Those other companies may be larger in total sales, but they have a smaller online presence; Newegg is currently the 13th largest online retailer. Probably explains why the other companies settled rather than go to trial; online sales likely weren't enough of a priority to them.
RTFA. The patents were overly broad/obvious, and there was prior art. Other companies who were approached by Soverain settled because they didn't want to get drawn into a lengthy and expensive legal battle. Newegg stood their ground, and ultimately prevailed.
Yes, Soverain was a "legitimate patent holder" in the sense that they legitimately owned the patents in question. But the patents themselves were not legitimate (in the sense of embodying anything original or unique).
I suspect that one of the reasons Newegg stood their ground is that -- unlike most of the other companies mentioned in the article -- they are exclusively an online operation, and therefore had more at stake.
I just got rid of my Belkin UPS. The batteries gave out, and I do not feel like spending any money to replace them. There was something seriously wrong with the design of the charging circuitry; it ran hot even when it wasn't using the inverter.
Besides being heavily into computer tech, I also brew beer. For the past decade or so New Years Eve has been spent with a group of local beer geeks (brewers and judges). The guy who hosts it always has a keg of something great (this year it was Firestone Walker Union Jack), and also does a 10+ year vertical of Sierra Nevada Bigfoot. Everybody else brings a few bottles of something to share. We've probably got better beer than any other New Years party in town!
FWIW I've had good luck with Asus, for both servers and desktops. Over a dozen Linux systems built using Asus motherboards so far. Guess I just haven't hit any of the ones with problematic chips on them.
"It always takes longer than you expect, even when you take into account Hofstadter's Law."
Re:A good example of a bad summary
on
Qt 5.0 Released
·
· Score: 2
Agreed. I've never developed anything using Qt, but have a pretty good idea of what it is. Even if I'd never heard of it before, a couple of mouse clicks and 3 keystrokes (assuming I've got Google or Wikipedia bookmarked) would've yielded an answer in seconds. Given this, and the fact that most people on/. probably already have at least a vague notion of what Qt does, I don't think it is necessary to explain it every time it comes up.
I may have mis-interpreted the post I replied to. I thought he was referring to sanity checks in software; upon re-reading he may have been referring to ECC RAM. I agree, the performance penalty for ECC RAM is minimal (and I use it in most of the systems I build).
It's just a matter of whether you realize it or not.
The blatant ones cause an application or OS crash. But depending on what got corrupted, it might just cause a momentary application glitch, or even cause an alteration in the contents of a file that you won't notice for weeks... if ever.
When I build PCs, they get an overnight Memtest run at a minimum. Most of the time I also use ECC RAM to protect against random flipped bits and DIMMs that fail after being in use for a while.
If every value read from memory had to be sanity checked there would be little CPU horsepower left to perform useful work. Furthermore, if the bad value happens to be code instead of data, the application (or OS) is probably gonna crash before you even have a chance to check anything.
Subtle point here: Under 3b it seems to me that the offer only needs to be made directly to the person who paid for the binary. That offer can indeed be passed along by that purchaser (and if this is done it must be honored from 3rd parties by the original seller), but the original seller has no obligation to make the offer directly to anyone else. So under this interpretation, a random Joe can't just contact the seller and demand a copy of the source code unless they've received a copy of the original offer via someone who bought the binary.
I admit it's a distinction that means little in today's highly connected world, given that the offer (and/or corresponding source code) can just be posted on the Internet by anyone; but it mattered more back when the GPL was originally devised.
Good point. However this also shows that the "Canonical doesn't put enough back to the Linux community" complaint we hear so much is silly. If this was true then distros like Mint would not use them as a starting point.
Replying to my own post... oops, I see the author of DOSBox turbo had already replied. So I guess the answer is, source was not being supplied with the binary, but will be shortly. So any concerns about DOSBox Turbo being GPL compliant - regardless of how anyone interprets section 3 - will soon be put to rest.
But regardless of who it is catering to, by choosing to create a derivative work based on GPL code you are implicitly agreeing to the GPL's terms. The original author(s) of DOSBox chose GPL as its license; whether or not most DOSBox users actually care about the source code is irrelevant.
It is unclear from the original Ask Slashdot whether source is included with the binary distribution of DOSBox Turbo.
I also don't see how this prevents anyone from charging a full binary license fee for the source, since GPL can be satisfied by providing the source with the binary -- do that, and you've met the requirement of Section 3 without any requirement to distribute source to anyone else.
Bottom line: The requirement to provide source code does not kick in unless you're distributing the binary to the person who is asking for the source code; the person you've given the source code is then free to redistribute it if they wish. Most people deal with this part of the GPL by simply posting the code publicly (accessible to anyone who wants it), but there is no requirement to do so.
So DOSBox Turbo is in fact abiding by the GPL, while DOSBot is not.
I think you're mis-reading Section 3. It doesn't say you need to distribute the source on demand to ANYBODY; it says you must either distribute it with the binary, provide an offer with the binary to distribute it for the cost of the media, or pass along the offer you received from upstream (if you're just re-distributing a binary you didn't compile). It seems pretty clear to me that this is only meant to apply if you're giving/selling the binary to someone. You are not compelled to give the source code away for free independent of the binary. (But anyone you sell the binary to can ask for a copy of the source, then re-distribute the source themselves if they wish.)
Based on a literal reading of Section 3 of the GPLv2, I think the guy you're replying to is actually correct. The offer of source code only needs to be made to the person who bought the object code. But... and it's a big but... that buyer of the object code has the right to extend the source code offer to anyone else if they want.
True. But the way that is worded, you still need to find someone else who had already bought the binary and received the source code offer, since it only says the offer must accompany the binary.
So I think the DOSBox Turbo guy is technically still within the letter of the GPL. DOSBot definitely sounds like a GPL violation though.
It's not being burned, it's only being used as a heat carrier. Seems to me it would be more efficient to just heat the water directly, and use it in a steam turbine. What am I missing here?
"Software Engineering" implies a more practical focus, with emphasis on using real teams to create working systems with real-world applicability. "Computer Science" implies a more theoretical approach, perhaps a bit removed from practical application. Both have their place, and there's huge overlap (one institution's Comp Sci program may even be more "practical" than another's Software Engineering program), so it is hard to make snap judgements.
But in general, putting someone with the Software Engineering degree in charge sounds like a good idea to me, as long as their are competent Comp Sci people working under her.
Slashdot Mirror
Those other companies may be larger in total sales, but they have a smaller online presence; Newegg is currently the 13th largest online retailer. Probably explains why the other companies settled rather than go to trial; online sales likely weren't enough of a priority to them.
What is this strange phrase you use, "don't need"? There's always something any self-respecting geek needs from Newegg!
RTFA. The patents were overly broad/obvious, and there was prior art. Other companies who were approached by Soverain settled because they didn't want to get drawn into a lengthy and expensive legal battle. Newegg stood their ground, and ultimately prevailed.
Yes, Soverain was a "legitimate patent holder" in the sense that they legitimately owned the patents in question. But the patents themselves were not legitimate (in the sense of embodying anything original or unique).
I suspect that one of the reasons Newegg stood their ground is that -- unlike most of the other companies mentioned in the article -- they are exclusively an online operation, and therefore had more at stake.
I just got rid of my Belkin UPS. The batteries gave out, and I do not feel like spending any money to replace them. There was something seriously wrong with the design of the charging circuitry; it ran hot even when it wasn't using the inverter.
Besides being heavily into computer tech, I also brew beer. For the past decade or so New Years Eve has been spent with a group of local beer geeks (brewers and judges). The guy who hosts it always has a keg of something great (this year it was Firestone Walker Union Jack), and also does a 10+ year vertical of Sierra Nevada Bigfoot. Everybody else brings a few bottles of something to share. We've probably got better beer than any other New Years party in town!
FWIW I've had good luck with Asus, for both servers and desktops. Over a dozen Linux systems built using Asus motherboards so far. Guess I just haven't hit any of the ones with problematic chips on them.
SVM enabled in your BIOS?
"It always takes longer than you expect, even when you take into account Hofstadter's Law."
Agreed. I've never developed anything using Qt, but have a pretty good idea of what it is. Even if I'd never heard of it before, a couple of mouse clicks and 3 keystrokes (assuming I've got Google or Wikipedia bookmarked) would've yielded an answer in seconds. Given this, and the fact that most people on /. probably already have at least a vague notion of what Qt does, I don't think it is necessary to explain it every time it comes up.
I may have mis-interpreted the post I replied to. I thought he was referring to sanity checks in software; upon re-reading he may have been referring to ECC RAM. I agree, the performance penalty for ECC RAM is minimal (and I use it in most of the systems I build).
It's just a matter of whether you realize it or not.
The blatant ones cause an application or OS crash. But depending on what got corrupted, it might just cause a momentary application glitch, or even cause an alteration in the contents of a file that you won't notice for weeks... if ever.
When I build PCs, they get an overnight Memtest run at a minimum. Most of the time I also use ECC RAM to protect against random flipped bits and DIMMs that fail after being in use for a while.
If every value read from memory had to be sanity checked there would be little CPU horsepower left to perform useful work. Furthermore, if the bad value happens to be code instead of data, the application (or OS) is probably gonna crash before you even have a chance to check anything.
Subtle point here: Under 3b it seems to me that the offer only needs to be made directly to the person who paid for the binary. That offer can indeed be passed along by that purchaser (and if this is done it must be honored from 3rd parties by the original seller), but the original seller has no obligation to make the offer directly to anyone else. So under this interpretation, a random Joe can't just contact the seller and demand a copy of the source code unless they've received a copy of the original offer via someone who bought the binary.
I admit it's a distinction that means little in today's highly connected world, given that the offer (and/or corresponding source code) can just be posted on the Internet by anyone; but it mattered more back when the GPL was originally devised.
Good point. However this also shows that the "Canonical doesn't put enough back to the Linux community" complaint we hear so much is silly. If this was true then distros like Mint would not use them as a starting point.
Erm... if you're running Mint, technically you are using Ubuntu. Unless you're running the Mint Debian Edition.
Replying to my own post... oops, I see the author of DOSBox turbo had already replied. So I guess the answer is, source was not being supplied with the binary, but will be shortly. So any concerns about DOSBox Turbo being GPL compliant - regardless of how anyone interprets section 3 - will soon be put to rest.
But regardless of who it is catering to, by choosing to create a derivative work based on GPL code you are implicitly agreeing to the GPL's terms. The original author(s) of DOSBox chose GPL as its license; whether or not most DOSBox users actually care about the source code is irrelevant.
It is unclear from the original Ask Slashdot whether source is included with the binary distribution of DOSBox Turbo.
I also don't see how this prevents anyone from charging a full binary license fee for the source, since GPL can be satisfied by providing the source with the binary -- do that, and you've met the requirement of Section 3 without any requirement to distribute source to anyone else.
Bottom line: The requirement to provide source code does not kick in unless you're distributing the binary to the person who is asking for the source code; the person you've given the source code is then free to redistribute it if they wish. Most people deal with this part of the GPL by simply posting the code publicly (accessible to anyone who wants it), but there is no requirement to do so.
So DOSBox Turbo is in fact abiding by the GPL, while DOSBot is not.
I think you're mis-reading Section 3. It doesn't say you need to distribute the source on demand to ANYBODY; it says you must either distribute it with the binary, provide an offer with the binary to distribute it for the cost of the media, or pass along the offer you received from upstream (if you're just re-distributing a binary you didn't compile). It seems pretty clear to me that this is only meant to apply if you're giving/selling the binary to someone. You are not compelled to give the source code away for free independent of the binary. (But anyone you sell the binary to can ask for a copy of the source, then re-distribute the source themselves if they wish.)
Nobody is being forced to do anything. If you don't want to distribute under the GPL, then don't distribute code that is derived from other GPL code.
Based on a literal reading of Section 3 of the GPLv2, I think the guy you're replying to is actually correct. The offer of source code only needs to be made to the person who bought the object code. But... and it's a big but... that buyer of the object code has the right to extend the source code offer to anyone else if they want.
True. But the way that is worded, you still need to find someone else who had already bought the binary and received the source code offer, since it only says the offer must accompany the binary.
So I think the DOSBox Turbo guy is technically still within the letter of the GPL. DOSBot definitely sounds like a GPL violation though.
It's not being burned, it's only being used as a heat carrier. Seems to me it would be more efficient to just heat the water directly, and use it in a steam turbine. What am I missing here?
"Software Engineering" implies a more practical focus, with emphasis on using real teams to create working systems with real-world applicability. "Computer Science" implies a more theoretical approach, perhaps a bit removed from practical application. Both have their place, and there's huge overlap (one institution's Comp Sci program may even be more "practical" than another's Software Engineering program), so it is hard to make snap judgements.
But in general, putting someone with the Software Engineering degree in charge sounds like a good idea to me, as long as their are competent Comp Sci people working under her.