If my completion is predictable, I know what I have to type without looking at the screen.
If it is smart, I must always check if what I want to do is in the completion list.
If it is too smart, it will propose complex completions, that are correct 80% of the time, and look correct 90% of the time at first glance.
The "typos" I make today with smart typing correction in my smart phone look to other people like correct sentences. It's just confusing. No one can guess what I meant to write.
I've seen them, and I hate them. I hated them from the first time I saw them. Actually, so much that I always (successfully) tried to avoid writing one. I think once I used a template.
Just look at this fuckery. I have to write the PID logic myself? Or I have to copy code? This is bullshit. I don't wan't to do this.
Once I've read a thread about X, and one comment said "If you have a UI, it's not Linux anymore!". Init scripts are for those people...
Never having to deal with init scripts before, I had to install a custom service on a server. My company was using custom bash scripts to restart processes that die. They also have some wierd mechanism to suspend the restart behavior, and some really bad email reporting. I wrote a systemd script for my service and was quite happy with it. I could even configure sending an email on crash rather easily.
When talking to the sysadmin of our company, it appears to me that he doesn't like systemd, just because it does some things differently compared to as it was before. That's a pretty weak argument.
Just buy a phone with a replaceable battery. Done! There are many. I have a Moto G4 Play, which has replaceable battery. No need to bother with AAA batteries – that's really a stupid idea.
This sounds much like the "If you know how to program, you can pick up any language quickly." kind of thinking. I encounter it often, and I also thought this way right after finishing university. But it's simply not this way. Maybe all imperative languages are alike, yes. And if you know imperative, and OOP you can pick up many languages quickly. You will not learn anything new doing so, though. But if you change paradigm, you will find out that there is much you don't know.
SEL is a 1998 anime. It is full of power line shots. I'd estimate that around 2-5 percent of the series consist of power lines. It would be interesting whether this was the start of the trend. Can someone please categorize this Tumblr thing into pre and post 1998, please?
I'm about to finish my PhD in a subject closely related to Deep Learning. I don't find an appealing job here in Germany without relocation. So I'll go on doing embedded development.
Even with the best AI, we need to give it an objective to try to accomplish.
There are some very simple objectives, like "Don't die." or "Replicate.", for which this is not exactly true. Unfortunately these are already very problematic objectives.
Already today you can create dumb agents in evolutional experiments that follow these objectives. And you do not have to implement those objectives. They are intrinsic to (complex) dynamical systems. Systems that change over time. It's not even that you need the physical laws of our universe to spawn evolution. Evolution is an effect that simply exists in every dynamical, i.e., time changing, system. Things that survive and replicate tend to dominate things that don't, because they survive and replicate. Simple, right? You do not have to program things to follow these objectives, they are simply there.
This is true. Just two days ago I did a medium sized refactoring on a Scala project.It's a personal project, and I started moving some stuff around, and then I did more and more, renaming things, changing method signatures. And right in the middle I thought "Damn, what's this mess I just made? You should have made this one little step at a time.". But I just went on fixing compile errors. And in the end everything worked again, and I was fine. With strong, static typing, you can just make the change you want to make, and then fix all the type errors–and most of the time it will work again. No way this is possible in a dynamically typed language. But then, I probably wouldn't have done this with Python.
I doubt that the authors of the paper can build a 1eV transistor right now. It can be done in theory. In theory you can also make the irreversible transistor much better.
And now that I skimmed over the article, it says that only 1 meV is theoretically lost per bit. This makes my point even more valid. We can improve current technology to be one million times more efficient before hitting this thermodynamical barrier.
Also, you finish your post with an unsupported statement. It might also be less plausible to shuffle 1000eV wasting 1eV than trying to make 1eV switches directly. And being a bit picky: eV is unit for energy, and not for charge.
I did not read this article, so I'll just make up some numbers I find plausible.
Somehow we need to equate entropy (or information loss) with energy. The assumption of 1eV per bit is probably ok. One electron, either changing potential of 1V or not.
So by making computations reversible, we could avoid this inevitable 1ev loss if the computation is not reversible. Nice. But if we currently burn 1keV per switch, there is no point talking about this technology right now. Let's shave off another 990eV first. Then we can think about reversibility.
You have to understand that making the computation reversible (not loosing this bit of knowledge) gets you from 1000eV to 999eV. That's premature optimization. And you all know that this is the root of all evil. Though we might keep an eye on it.
The goal is to convert the electric energy mostly into kinetic energy, and not into heat. The second law of thermodynamics implies that you cannot recover heat energy completely.
Maglev in evacuated tubes can, in theory, be one of the most energy efficient ways of transportation. There is no loss to friction–so not much to fear from Thermodynamic's second law, making the process reversable in theory. And if you can then use the Maglev technology to recover most of the kinetic energy, you're there.
The reported error rate is for conversational English. This means that you cannot throw meaningless words at it. Modern speech recognition exploits grammatical and semantical structure. The stock recognizers can't do this for programming languages. You could train the model on a programming language, and certain constructs (like brackets, if-then-else) will see an improvement in recognition.
I'm totally not pro computer/smartphone for kids. But I doubt that Waldorf schools are much better than ordinary schools (at least here in Germany). If you draw such a conclusion, you must control for other differences, e.g., having wealthier parents and probably other stuff.
The problem is that there is never hard evidence for such things. From a philosophical point of view, you can actually proof nothing about the real world. There are only cues, and you take them together to form a belief that lets you be certain to some degree. This degree should never reach total certainty, otherwise you made a mistake.
Now come the Putin guys shouting "Evidence! Evidence!" all the time, just to cast into doubt their rather probable involvement. That's exactly like conspiracy theories work.
MH-17 is really, really probably mostly the fault of Russia. There are tons of cues, like social media posts, radio calls, and more technical stuff, all pointing towards that it had been shot down with a Russian Buk, operated either by Russian soldiers (the more probable explanation), or Ukrainian separatists (improbable, and these also appear to be largely Russian soldiers on "vacation"). All the cues give a very consistent picture of a major Russian involvement, but you cannot prove it. Yes, you'll never be able to do this. Still it is reasonable to belief it was Russia, and not to believe Russia telling you that it was someone else (btw, Russia's story was changing on a weekly basis, while the story of the west was always the same, the consistent one).
Modern SAT solvers are able to solve SAT problems with millions of variables in seconds. Yes, there are hard problems with some hundred variables that are too hard to solve. But as it turned out, most useful problems are easy to solve. So if you have an NP-complete problem, you should just try to put it into minisat. If it can't be solved easily, there's still time left to despair.
Darktable is a primarily a great Raw editor. But over time it has become a decent photo manager, too. Darktable supports lossless edits, so you can store your untouched original files, and all derivations are stored by their edit history in sidecar files.
I used to use digikam, which has many good features. But digikam simply crashes way too often.
If you carefully read my comment, you will notice that it does not contain any appraisal. Certainly, there are intricate trade-offs involved in coming up with a good policy.
What I wanted to state is the following: Exercising the existing laws that prohibit killing and bombing people is of no use if you want to prevent terrorism. Terrorists do not care about the punishment when they blow themselves up in the end anyway. Thus you need prevention, and prevention only works when you have information.
Slashdot Mirror
If my completion is predictable, I know what I have to type without looking at the screen.
If it is smart, I must always check if what I want to do is in the completion list.
If it is too smart, it will propose complex completions, that are correct 80% of the time, and look correct 90% of the time at first glance.
The "typos" I make today with smart typing correction in my smart phone look to other people like correct sentences. It's just confusing. No one can guess what I meant to write.
I've seen them, and I hate them. I hated them from the first time I saw them. Actually, so much that I always (successfully) tried to avoid writing one. I think once I used a template.
Just look at this fuckery. I have to write the PID logic myself? Or I have to copy code? This is bullshit. I don't wan't to do this.
Once I've read a thread about X, and one comment said "If you have a UI, it's not Linux anymore!". Init scripts are for those people...
Never having to deal with init scripts before, I had to install a custom service on a server. My company was using custom bash scripts to restart processes that die. They also have some wierd mechanism to suspend the restart behavior, and some really bad email reporting. I wrote a systemd script for my service and was quite happy with it. I could even configure sending an email on crash rather easily.
When talking to the sysadmin of our company, it appears to me that he doesn't like systemd, just because it does some things differently compared to as it was before. That's a pretty weak argument.
Just buy a phone with a replaceable battery. Done! There are many. I have a Moto G4 Play, which has replaceable battery. No need to bother with AAA batteries – that's really a stupid idea.
This study must be fake news. Trump is both extremely intelligent, smartest president EVER! Also he is best president, if not bestest! Fake news!
This sounds much like the "If you know how to program, you can pick up any language quickly." kind of thinking. I encounter it often, and I also thought this way right after finishing university. But it's simply not this way. Maybe all imperative languages are alike, yes. And if you know imperative, and OOP you can pick up many languages quickly. You will not learn anything new doing so, though. But if you change paradigm, you will find out that there is much you don't know.
SEL is a 1998 anime. It is full of power line shots. I'd estimate that around 2-5 percent of the series consist of power lines. It would be interesting whether this was the start of the trend. Can someone please categorize this Tumblr thing into pre and post 1998, please?
I'm about to finish my PhD in a subject closely related to Deep Learning. I don't find an appealing job here in Germany without relocation. So I'll go on doing embedded development.
Even with the best AI, we need to give it an objective to try to accomplish.
There are some very simple objectives, like "Don't die." or "Replicate.", for which this is not exactly true. Unfortunately these are already very problematic objectives.
Already today you can create dumb agents in evolutional experiments that follow these objectives. And you do not have to implement those objectives. They are intrinsic to (complex) dynamical systems. Systems that change over time. It's not even that you need the physical laws of our universe to spawn evolution. Evolution is an effect that simply exists in every dynamical, i.e., time changing, system. Things that survive and replicate tend to dominate things that don't, because they survive and replicate. Simple, right? You do not have to program things to follow these objectives, they are simply there.
I love this quote.
Though it's totally unclear where it originated.
This is true. Just two days ago I did a medium sized refactoring on a Scala project.It's a personal project, and I started moving some stuff around, and then I did more and more, renaming things, changing method signatures. And right in the middle I thought "Damn, what's this mess I just made? You should have made this one little step at a time.". But I just went on fixing compile errors. And in the end everything worked again, and I was fine. With strong, static typing, you can just make the change you want to make, and then fix all the type errors–and most of the time it will work again. No way this is possible in a dynamically typed language. But then, I probably wouldn't have done this with Python.
I doubt that the authors of the paper can build a 1eV transistor right now. It can be done in theory. In theory you can also make the irreversible transistor much better.
And now that I skimmed over the article, it says that only 1 meV is theoretically lost per bit. This makes my point even more valid. We can improve current technology to be one million times more efficient before hitting this thermodynamical barrier.
Also, you finish your post with an unsupported statement. It might also be less plausible to shuffle 1000eV wasting 1eV than trying to make 1eV switches directly. And being a bit picky: eV is unit for energy, and not for charge.
I did not read this article, so I'll just make up some numbers I find plausible.
Somehow we need to equate entropy (or information loss) with energy. The assumption of 1eV per bit is probably ok. One electron, either changing potential of 1V or not.
So by making computations reversible, we could avoid this inevitable 1ev loss if the computation is not reversible. Nice. But if we currently burn 1keV per switch, there is no point talking about this technology right now. Let's shave off another 990eV first. Then we can think about reversibility.
You have to understand that making the computation reversible (not loosing this bit of knowledge) gets you from 1000eV to 999eV. That's premature optimization. And you all know that this is the root of all evil. Though we might keep an eye on it.
The goal is to convert the electric energy mostly into kinetic energy, and not into heat. The second law of thermodynamics implies that you cannot recover heat energy completely.
Maglev in evacuated tubes can, in theory, be one of the most energy efficient ways of transportation. There is no loss to friction–so not much to fear from Thermodynamic's second law, making the process reversable in theory. And if you can then use the Maglev technology to recover most of the kinetic energy, you're there.
The reported error rate is for conversational English. This means that you cannot throw meaningless words at it. Modern speech recognition exploits grammatical and semantical structure. The stock recognizers can't do this for programming languages. You could train the model on a programming language, and certain constructs (like brackets, if-then-else) will see an improvement in recognition.
There are some very nice games that require/teach programming skills. You can check out virtually every game by Zachtronics.
I'm totally not pro computer/smartphone for kids. But I doubt that Waldorf schools are much better than ordinary schools (at least here in Germany). If you draw such a conclusion, you must control for other differences, e.g., having wealthier parents and probably other stuff.
The problem is that there is never hard evidence for such things. From a philosophical point of view, you can actually proof nothing about the real world. There are only cues, and you take them together to form a belief that lets you be certain to some degree. This degree should never reach total certainty, otherwise you made a mistake.
Now come the Putin guys shouting "Evidence! Evidence!" all the time, just to cast into doubt their rather probable involvement. That's exactly like conspiracy theories work.
MH-17 is really, really probably mostly the fault of Russia. There are tons of cues, like social media posts, radio calls, and more technical stuff, all pointing towards that it had been shot down with a Russian Buk, operated either by Russian soldiers (the more probable explanation), or Ukrainian separatists (improbable, and these also appear to be largely Russian soldiers on "vacation"). All the cues give a very consistent picture of a major Russian involvement, but you cannot prove it. Yes, you'll never be able to do this. Still it is reasonable to belief it was Russia, and not to believe Russia telling you that it was someone else (btw, Russia's story was changing on a weekly basis, while the story of the west was always the same, the consistent one).
But, but, but...
72. Reinvent the Wheel Often
*Few* compilers would be more be even more correct.
Modern SAT solvers are able to solve SAT problems with millions of variables in seconds. Yes, there are hard problems with some hundred variables that are too hard to solve. But as it turned out, most useful problems are easy to solve. So if you have an NP-complete problem, you should just try to put it into minisat. If it can't be solved easily, there's still time left to despair.
I patent hydrogen. Now it's your turn
But don't you dare to patent elementary particles! This would ruin the fun of patenting atoms, and I can't have my hydrogen.
And don't complain that you wanted to patent molecules before I had the idea of patenting atoms! That's not how the game's supposed to be played
Darktable is a primarily a great Raw editor. But over time it has become a decent photo manager, too. Darktable supports lossless edits, so you can store your untouched original files, and all derivations are stored by their edit history in sidecar files.
I used to use digikam, which has many good features. But digikam simply crashes way too often.
If you carefully read my comment, you will notice that it does not contain any appraisal. Certainly, there are intricate trade-offs involved in coming up with a good policy.
What I wanted to state is the following: Exercising the existing laws that prohibit killing and bombing people is of no use if you want to prevent terrorism. Terrorists do not care about the punishment when they blow themselves up in the end anyway. Thus you need prevention, and prevention only works when you have information.