Last week in my country, a new political party overrun the previous party in charge of the municipality for about 30 years (yes, those thing happen in Europe sometimes).
The day the new government took charge, the displaced party dug out some four-year-old tweets containing a silly joke about nazis (the kind that would gather a +5 funny and some grammar nazi "corrected for you" replies around here) when the man had not even a politician. The same day, all the traditional media were reporting on their front pages as if it was the man's true opinion instead of a joke, reaching international press and forcing the councillor to resign (you may have heard about it as the "communist politician supporting the holocaust").
As long as the public falls for such obvious tactics, and until politicians learn to trim their twitter and facebook timelines when they run for office, this is bound to happen again and again.
Uhh, precedent. Double the resources, double the ability. This is well known.
That is magical thinking. It has no place in proper engineering practice.
There are zillions of reasons that interfere with ability to work faster in larger problems - yet they can be summarized with the words "non-linear growth". Try solving the Travelling salespeople problem twice as big with merely twice as fast hardware, it will slow to a grinch.
It's not like AI is going to run on some unknown substrate.
We know the substrate of brain power, gray cells. This doesn't mean that we understand the way they work together to create intelligence. If we ever create an AI, it will be so complex that we'll likely be in the same situation with respect to how it works.
"We" don't have to make one. All we have to do is set an AI towards self improvement/production of better AIs. THAT is where superintelligence comes from. All we have to do is make one that is an idiot savant geared toward AI design.
We do that with every new generation of babies, and it hasn't produced a super intelligence yet. What makes you think doing it ON A COMPUTER would make any difference?
There's no need for it to be abused. Just tag each POI with an <unconfirmed> tag and hide them on the map by default (but show them in searches) until they get reported by multiple users.
I couldn't care less about the TV except that it has to be diverting resources from improving the browser.
Does it? Improvements required to run the browser on a TV will necessarily involve making it more lightweight and portable, i.e. less dependent on the quirks of specific platforms.
No one else can because it will cause Google's CTS tool to fail verifying which won't allow you to ship with Google Play.
Conversely, if someone else built such system and it worked to keep all vendors updated, it wouldn't matter much that it failed to validate in Google's CTS. In that situation it would be relatively easy to migrate everyone away from Google Play -developers first, and users would follow- to an alternate app market supported by the maintainer of such successful system.
Now that I think of it, that would be the most likely way a strong contender might use to take control of Android from Google - in fact, that may be precisely what Microsot has in mind for their recent partnership with Cyanogen - the old Embrace, Extend, Extinguish.
I don't want a nanny-search moving the things I'm looking for down the page. Just give me what I searched for, nothing more, nothing less, no "judgment" about what I want to see.
So you'd rather have a complete database dump of all the web pages that contain your search terms, in random order, to do your own filtering among petabytes of data each time you look for "what you searched for, nothing less"?
'Cause any time you use a web search engine that provides just a few results, there *is* a judgement involved of which ones should appear at the first page; and Google is in the place it's now because their judgement was much better than any other search engine at the time. Including "mobile friendly" is only adding one more criterion to their ordering that they think will work well for a majority of their users.
the only reason people use Google is that it provides better search results
See my other comment above - the only reason it provides better search results is because they have more people using it, so other players can't provide a better nor cheaper product with respect to search.
That is not inherently bad as long as Google remains a monopoly by having better quality and limiting it to web search. But if they use that advantage to compete at other markets (such as advertising) not by having a better/cheaper product, but by exploiting all those users achieved through their natural monopoly, applying anti-monopoly laws makes sense.
First mover advantage plus the effect that having the majority of users can improve the quality of their results (in fact Google was *not* a first mover in the search space, but now they are entrenched). In the internet, code is law, and Google has a good amount of defining many technologies in widespread use - and more importantly, the way to learn about them.
A few weeks ago I read an analysis by a Mozilla blogger (which I can no longer find) of how, now that pagerank is less and less useful due to link farms and spammers otherwise attacking their algorithm, search quality depends largely on analysis of search terms introduced by users and the results they find interesting. This is a chicken-and-egg situation for any competitor: you can only improve your results by having more users, but they won't come if your results are not better than the market leader's. This is a natural monopoly, but one created by network effects and thus of the kind that can only be displaced by a disruptive process, not by regular competition.
And there's a similar effect for advertising - if there's a natural monopoly over the space were all users reside, then you must advertise in that platform in order to have enough eyeballs. It's the same mechanism that produced a lock-in for Facebook and Microsoft platforms back in the day - you go there not because the product is better, but because you need to interact with everyone who is doing the same.
In Europe we already tried allowing a winner-takes-it-all strategy where a very good leader keeps the monopoly over a (market/region/population), it was called an absolute monarchy.
It looks good for as long as the original manager (who reached the position as the best in a meritocracy) stays in place. It lasts for a generation, when the competent leader legates the role to their heirs, who may or may not be prepared to maintain the same level of quality service.
By that time, it is too late to displace the incompetent newcomers - all the network effects that entrenched the original leader as a monopoly are still in place and are too strong to overcome even when there are better alternatives, except by a disruptive process that redefines the rules of the game in full. I heard you Americans didn't like absolute monarchies? You should then understand the EU's position.
Every time some links to the obligatory xkcd, I remember the numbered jokes joke. Except that usually I don't need to follow the link nor see the number to know what joke it is referring to.
60% of the DNA is _definitely_ junk, as they consist of known repeated elements (LINEs, SINEs and others) and defunct genes. This is not an 'absence of evidence', we know exactly how this DNA has happened.
How do we know that those repetitions are not needed to accelerate (by parallel processing) some important process which, with a single expression, would otherwise be too slow to survive?
We don't fully understand how the phenotype is developed from the genotype, and it very well might depend on statistical properties of gene appearances in the genome, and not just their presence or absence. Is there something in biology science that could discard such possibility?
The world could have collaborated and built the modern Internet just fine on BSD licensed software, which is itself a variation of public domain.
True, but the nature of collaboration with BSD software would have been much more enterprise-y and committee-centered.
The idea of grass-roots FLOSS development only happened after Stallman's ideals of "giving back to the community" spread around with strong guarantees that their contributions would remain open, which didn't happen with the "public-domain-but-not-quite" that plagued the BSD-licensed but patent-encumbered UNIX systems.
I believe Stallman is credited for this because the average user never heard of Open Source or Free Software until the arrival of the GPL and its enabling of systems built with the Linux kernel and GNU userland.
You make it sound as if it was a coincidence that FLOSS took off at the very precise time that the GPL was created. It was not. The GPL is different from other permissive licenses in that it perpetuates the openness of the system it's used in; therefore it has a natural mechanism for survival of the project, that confers it an competitive advantage and makes it more robust against closed forks.
for whatever reason, FOSS and similar ideas were completely unknown to average users until the GPL took off.
This "whatever" reason may very well be that the GPL was created. The fact that source for all published modifications had to be released back to the project was a strong incentive way back for FLOSS developers to prefer contributing in GPL projects over BSD ones. It may not be that much relevant nowadays because the sharing culture has grown stronger, but the idea of "copyleft" and the "share and share-alike" was a necessary element to make it happen.
The Unix wars were still fresh in our minds, so we were all painfully aware of what happens when Big Megacorp forks a free system and releases a strong closed fork with commercial support, competing with the original project. In that context, copyleft is a guarantee that the software you've volunteered to will not be strangled by a strong vendor using the software without giving anything back.
The idea that big companies should support open source software as proof of their good will, even if the license doesn't require it, was mostly a consequence of the dynamics reinforced by the GPL mindset. Prior to that, BSD licensed UNIX vendors fought over the patents achieved on their particular variant and treated it as intellectual property to be defended, not as a shared resource to be cultivated.
There's a frequent misunderstanding when people talk about freedom with respect to the GPL. The concept of "freedom" is itself not well defined, and historically there are at least two competing and somewhat opposing definitions, "positive" freedom (which is about maximizing the amount of things people are able to do) and "negative" freedom (not interfering with things that others want to do). The GPL is primarily concerned about the former, and your complaints are about the latter.
The goal of the GPL is that everybody can use the software for any purpose, and learn how any changes and modifications work; this is seen as a requirement to increase the amount of things that can be done with the software, guaranteeing that it can be adapted to any hardware or platform, with no commercial secrets getting in the way.
In order to achieve this goal, the GPL doesn't come "free" (as in "gratis", i.e. no cost): it has a cost that you must pay if you want to use it; but for anyone willing to pay it, there are no further restrictions imposed by anyone, for any modified version. In your case you *could* have merged your application code with the GPL library* for any purpose**, but you should be willing to pay the cost, which is to release your own source code when you republish the software. So, your negative freedom is reduced (you are forbidden from keeping your version of the software hidden and publish just the binaries), but the positive freedom of the system is increased - overall there are more people who know how to use your modifications and adapt them for other uses, which couldn't happen if you kept your modifications secret.
The expectation is that by adding contributions from many users to the pool of knowledge, the whole society sees an increase in the amount of possibilities to use the software ("positive freedom"); it's the same principle that motivated the patent system in the Renaissance. The "release what you know" cost is intended to publish knowledge that otherwise would not be shared, and thus cumulatively improve the whole system. Now, there are valid concerns that the upfront cost may instead work as a disincentive to participate in the system (both with patents and copyleft software), but that argument doesn't make less free.
* Assuming the original license and copyright law allowed it. (This is why the FSF recommends using only GPL-compatible licenses). **(Including selling it, although with FLOSS software this typically only works once for each release).
In contrast, I've always thought that the primary concern was towards the interests of the software.
I agree with your view, but that terminology is anthropomorphizing an inanimate object, which IMHO makes it difficult to understand the benefits of that approach. If GPL achieves "what's best for the project", few people will care.
I've been recently describing what's good about the GPL by highlighting the knowledge about the software.
Compared to other FLOSS licenses, the GPL/copyleft ones, are the ones which best protect users interest to learn how the software-as-a project-works. Permissive licenses which allow users to close the source of their forks provide more individual freedom, at the cost of losing the knowledge about those published forks; with copyleft, that knowledge is preserved.
As you see, the logic of my explanation is the same (you maintain everything in the project), but providing a concrete reason why users of the software should care about keeping it evolving in the public sphere.
Ideals are not there to achieve convenience. They exist to steer us away from convenience, to avoid short-term gains that would push us into some long term dead-ends.
So ideals are not useful because we live by them on a day to day basis, but because they warn us when we deviate too far from them. Of course, having a few idealists that *do* live by their principles is a useful reminder for the rest of us that agree with them, but are nonetheless swayed by convenience.
Remember the Ubuntu phone? Remember what people were excited about regarding it? Notice how it hasn't been achieved due to various business cockblocks, thus leaving the gate wide open for someone bigger to step in? Hint hint.
Actually, there are a couple Ubuntu phones. Bq's Aquaris E4.5 was launched last month in Europe. (It's not the flagship device the Ubuntu Edge was suposed to be, but rather a semi-budget offering). And the Meizu MX4 Ubuntu Edition is a mid-range device.
Software testing doesn't protect against a user pressing the wrong button, which then works as expected. I agree it's a management error, but the failure in such cases is a lack of user testing.
Systems should be designed to follow the interactions that are more likely to be made by users, not the other way around - forcing users to follow the path that a developer thought would make sense. Unfortunately, user-centered design is still a foreign concept to a good chunk of developer houses.
Slashdot Mirror
Last week in my country, a new political party overrun the previous party in charge of the municipality for about 30 years (yes, those thing happen in Europe sometimes).
The day the new government took charge, the displaced party dug out some four-year-old tweets containing a silly joke about nazis (the kind that would gather a +5 funny and some grammar nazi "corrected for you" replies around here) when the man had not even a politician. The same day, all the traditional media were reporting on their front pages as if it was the man's true opinion instead of a joke, reaching international press and forcing the councillor to resign (you may have heard about it as the "communist politician supporting the holocaust").
As long as the public falls for such obvious tactics, and until politicians learn to trim their twitter and facebook timelines when they run for office, this is bound to happen again and again.
That is magical thinking. It has no place in proper engineering practice.
There are zillions of reasons that interfere with ability to work faster in larger problems - yet they can be summarized with the words "non-linear growth". Try solving the Travelling salespeople problem twice as big with merely twice as fast hardware, it will slow to a grinch.
We know the substrate of brain power, gray cells. This doesn't mean that we understand the way they work together to create intelligence. If we ever create an AI, it will be so complex that we'll likely be in the same situation with respect to how it works.
"We" don't have to make one. All we have to do is set an AI towards self improvement/production of better AIs. THAT is where superintelligence comes from. All we have to do is make one that is an idiot savant geared toward AI design.
We do that with every new generation of babies, and it hasn't produced a super intelligence yet. What makes you think doing it ON A COMPUTER would make any difference?
Let me guess. It involves the hunch of the elephant, the baking temperature of the refrigerator and the airspeed velocity of the unladen plane?
There's no need for it to be abused. Just tag each POI with an <unconfirmed> tag and hide them on the map by default (but show them in searches) until they get reported by multiple users.
Does it? Improvements required to run the browser on a TV will necessarily involve making it more lightweight and portable, i.e. less dependent on the quirks of specific platforms.
> Pixelated graphics do NOT look good.
Say what again?
Conversely, if someone else built such system and it worked to keep all vendors updated, it wouldn't matter much that it failed to validate in Google's CTS. In that situation it would be relatively easy to migrate everyone away from Google Play -developers first, and users would follow- to an alternate app market supported by the maintainer of such successful system.
Now that I think of it, that would be the most likely way a strong contender might use to take control of Android from Google - in fact, that may be precisely what Microsot has in mind for their recent partnership with Cyanogen - the old Embrace, Extend, Extinguish.
So you'd rather have a complete database dump of all the web pages that contain your search terms, in random order, to do your own filtering among petabytes of data each time you look for "what you searched for, nothing less"?
'Cause any time you use a web search engine that provides just a few results, there *is* a judgement involved of which ones should appear at the first page; and Google is in the place it's now because their judgement was much better than any other search engine at the time. Including "mobile friendly" is only adding one more criterion to their ordering that they think will work well for a majority of their users.
the only reason people use Google is that it provides better search results
See my other comment above - the only reason it provides better search results is because they have more people using it, so other players can't provide a better nor cheaper product with respect to search.
That is not inherently bad as long as Google remains a monopoly by having better quality and limiting it to web search. But if they use that advantage to compete at other markets (such as advertising) not by having a better/cheaper product, but by exploiting all those users achieved through their natural monopoly, applying anti-monopoly laws makes sense.
First mover advantage plus the effect that having the majority of users can improve the quality of their results (in fact Google was *not* a first mover in the search space, but now they are entrenched). In the internet, code is law, and Google has a good amount of defining many technologies in widespread use - and more importantly, the way to learn about them.
A few weeks ago I read an analysis by a Mozilla blogger (which I can no longer find) of how, now that pagerank is less and less useful due to link farms and spammers otherwise attacking their algorithm, search quality depends largely on analysis of search terms introduced by users and the results they find interesting. This is a chicken-and-egg situation for any competitor: you can only improve your results by having more users, but they won't come if your results are not better than the market leader's. This is a natural monopoly, but one created by network effects and thus of the kind that can only be displaced by a disruptive process, not by regular competition.
And there's a similar effect for advertising - if there's a natural monopoly over the space were all users reside, then you must advertise in that platform in order to have enough eyeballs. It's the same mechanism that produced a lock-in for Facebook and Microsoft platforms back in the day - you go there not because the product is better, but because you need to interact with everyone who is doing the same.
In Europe we already tried allowing a winner-takes-it-all strategy where a very good leader keeps the monopoly over a (market/region/population), it was called an absolute monarchy.
It looks good for as long as the original manager (who reached the position as the best in a meritocracy) stays in place. It lasts for a generation, when the competent leader legates the role to their heirs, who may or may not be prepared to maintain the same level of quality service.
By that time, it is too late to displace the incompetent newcomers - all the network effects that entrenched the original leader as a monopoly are still in place and are too strong to overcome even when there are better alternatives, except by a disruptive process that redefines the rules of the game in full. I heard you Americans didn't like absolute monarchies? You should then understand the EU's position.
The last one about centralized email is prophetic, if you change "penis enlargement" to "terrorism".
Every time some links to the obligatory xkcd, I remember the numbered jokes joke.
Except that usually I don't need to follow the link nor see the number to know what joke it is referring to.
How do we know that those repetitions are not needed to accelerate (by parallel processing) some important process which, with a single expression, would otherwise be too slow to survive?
We don't fully understand how the phenotype is developed from the genotype, and it very well might depend on statistical properties of gene appearances in the genome, and not just their presence or absence. Is there something in biology science that could discard such possibility?
True, but the nature of collaboration with BSD software would have been much more enterprise-y and committee-centered.
The idea of grass-roots FLOSS development only happened after Stallman's ideals of "giving back to the community" spread around with strong guarantees that their contributions would remain open, which didn't happen with the "public-domain-but-not-quite" that plagued the BSD-licensed but patent-encumbered UNIX systems.
You make it sound as if it was a coincidence that FLOSS took off at the very precise time that the GPL was created. It was not. The GPL is different from other permissive licenses in that it perpetuates the openness of the system it's used in; therefore it has a natural mechanism for survival of the project, that confers it an competitive advantage and makes it more robust against closed forks.
This "whatever" reason may very well be that the GPL was created. The fact that source for all published modifications had to be released back to the project was a strong incentive way back for FLOSS developers to prefer contributing in GPL projects over BSD ones. It may not be that much relevant nowadays because the sharing culture has grown stronger, but the idea of "copyleft" and the "share and share-alike" was a necessary element to make it happen.
The Unix wars were still fresh in our minds, so we were all painfully aware of what happens when Big Megacorp forks a free system and releases a strong closed fork with commercial support, competing with the original project. In that context, copyleft is a guarantee that the software you've volunteered to will not be strangled by a strong vendor using the software without giving anything back.
The idea that big companies should support open source software as proof of their good will, even if the license doesn't require it, was mostly a consequence of the dynamics reinforced by the GPL mindset. Prior to that, BSD licensed UNIX vendors fought over the patents achieved on their particular variant and treated it as intellectual property to be defended, not as a shared resource to be cultivated.
And thus comes the danger of having all your logged-in email addresses accessible to whomever steals you phone, which was my original point.
So, you've never encountered a site with a "I've forgotten my password" option that sends you a mail to log in?
Anyway, it's bad enough that a thief can access all data in the logged in service even if they can't change the password.
But it's more than enough time to access all the services to which you're logged in in your browser, and possibly change your password in them.
There's a frequent misunderstanding when people talk about freedom with respect to the GPL. The concept of "freedom" is itself not well defined, and historically there are at least two competing and somewhat opposing definitions, "positive" freedom (which is about maximizing the amount of things people are able to do) and "negative" freedom (not interfering with things that others want to do). The GPL is primarily concerned about the former, and your complaints are about the latter.
The goal of the GPL is that everybody can use the software for any purpose, and learn how any changes and modifications work; this is seen as a requirement to increase the amount of things that can be done with the software, guaranteeing that it can be adapted to any hardware or platform, with no commercial secrets getting in the way.
In order to achieve this goal, the GPL doesn't come "free" (as in "gratis", i.e. no cost): it has a cost that you must pay if you want to use it; but for anyone willing to pay it, there are no further restrictions imposed by anyone, for any modified version. In your case you *could* have merged your application code with the GPL library* for any purpose**, but you should be willing to pay the cost, which is to release your own source code when you republish the software. So, your negative freedom is reduced (you are forbidden from keeping your version of the software hidden and publish just the binaries), but the positive freedom of the system is increased - overall there are more people who know how to use your modifications and adapt them for other uses, which couldn't happen if you kept your modifications secret.
The expectation is that by adding contributions from many users to the pool of knowledge, the whole society sees an increase in the amount of possibilities to use the software ("positive freedom"); it's the same principle that motivated the patent system in the Renaissance. The "release what you know" cost is intended to publish knowledge that otherwise would not be shared, and thus cumulatively improve the whole system. Now, there are valid concerns that the upfront cost may instead work as a disincentive to participate in the system (both with patents and copyleft software), but that argument doesn't make less free.
* Assuming the original license and copyright law allowed it. (This is why the FSF recommends using only GPL-compatible licenses).
**(Including selling it, although with FLOSS software this typically only works once for each release).
In contrast, I've always thought that the primary concern was towards the interests of the software.
I agree with your view, but that terminology is anthropomorphizing an inanimate object, which IMHO makes it difficult to understand the benefits of that approach. If GPL achieves "what's best for the project", few people will care.
I've been recently describing what's good about the GPL by highlighting the knowledge about the software.
Compared to other FLOSS licenses, the GPL/copyleft ones, are the ones which best protect users interest to learn how the software-as-a project-works. Permissive licenses which allow users to close the source of their forks provide more individual freedom, at the cost of losing the knowledge about those published forks; with copyleft, that knowledge is preserved.
As you see, the logic of my explanation is the same (you maintain everything in the project), but providing a concrete reason why users of the software should care about keeping it evolving in the public sphere.
Convenience trumps ideals more often than not.
Ideals are not there to achieve convenience. They exist to steer us away from convenience, to avoid short-term gains that would push us into some long term dead-ends.
So ideals are not useful because we live by them on a day to day basis, but because they warn us when we deviate too far from them. Of course, having a few idealists that *do* live by their principles is a useful reminder for the rest of us that agree with them, but are nonetheless swayed by convenience.
Actually, there are a couple Ubuntu phones. Bq's Aquaris E4.5 was launched last month in Europe. (It's not the flagship device the Ubuntu Edge was suposed to be, but rather a semi-budget offering). And the Meizu MX4 Ubuntu Edition is a mid-range device.
Software testing doesn't protect against a user pressing the wrong button, which then works as expected. I agree it's a management error, but the failure in such cases is a lack of user testing.
Systems should be designed to follow the interactions that are more likely to be made by users, not the other way around - forcing users to follow the path that a developer thought would make sense. Unfortunately, user-centered design is still a foreign concept to a good chunk of developer houses.