing from the content provider's perspective; it's a reflection of how immature software is that we can't do the same with things like web browsers.
In answer to point 1; I expect the application to load its UI on startup. I'd rather lose an extra second when I'm already waiting for the application, than lose a second when I'm trying to do things.
In this case, the application should load its icons when it loads; then, when the menu comes up, it's instantaneous. Load time is meaningless to me when I work in (maybe) 5 or 6 applications that are left running all the time. Render time when I click on a menu is important.
And don't try and argue that loading all the icons is going to push me into swap; even if it did, swapping applications back in takes far less than a second, even though I don't go into swap.
What happens if I want to be notified of an incoming IM, and I have sound disabled, while working in another desktop?
Certainly with Kopete, and presumably with other multiple desktop aware IM programs, a new message notification can be made to appear briefly on all desktops; this can be configured on a per-contact basis.
Thus, when my close friends IM me, I get a prompt no matter what virtual desktop I'm on. When it's IM spam, or a stranger, I don't get prodded.
In other words, as with all PKI systems, TCPA boils down to whether the private keys can be kept secret. If just one gets stolen, you've got trouble.
Although DVD-CSS wasn't PKI-based, it was broken initially in exactly that fashion; Xing failed to keep their player key secret, and the knowledge gained by using Xing's key showed that the system was weak. Xing's key was promptly revoked, and as a minor player, this caused no major hassle for anyone, but since the system was weak, DVD-CSS was dead.
In this system, the danger point is someone leaking a major manufacturer's key; killing a small player like Xing over poor security didn't upset many people. If Sony used one key in all their devices, from cheap portables to expensive HiFi players, and I obtained it, I can put the industry in a nasty situation. Do you stop all the millions of Sony players from working, and upset consumers, or do you just release stuff that you know is hacked?
Based on past evidence (DVD-CSS), I suspect that sooner or later, TCPA will hit that dilemma.
BTW, I am aware that the public key is guaranteed to leak; however, not everyone in this thread acts like they have a basic understanding of PKI, and I was attempting to make it clear that the public key alone isn't enough.
As far as I can see, the weak point in this scenario is that there is one trusted root key (the Trusted Computing Group key); as soon as that key is leaked (both private and public halves), it is possible to generate fake data that I can use to intercept the media.
Also, the highest profit margin customers are the ones with most options when it comes to changing for better service; if your quality of service drops too low, it's worth your competitors' time competing for high-margin customers on quality of service rather than price.
But modern caches are set-associative, and LRU; hence, in the worst case where two cores are accessing different data with the same tags, they split each set evenly between them. Thus, a dual core with shared cache is never any slower than a dual core with the same amount of cache split in two.
Practically, a dual core with shared cache will probably have less cache than a dual core with two caches, and thus will be slower, but there is no reason why the cache controller has to be designed in such a way that two memory intensive programs can trample each others' cache elements.
Having seen the advert, I'm not surprised they got told off; the gist of it was that Linux had to be worse than Windows, since Windows on a dual Xeon was as fast as Linux on an S/390 mainframe, but at 1/10th the cost.
If you didn't read the website the advert pointed you at very carefully, you would be led to believe that Linux needed much more expensive hardware than Windows to even match capabilities; in fact, the study made no such claims.
How about open-book exams? I did them as part of my CS degree in the Perl, Java and data structures modules.
No Internet access, but you can bring in any books, notes, and other materials (excluding computer materials). You have access to the appropriate language reference guide. You've no idea what the actual problem set will be, and thus you can only really bring reference material in; people who know more are at an advantage because they spend less time consulting references.
The reason I'm happy to swap CDs/DVDs/cartridges/whatever in a console, and not happy to swap things on a PC is that on a console, I do not have to install many gigabytes of stuff just to play.
I would be quite happy to swap CDs/DVDs on a PC if the game could be played entirely from that disc. I am not happy about copying a couple of gigabytes of data to my hard disc, then inserting the CD every time I want to play.
If you're a games developer, choose one: Either require the CD to be inserted, but don't put anything other than savegames and other personalised data on my hard disc, or install to the hard disc, but don't require the CD. Whichever you choose, I'll be happy.
On the contrary, there are satellites above Africa; IIRC, M-Net in South Africa serves the whole continent with digital satellite TV from PanAmSat's PAS-10 and PAS-7 satellites. Given that it's possible to get DST, it's also possible to transfer data via satellite using the same satellite.
However, you can split an application into parts that come from the GPL library, and parts that you wrote, which you can release under the BSD licence if you wish. Under copyright law, the combination of the two is licenced under the most restrictive combination of the two licences (in this case, the GPL). If I then rewrote the GPL library parts in a clean room (thus avoiding any copying of code), and released my rewrite under the BSD licence, you could then combine my library and your code, and have the application released under the BSD licence.
The only reason the GPL creates issues for compatibility is that the GPL both adds restrictions (thus, a GPL-compatible licence must permit those restrictions to be added), and prohibits you from adding further restrictions (thus, a compatible licence must not add restrictions the GPL doesn't). You can release an application built on GPL libraries under the BSD licence, so long as you are aware that it is only the application code that is BSD licensed; the combination of application code and essential libraries is GPL licensed, but your code alone does not have to be.
I would assume that they're compliant with section 3a. Every time they've supplied a binary to a developer internally, they've supplied the source code too. You can't get the source, because you possess neither a legally acquired (not stolen - perhaps by stealing a disc, or a laptop) binary, nor a copy of a written offer to supply source.
This is assuming that they have distributed within the eyes of the law to their developers. It may well be the case that all their copies count legally as distributed to SpecOps Labs, in which case, so long as SpecOps Labs have ensured that they have source as well as binary, they're in compliance with the GPL, because all distribution has taken place under GPL section 3.a)
Slashdot Mirror
Thus, you only get credits from people you've uploaded to; this rewards you most if you're usually downloading the same type of stuff, as you'll probably find your credits put you in a group of people who boost you up.
This is why the judge gets discretion to cap costs, and to refuse to award costs; if I sued IBM, and lost on a technicality, IBM would probably still pay costs here in the UK.
Oddly enough, there were forms available for people who needed access to potentially offensive research material; they allowed you to specify what sort of material (and bestiality was one option), and how long you'd need access for.
I'm not connected to Internet 2, so I can't comment on what it uses, but if it's exclusively IPv6, how come its website is pure IPv4?
Not all of us use illegal services to obtain stuff that we're not happy to buy. And some of us even support the idea behind the current crackdown on illegal P2P filesharers, even if the implementation goes awry.
So, PCI has more than enough bandwidth for silly numbers of audio streams; it's just that no-one bothers with serious audio.
Put briefly, when your MTA connects to my mail server, my mail server knows the IP that's connected to it. I then do a PTR lookup on that IP to get a name for the mailserver; if this is a name that's in a blacklist, I reject the mail. I then do a forward lookup on the name I've acquired, to obtain all A, AAAA, or A6 records (as appropriate). If none of these records match (after follwing CNAME chains and the like), then I flag the server as suspicious.
Most of the servers I do catch are Chinese IP address, or ATTBI addresses, claiming to be hotmail.com or yahoo.co.uk servers. To follow your point about RFC2821; DNS behaviour is not specified in RFC2821, or elsewhere in the RFCs. However, my empirical experience is that I don't catch wanted mail this way, but I do catch unwanted mail; YMMV, but it works for me.
As a simple test, aimed at making spoofing a little more awkward, my mail server does a reverse lookup on the IP of the connecting mail server, then does a forward lookup on that name; if the connecting IP does not match the IP looked up, a special header is inserted, which SpamAssassin uses to score the mail a little more strongly. No mail lost, but you cannot easily spoof someone else's domain (set your reverse lookup to mail.hotmail.com, and my server will increase your spam score).
Sendmail's licence is a hybrid between the GPL and the BSD licences. I think it lets you get away from the patent issue though (ask a lawyer to be certain).
Does it really cause people nasty problems that there are over ten different manufacturers of NTSC decoding circuits, and over 100 different combinations of NTSC decoder and display device? Do people really find it difficult to produce content that conforms to all the combinations of NTSC receivers that exist?
The problem in software is simply that we're not very good at producing software standards, or at keeping to them. If we had good standards properly implemented for the WWW, the user's choice of browser would be a non-issue. It doesn't bother people that there's more than one different DVD player, or TV set design. Each still does the same thing from the content provider's perspective; it's a reflection of how immature software is that we can't do the same with things like web browsers.
In this case, the application should load its icons when it loads; then, when the menu comes up, it's instantaneous. Load time is meaningless to me when I work in (maybe) 5 or 6 applications that are left running all the time. Render time when I click on a menu is important.
And don't try and argue that loading all the icons is going to push me into swap; even if it did, swapping applications back in takes far less than a second, even though I don't go into swap.
Certainly with Kopete, and presumably with other multiple desktop aware IM programs, a new message notification can be made to appear briefly on all desktops; this can be configured on a per-contact basis.
Thus, when my close friends IM me, I get a prompt no matter what virtual desktop I'm on. When it's IM spam, or a stranger, I don't get prodded.
So, having a sofa isn't something special, and nor is being on the Internet. Owning an iPod is, thus the iPod is not part of the furniture.
Although DVD-CSS wasn't PKI-based, it was broken initially in exactly that fashion; Xing failed to keep their player key secret, and the knowledge gained by using Xing's key showed that the system was weak. Xing's key was promptly revoked, and as a minor player, this caused no major hassle for anyone, but since the system was weak, DVD-CSS was dead.
In this system, the danger point is someone leaking a major manufacturer's key; killing a small player like Xing over poor security didn't upset many people. If Sony used one key in all their devices, from cheap portables to expensive HiFi players, and I obtained it, I can put the industry in a nasty situation. Do you stop all the millions of Sony players from working, and upset consumers, or do you just release stuff that you know is hacked?
Based on past evidence (DVD-CSS), I suspect that sooner or later, TCPA will hit that dilemma.
BTW, I am aware that the public key is guaranteed to leak; however, not everyone in this thread acts like they have a basic understanding of PKI, and I was attempting to make it clear that the public key alone isn't enough.
Am I missing something important?
Also, the highest profit margin customers are the ones with most options when it comes to changing for better service; if your quality of service drops too low, it's worth your competitors' time competing for high-margin customers on quality of service rather than price.
Don't worry, the default is still plain text. No need for unwanted HTML mail.
Practically, a dual core with shared cache will probably have less cache than a dual core with two caches, and thus will be slower, but there is no reason why the cache controller has to be designed in such a way that two memory intensive programs can trample each others' cache elements.
If you didn't read the website the advert pointed you at very carefully, you would be led to believe that Linux needed much more expensive hardware than Windows to even match capabilities; in fact, the study made no such claims.
No Internet access, but you can bring in any books, notes, and other materials (excluding computer materials). You have access to the appropriate language reference guide. You've no idea what the actual problem set will be, and thus you can only really bring reference material in; people who know more are at an advantage because they spend less time consulting references.
I would be quite happy to swap CDs/DVDs on a PC if the game could be played entirely from that disc. I am not happy about copying a couple of gigabytes of data to my hard disc, then inserting the CD every time I want to play.
If you're a games developer, choose one: Either require the CD to be inserted, but don't put anything other than savegames and other personalised data on my hard disc, or install to the hard disc, but don't require the CD. Whichever you choose, I'll be happy.
On the contrary, there are satellites above Africa; IIRC, M-Net in South Africa serves the whole continent with digital satellite TV from PanAmSat's PAS-10 and PAS-7 satellites. Given that it's possible to get DST, it's also possible to transfer data via satellite using the same satellite.
The only reason the GPL creates issues for compatibility is that the GPL both adds restrictions (thus, a GPL-compatible licence must permit those restrictions to be added), and prohibits you from adding further restrictions (thus, a compatible licence must not add restrictions the GPL doesn't). You can release an application built on GPL libraries under the BSD licence, so long as you are aware that it is only the application code that is BSD licensed; the combination of application code and essential libraries is GPL licensed, but your code alone does not have to be.
This is assuming that they have distributed within the eyes of the law to their developers. It may well be the case that all their copies count legally as distributed to SpecOps Labs, in which case, so long as SpecOps Labs have ensured that they have source as well as binary, they're in compliance with the GPL, because all distribution has taken place under GPL section 3.a)