While this is true, there's no reason for the offer itself to be extended to another third-party, without going through a direct recipient first. Think of it as a call option that can be copied, and doesn't exhaust itself after one use. You need to be in possession of this contract (either by directly receiving it, or receiving a copy from another person) in order to exercise your option to receive the source.
IANAL though. Another way to think about it is, as a third party, you don't have the order form until someone gives it to you:)
If they didn't modify the source, they must pass on the written offer to provide source code they received, provided they're distributing the binary noncommercially. For commercial uses, or if they don't (or can't) pass on said written offer, then they must provide source code, even if it is unmodified. Of course, they need only provide source to people they give the binary to (including free demos or whatever).
Amazon's FPS can do recurring subscriptions (with multi-use tokens) as well - it's a bit more complex, of course, they give you the low-level stuff and leave it to you to build an application on top of it. The advantage here is you're the one triggering the billing each month - you don't have to rely on them to re-submit it for you. Of course, this is also a disadvantage, as your app needs to keep track of such things, and handle triggering the transaction at the proper times.
I don't know if a convenient 'trickle' uploader exists yet, but I'm setting up a backup scheme for myself that uses duplicity to upload to Amazon's S3, and uses an EC2 instance for a few hours each month to coalesce the incremental backups into a full backup. Since this is for my VPS, I don't worry too much about using a lot of bandwidth when it runs the backup (the incrementals are usually small anyway).
Problem one is there's no proven routing algorithm for flat networks. Also, how do you screen out file sharers? Most routers under this system will be run by ordinary people who don't want to sysadmin, so they won't screen them out, and with no central authority getting a new address will have to be trivial.
While the OP has the issue of not having a/debt/, and thus the university is not required to accept currency, it seems like in your case the legal tender rule applies - what other basis do they have to require you to pay but a debt, or a contract - and in the latter case, since you were not made aware of the need of the metro card, I don't know if that'd be legal either.
That's not a restriction of the GPL - it's pure copyright law. The GPL steps in if you want to do something that's beyond the bounds of copyright law - and the privilege of including GPL'd code freely into proprietary code is not one that is granted; thus it's denied by default, just like with any proprietary license.
If I'm not very mistaken, the very basis of contract law is that each side fully understands the terms of the contract. Moreover, since he didn't directly sign the document, wouldn't the burden of proof be on the side that is asserting that he did agree to it?
IANAL, etc.
A lot of installer packages on windows are designed for the proprietary EULA system, so what happens is the Free Software author sees 'licence file:', drops the GPL in, and now it comes up as a 'you-must-agree' screen.
Not that it matters; the GPL doesn't obligate or restrict you if you don't take advantage of its extra permissions.
Using 4MB blocks for everything would kill memory performance - and more specifically, mmap performance. Each library loaded in your system would require at least 4MB of ram - probably more, as they have code, data, and zeroed data segments. Additionally, each process would require another 4MB*n. There's no gain for doing this either, except under specialized circumstances, as the OS can already request a batch of sectors from the drive in one operation.
Operating systems tend to use 4096-byte blocks already, as that's the size of a memory page on x86 and amd64. If you were to require 16kb transfers, then the block cache would have to start allocating contiguous four-page groups for DMA transfers and the like, which could be difficult if memory is fragmented; in comparison, pages are the basic allocation unit for RAM, so 4kb's easy to find.
That's init=/bin/sh, and it's hardly a bug - you're just changing the kernel configuration to skip all that silly authentication stuff. Someone who has access to the bootloader (you do set a bootloader password, right?) probably has enough access to override whatever security - short of disk encryption - you might have by force if necessary.
While I understand the importance of web searches, would it be so hard to change the nature of web crawlers so that they didn't spider a site unless they were specifically allowed? You know, like you don't generally walk into somebody's house or cube at work and start rummaging through their stuff just because the door is open. Unless a web-crawler is invited into a site (via robots.txt) then it doesn't index it.
It would be trivially easy. It would also render the internet unsearchable, as there are virtually no sites which go out of their way to invite in spiders. In the end, most sites (including, probably, the ones you're searching for) would be invisible to spiders, not because their owners don't want spiders in, but because their owners don't know about robots.txt, or didn't bother whitelisting everything, or because all of the sites which link to them didn't bother whitelisting themselves. Spiders only work if most of the internet is accessible, not if most is blocked off.
If you really care about spiders, there is already a de-facto standard in place for communicating to them. Don't break something that already works, and don't expect an established community and system to change as you like just because you can't be bothered to read any major spider's FAQ.
Compression algorithms designed around block repetition and relative frequency of occurrence of particular byte values don't typically work too well with audio files - when you've got a waveform, you'll have jitter that wrecks any exact repetition of data, and since it's basically going back and forth, you won't see much difference between how often 0x10 and 0x40 occur. Audio lossless compression takes advantage of the specific properties of audio - for example, FLAC does things like taking a linear prediction of the waveform over a small space, then recording the differences between its prediction and the actual value, in a form which optimizes for numbers close to zero.
The information in question is not reported back to the tracker; the tracker could only detect bad peers by running its own trusted peer to probe the network - however this would probably be impractical in practice.
In actual use, individual peers generally detect a bad peer quickly enough, and then ban it locally. This tends to work well enough in practice.
Slashdot Mirror
That screenshot is of dolphin, not konqueror. Additionally, it's not the KDE 4 version, so the bug may well already have been fixed.
Nitpick: It is possible for 105% of the people required to take the testing to take it, iff people who are not required take it as well.
By default firefox does not send URLs to google. It downloads a static list from google periodically, and checks against that.
While this is true, there's no reason for the offer itself to be extended to another third-party, without going through a direct recipient first. Think of it as a call option that can be copied, and doesn't exhaust itself after one use. You need to be in possession of this contract (either by directly receiving it, or receiving a copy from another person) in order to exercise your option to receive the source. IANAL though. Another way to think about it is, as a third party, you don't have the order form until someone gives it to you :)
If they didn't modify the source, they must pass on the written offer to provide source code they received, provided they're distributing the binary noncommercially. For commercial uses, or if they don't (or can't) pass on said written offer, then they must provide source code, even if it is unmodified. Of course, they need only provide source to people they give the binary to (including free demos or whatever).
Amazon's FPS can do recurring subscriptions (with multi-use tokens) as well - it's a bit more complex, of course, they give you the low-level stuff and leave it to you to build an application on top of it. The advantage here is you're the one triggering the billing each month - you don't have to rely on them to re-submit it for you. Of course, this is also a disadvantage, as your app needs to keep track of such things, and handle triggering the transaction at the proper times.
On the contrary - Microsoft can make anything hard, if they try.
All joking aside, a silicon-based lifeform would find us about as appetizing as we find sand.
I don't know if a convenient 'trickle' uploader exists yet, but I'm setting up a backup scheme for myself that uses duplicity to upload to Amazon's S3, and uses an EC2 instance for a few hours each month to coalesce the incremental backups into a full backup. Since this is for my VPS, I don't worry too much about using a lot of bandwidth when it runs the backup (the incrementals are usually small anyway).
Problem one is there's no proven routing algorithm for flat networks. Also, how do you screen out file sharers? Most routers under this system will be run by ordinary people who don't want to sysadmin, so they won't screen them out, and with no central authority getting a new address will have to be trivial.
What? 4chan memes in my slashdot?
Captcha: dignity, which clearly neither of us has anymore.
While the OP has the issue of not having a /debt/, and thus the university is not required to accept currency, it seems like in your case the legal tender rule applies - what other basis do they have to require you to pay but a debt, or a contract - and in the latter case, since you were not made aware of the need of the metro card, I don't know if that'd be legal either.
IANAL etc
That's not a restriction of the GPL - it's pure copyright law. The GPL steps in if you want to do something that's beyond the bounds of copyright law - and the privilege of including GPL'd code freely into proprietary code is not one that is granted; thus it's denied by default, just like with any proprietary license.
The Japanese may have beat you to it. Sort of. Well, the world doesn't get rebuilt every night, but how's about ~3 years ago?
If I'm not very mistaken, the very basis of contract law is that each side fully understands the terms of the contract. Moreover, since he didn't directly sign the document, wouldn't the burden of proof be on the side that is asserting that he did agree to it? IANAL, etc.
A lot of installer packages on windows are designed for the proprietary EULA system, so what happens is the Free Software author sees 'licence file:', drops the GPL in, and now it comes up as a 'you-must-agree' screen.
Not that it matters; the GPL doesn't obligate or restrict you if you don't take advantage of its extra permissions.
mka's just a container; write a script to mux/demux the track as needed when you load music into the player.
Using 4MB blocks for everything would kill memory performance - and more specifically, mmap performance. Each library loaded in your system would require at least 4MB of ram - probably more, as they have code, data, and zeroed data segments. Additionally, each process would require another 4MB*n. There's no gain for doing this either, except under specialized circumstances, as the OS can already request a batch of sectors from the drive in one operation.
Operating systems tend to use 4096-byte blocks already, as that's the size of a memory page on x86 and amd64. If you were to require 16kb transfers, then the block cache would have to start allocating contiguous four-page groups for DMA transfers and the like, which could be difficult if memory is fragmented; in comparison, pages are the basic allocation unit for RAM, so 4kb's easy to find.
You do know that adsense keeps tracking cookies too, right?
That's init=/bin/sh, and it's hardly a bug - you're just changing the kernel configuration to skip all that silly authentication stuff. Someone who has access to the bootloader (you do set a bootloader password, right?) probably has enough access to override whatever security - short of disk encryption - you might have by force if necessary.
Compression algorithms designed around block repetition and relative frequency of occurrence of particular byte values don't typically work too well with audio files - when you've got a waveform, you'll have jitter that wrecks any exact repetition of data, and since it's basically going back and forth, you won't see much difference between how often 0x10 and 0x40 occur. Audio lossless compression takes advantage of the specific properties of audio - for example, FLAC does things like taking a linear prediction of the waveform over a small space, then recording the differences between its prediction and the actual value, in a form which optimizes for numbers close to zero.
The information in question is not reported back to the tracker; the tracker could only detect bad peers by running its own trusted peer to probe the network - however this would probably be impractical in practice.
In actual use, individual peers generally detect a bad peer quickly enough, and then ban it locally. This tends to work well enough in practice.
Still munged I'm afraid.