I used to work for a defence contractor here in Australia, and even though our social media policies explicitly said not to advertise who you worked for or which projects you worked on, there was this strange blind spot where LinkedIn was concerned.
Even though heaps of staff had their affiliation clearly displayed there, management couldn't really complain about it because they were doing the same thing. It was thoroughly bizarre.
This ad that the UK's "The Guardian" put together last year is still one of the most powerful representations I've seen of the future role of the media as summarising and providing a window into a broader public conversation happening online.
One thing to keep in mind is that good definitions of "consumer harm" include long term consequences and systemic risk, not just immediate effects. That's why predatory pricing is illegal after all - the short term effects for consumers are positive (hey, cheap stuff!), but the longer term ones are harmful (driving the competition out of business in order to later obtain monopoly rents)
There's a lot of risk to be found in outsourcing key parts of your infrastructure to a foreign corporation. The unfortunate part is most of this isn't addressed in the law, so other laws (like anti-trust) get twisted to suit :P
people loathe software that requires internet access to use.
We don't loathe all software that requires on internet connection, just software that requires one for no good reason.
In this case, if you buy this mouse to use with one gaming computer that never leaves your house... you still have to register it to get it to work properly. WTF?
All they needed to do to avoid this problem is to provide an option to store settings locally *or* upload them to the cloud for portability. This is not hard, and the fact they don't offer this capability makes their attempted justifications of the software as being for the benefit of the user self-evidently BS.
There's a much better solution to the "reading times" problem: just include an estimated reading time at the top of the article! Even a naive estimate based on "word count / typical reading speed" is enough of a hint.
I also like how their two "justifications" fight each: a long scroll bar is scary and intimidating and puts people off, but a high page count is a useful indicator of the reading time needed...
You're confusing a couple of different things there. Open source is a *production* method, safety checks are a *certification* requirement.
A certification may cover your production methods (and certainly many safety certifications will include such requirements). If a particular design has been certified, but your production tools have not, then your plane won't be certified. Similarly, if you change your design, then the cert will no longer apply.
So you're unlikely to see the FAA certifying an open source passenger plane design as safe, as FAA certification for passenger planes involves a lot more than that. However, you may see open source entries into the "kit plane" market, as the quality and certification standards are much lower in that market (since fewer people are put at risk in personal aviation than in commercial aviation - it's more about personal trust than it is about certification).
Similar issues exist with open source software - many certs only apply to particular builds of software from particular vendors. If you rebuild it yourself, the cert technically no longer applies, so if you have a strict "must be certified" requirement, you either have to seek recertification for your rebuilt copy or else just use one of the certified versions.
(Disclosure: I currently work for Red Hat, and the validity of various certifications is one of the selling points of our subscriptions, even though the source code of our products is freely available and redistributable through open source licensing)
I've noticed lately that I almost never type out a full domain name manually. Instead, I will either:
1. Start typing the domain, then choose the actual site from Firefox's internal history search
2. Follow a link
3. Do a Google search for the company name
The only case where I even consider typing out the full URL is if I got it from an offline source, and even there I'm more likely to use a search for the core name rather than typing out the whole thing.
Phones are also getting to the point where they can read a URL through the camera, without the assistance of a QR code.
On top of that, we can add the fact that for many "web applications" the tool we actually use to access it is a dedicated smartphone utility - the namespace is the name in the relevant app store, not the DNS host of the site itself.
All ICANN is doing is hastening the conversion of DNS from a scheme that is a visible part of the main UI of the internet, to an abstraction layer over IP addresses that is used as a hint by search engines and a stable record for consumption by other software (such as other web services, client applications, browser history recording and automatic URL recognition tools)
While there would still be a lot of dross to clear out, one key part of an independent invention defence is that "Will this prevail over an independent invention defence?" will become one of the questions companies ask themselves before going to the expense of filing for a patent.
One part of fixing the examination process is to get companies to stop filing for patents that they know are rubbish because proving obviousness is often more expensive than just taking a license for a bogus patent. If there is an onus on the patent holder to prove "they copied our idea" rather than "they had the same idea we had", then getting a patent just for the sake of having the patent becomes significantly less attractive.
The excesses of GEMA are so outlandish, I sometimes wonder if other copyright maximalists want to yell at them "YOU'RE NOT HELPING! PLEASE QUIT SCREWING IT UP FOR THE REST OF US!"
As Beech said, a lot of people genuinely buy into the "property" model for abstract ideas.
"I said it, therefore it's mine and you can't use it without paying me or getting my permission"
"I thought it, therefore it's mine and you can't use it without paying me or getting my permission"
Only once we fully recognise that it isn't *copying itself* that's morally wrong, but *deception regarding the origins of a work*, can we ever move on to creating a legal system that makes more sense.
Trademark law, when properly applied to the task of preventing consumer confusion, is already there. It's primarily about it being illegal to mislead people about the source of goods and services in order to misappropriate someone else's reputation.
Patent law would take a long step down that road by including an "independent invention" defence, with plaintiffs having to provide reasonable evidence that copying was likely before a court case could proceed.
Copyright law needs to be refocused more on preventing plagiarism, rather than all copying. Such an approach would essentially make "CC-BY" the only state enforced level of agreement, with any further rights being handled either by contractual agreements, or else by becoming part of the realm of social, rather than legal, enforcement.
Much of the open source world is moving this way: a lot of the *code* is available under very permissive licenses, but the *names* are controlled under trademark law. If all you want to do is use the software, no problem, but if you want to make use of the *name* without complying with the organisation's trademark usage policy, then you're asking for legal trouble.
Sure, trademark law is open to its own abuses (especially aggressive trademark holders starting lawsuits over nominative use, or other cases where there is no chance for confusion), but it's anti-fraud foundation means it is in much better shape than either patent or copyright law.
Thinking about it further, it appears from some of the notes here and on Groklaw that Google have *conceded* that the details of that argument checking function were the result of a cleanroom breach, so they're clearly not planning to argue the independent development angle.
They may figure it isn't worth trying to persuade the jury of the "not even creative" aspect when there are so many simpler points they can use to prevail (such as the 9 lines out of 15 million aspect). Definite shades of SCO there.
No, I had a look - the structure of the C code in Tim's original implementation for CPython is completely different.
The only relation the lawsuit has to timsort is that it happens to be the file where the allegedly infringing offending Android code lives.
It's just amusing to me that Oracle is claiming "OMG infringement" over a completely trivial range checking function that Google used as part of their implemention of Tim's genuinely innovative, but freely shared, sorting algorithm.
Ah, my mistake - that link is to the OpenJDK version of the Array.java file, so that's the likely origin of the "length" vs "arrayLen" argument name difference. Still, none of "length", "len", "arrayLen" and "arrayLength" can really be called creative when it comes to naming a parameter that is "the length of the array".
If that function is the only copying they can prove, then it's not a derivative work. There's zero creativity in that function - it's driven entirely be the exposed java.util.Array API, normal Java conventions and implementing the necesssary checks efficiently.
Exception types and message contents? Part of the API - visible just by *using* the API, not looking at its implementation.
"rangeCheck()" - no creativity. Literal description of the function's purpose, named according to Java conventions.
"fromIndex, toIndex" - taken from the corresponding argument names in the public Array.binarySearch() API
function structure - merely a logical way of minimising the number of checks needed:
- is the start index after the end index?
- is the start index negative?
- is the end index past the end of the array?
By doing the checks in this order, you don't have to check:
- is the end index negative? (if this is the case, one of the first two checks will already fail)
- is the start index past the end of the array? (if this is the case, either the first or last check will already fail)
Yeah, when I saw that slide my reaction was "Wait, what? That can't be copyright worthy - it's a completely trivial slice range checking operation."
Even the variable names aren't creative - they're literal descriptions of what the arguments are for.
Also, *chortle* at the fact TimSort is one of the files they're complaining about. The current Java sorting algorithm is a genuinely innovative one that Tim Peters wrote for Python and then made available to the world *for free* under an open source licence.
That's completely irrelevant to the arguments in court, but it's still hilariously ironic :)
The LinkedIn blind spot
Even though heaps of staff had their affiliation clearly displayed there, management couldn't really complain about it because they were doing the same thing. It was thoroughly bizarre.
Cynical interpretation
The Guardian's Three Little Pigs ad
Trial data?
See: http://www.alltrials.net/
Immediate consumer harm vs systemic risk
There's a lot of risk to be found in outsourcing key parts of your infrastructure to a foreign corporation. The unfortunate part is most of this isn't addressed in the law, so other laws (like anti-trust) get twisted to suit :P
So, powerful GOP supporters with a vested interest in the copyright status quo?
...for no good reason
In this case, if you buy this mouse to use with one gaming computer that never leaves your house... you still have to register it to get it to work properly. WTF?
All they needed to do to avoid this problem is to provide an option to store settings locally *or* upload them to the cloud for portability. This is not hard, and the fact they don't offer this capability makes their attempted justifications of the software as being for the benefit of the user self-evidently BS.
Estimated reading times
I also like how their two "justifications" fight each: a long scroll bar is scary and intimidating and puts people off, but a high page count is a useful indicator of the reading time needed...
Production methods and certification are two different things
A certification may cover your production methods (and certainly many safety certifications will include such requirements). If a particular design has been certified, but your production tools have not, then your plane won't be certified. Similarly, if you change your design, then the cert will no longer apply.
So you're unlikely to see the FAA certifying an open source passenger plane design as safe, as FAA certification for passenger planes involves a lot more than that. However, you may see open source entries into the "kit plane" market, as the quality and certification standards are much lower in that market (since fewer people are put at risk in personal aviation than in commercial aviation - it's more about personal trust than it is about certification).
Similar issues exist with open source software - many certs only apply to particular builds of software from particular vendors. If you rebuild it yourself, the cert technically no longer applies, so if you have a strict "must be certified" requirement, you either have to seek recertification for your rebuilt copy or else just use one of the certified versions.
(Disclosure: I currently work for Red Hat, and the validity of various certifications is one of the selling points of our subscriptions, even though the source code of our products is freely available and redistributable through open source licensing)
The era of typing full domain names manually is over
1. Start typing the domain, then choose the actual site from Firefox's internal history search
2. Follow a link
3. Do a Google search for the company name
The only case where I even consider typing out the full URL is if I got it from an offline source, and even there I'm more likely to use a search for the core name rather than typing out the whole thing.
Phones are also getting to the point where they can read a URL through the camera, without the assistance of a QR code.
We also know that there are already *many* people that don't even use DNS for sites they visit regularly: http://www.readwriteweb.com/archives/how_google_failed_internet_meme.php
On top of that, we can add the fact that for many "web applications" the tool we actually use to access it is a dedicated smartphone utility - the namespace is the name in the relevant app store, not the DNS host of the site itself.
All ICANN is doing is hastening the conversion of DNS from a scheme that is a visible part of the main UI of the internet, to an abstraction layer over IP addresses that is used as a hint by search engines and a stable record for consumption by other software (such as other web services, client applications, browser history recording and automatic URL recognition tools)
Re: Independent invention.
One part of fixing the examination process is to get companies to stop filing for patents that they know are rubbish because proving obviousness is often more expensive than just taking a license for a bogus patent. If there is an onus on the patent holder to prove "they copied our idea" rather than "they had the same idea we had", then getting a patent just for the sake of having the patent becomes significantly less attractive.
Re: Re:
Secretly anti-copyright?
I think I recall an actress saying something relevant in a film I saw...
More pessimistically, anyone want to place bets on how long it will be until Carrie Fisher has the legal right to get Mike to delete this comment? :P
Rejecting the property model
"I said it, therefore it's mine and you can't use it without paying me or getting my permission"
"I thought it, therefore it's mine and you can't use it without paying me or getting my permission"
Only once we fully recognise that it isn't *copying itself* that's morally wrong, but *deception regarding the origins of a work*, can we ever move on to creating a legal system that makes more sense.
Trademark law, when properly applied to the task of preventing consumer confusion, is already there. It's primarily about it being illegal to mislead people about the source of goods and services in order to misappropriate someone else's reputation.
Patent law would take a long step down that road by including an "independent invention" defence, with plaintiffs having to provide reasonable evidence that copying was likely before a court case could proceed.
Copyright law needs to be refocused more on preventing plagiarism, rather than all copying. Such an approach would essentially make "CC-BY" the only state enforced level of agreement, with any further rights being handled either by contractual agreements, or else by becoming part of the realm of social, rather than legal, enforcement.
Much of the open source world is moving this way: a lot of the *code* is available under very permissive licenses, but the *names* are controlled under trademark law. If all you want to do is use the software, no problem, but if you want to make use of the *name* without complying with the organisation's trademark usage policy, then you're asking for legal trouble.
Sure, trademark law is open to its own abuses (especially aggressive trademark holders starting lawsuits over nominative use, or other cases where there is no chance for confusion), but it's anti-fraud foundation means it is in much better shape than either patent or copyright law.
Re: Re: Jurors are no programmers
Thinking about it further, it appears from some of the notes here and on Groklaw that Google have *conceded* that the details of that argument checking function were the result of a cleanroom breach, so they're clearly not planning to argue the independent development angle.
They may figure it isn't worth trying to persuade the jury of the "not even creative" aspect when there are so many simpler points they can use to prevail (such as the 9 lines out of 15 million aspect). Definite shades of SCO there.
Re: Re: Re: De minimis or not...
The only relation the lawsuit has to timsort is that it happens to be the file where the allegedly infringing offending Android code lives.
It's just amusing to me that Oracle is claiming "OMG infringement" over a completely trivial range checking function that Google used as part of their implemention of Tim's genuinely innovative, but freely shared, sorting algorithm.
Re: Re:
Re:
Exception types and message contents? Part of the API - visible just by *using* the API, not looking at its implementation.
"rangeCheck()" - no creativity. Literal description of the function's purpose, named according to Java conventions.
"arrayLen" - this looks like it may actually be a dubious assertion on Oracle's part. It's called "length" in the version published here: http://www.docjar.com/html/api/java/util/Arrays.java.html
Google's lawyers should really look into that.
"fromIndex, toIndex" - taken from the corresponding argument names in the public Array.binarySearch() API
function structure - merely a logical way of minimising the number of checks needed:
- is the start index after the end index?
- is the start index negative?
- is the end index past the end of the array?
By doing the checks in this order, you don't have to check:
- is the end index negative? (if this is the case, one of the first two checks will already fail)
- is the start index past the end of the array? (if this is the case, either the first or last check will already fail)
Re: De minimis or not...
Even the variable names aren't creative - they're literal descriptions of what the arguments are for.
Also, *chortle* at the fact TimSort is one of the files they're complaining about. The current Java sorting algorithm is a genuinely innovative one that Tim Peters wrote for Python and then made available to the world *for free* under an open source licence.
That's completely irrelevant to the arguments in court, but it's still hilariously ironic :)
[1] http://hg.python.org/cpython/file/default/Objects/listsort.txt
More comments from Nick Coghlan >>
Techdirt has not posted any stories submitted by Nick Coghlan.
Submit a story now.
Tools & Services
TwitterFacebook
RSS
Podcast
Research & Reports
Company
About UsAdvertising Policies
Privacy
Contact
Help & FeedbackMedia Kit
Sponsor/Advertise
Submit a Story
More
Copia InstituteInsider Shop
Support Techdirt