I'd say it depends on where they were educated more than where they're from. The Indian education system seems to be designed to actively prohibit thinking, so if it's an Indian programmer they're more likely to be competent the quicker they were removed from the Indian education system. If they got away before university then they may be okay. If they escaped as children then they're a lot more likely to be competent.
I'd also add Russia to the original list. Some of the most competent developers I've worked with recently are Russian.
Yes, it's true (looks like the GNU fanboys have mod points today). Clang can build in Visual Studio, but there are still some rough corners to it. A few of the ReactOS developers and some Sony people are now actively working on LLVM / Clang in Windows, so it's likely to improve a lot in the 3.1 timeframe. Currently the big missing things are support for Microsoft's non-standard template instantiation (which means it can't quite parse all of the Windows C++ headers, only the C ones) and support for exceptions. Both of these are in progress.
C90 does not contain a standard type that has a 64-bit range. C99 defines long long, which must have a range greater than or equal to a 64-bit binary number. It also defines int64_t and uint64_t, which must have exactly the range of a 64-bit binary number. More usefully, it also defines [u]intptr_t, which must have the same size as void*. There is no C90 integer type that is guaranteed to be the same size as a pointer, and the fact that a lot of people assume that sizeof(long) == sizeof(void*) is one of the things most likely to be responsible for code not being portable.
I wouldn't hire anyone who wrote C90 these days. There's simply no excuse for it.
As with every other version of the C standard, you can read the draft yourself. A few things that are nice:
A detailed and well-thought-out set of atomic operations. I've got a diff for clang that implements these that should be committed in the next few days after a bit of tidying. Ed Schouten has written the supporting header to FreeBSD libc, so these can be used now (with fall back to GCC intrinsics for a marginally slower implementation).
Unicode string literals and a few functions for manipulating them.
_Generic() letting you write type-generic macros.
Anonymous structure and union members, so you can write things like struct { int tag; union { void *ptr; uintptr_t i}; } s; and then refer to s.ptr or s.i, rather than needing to provide a name for the union (this is already a GNU extension, but it's nice to have it in the standard).
Static assertions, so you can do things like _Static_assert(sizeof(int) == 4, "This code path should not be used in ILP64 platforms!"); and get a compile-time error if it is.
A _Thread_local storage qualifier for thread-local variables (equivalent to the __thread GNU extension).
Alignment checks and specifiers.
A few things from POSIX, like the x specifier in fopen() for exclusive open.
Some of the not-so-nice features include threads.h, which is equivalent to pthreads but with a different function names (and ones that seem quite likely to cause conflicts with existing code).
GCC? People still use that? Clang can now parse a lot of the standard windows headers. 3.1 should have finished implementing the required quirks to understand the Windows templates. There's also work underway to support the Win64 exception model, which will hopefully be done by the 3.1 release.
In this case it's entirely justified, although those functions have been an optional part of C as TR 24731-1 for a few years. They are entirely pointless (with the possible exception of gets_s) - they only address bugs that result from spectacularly bad code, and the sort of person who writes code that bad won't use the _s variants.
To be fair to the grandparent, glowing peas would be pretty awesome and if someone is not trying to splice firefly and pea DNA to achieve this then I think we should be looking hard at the genetics community and asking 'why not?'
The only smartphone feature I actually use on my phone is the SIP client. Oh, and very occasionally the OSM app. I'd be quite happy with a dumb phone as long as it could make SIP calls over WiFi. Unfortunately, the only phones I've seen like that are shockingly bad at being telephones. I want:
To be able to sync my contacts (maybe calendar, don't really care though) over Bluetooth.
To be able to leave the phone connected to the WiFi for long periods without flattening the battery so it can receive SIP calls
To be able to make SIP and GSM calls via the same UI.
To be able to send and receive SMS (well, actually, I don't, but other people I communicate with want me to).
To not want to stab the UI designer in the face every time I use the phone.
To not pay a silly amount for the phone.
That's it. There may be phones that meet these requirements, but I've not yet found one.
You can use heuristics like that. Or you can try the really novel approach of actually testing. On this machine, -j4 to -j8 are about the same speed (although use more RAM, so with C++ stuff that takes 512MB+ per compiler process it can start swapping, which completely kills performance). Any higher numbers and it gets slower. -j4 is the optimum.
Xerox PARC hasn't existed for a long time. PARC was sold off. It is not affiliated with Xerox and lacks most of what made Xerox PARC cool in the first place.
Oh, and it's a mistake to say, as TFA claims, that PARC was famous for failing to commercialise its inventions. I can think of at four things off the top of my head that made Xerox more money than the total operating costs of PARC for the entire time that Xerox ran it.
Overclocking my old Cyrix CPUs made a noticeable difference in a lot of games. These days? I just don't care. My laptop has a quad-core 2.2GHz i7. It is insanely fast. Even big compile jobs can run with -j4 and it's still responsive. My tablet has a 1.2GHz dual-core processor, and it's faster than the desktop that my mother uses - why would I bother overclocking it?
Add to that, a lot of work on mobile devices is bursty and the operating systems tend to be very aggressive about sleep states. If you make your CPU faster, it takes less time to do a small burst of work (e.g. rendering a web page) and then goes back to sleep. The faster processor - drawing more power when under load - can therefore use less power overall.
The test with HIV+ people is to check for side effects. They won't get any benefit from it, but they will experience all of the side effects that come with it (hopefully none, but the point of testing is to try to turn 'hopefully' into 'definitely'). The next round of testing (with the HIV- people) is to see if it actually works. These people are likely to be selected from demographics deemed to have a high risk of HIV infection, and if none of them become infected then it worked. If some of them, but a lower population than would be expected, become HIV+ then the vaccine is partially successful and may be used anyway if it has no side effects: reducing the probability of infection by 50% goes a long way towards eliminating the disease, because now there's a much higher chance that no one who is not immune will come into contact with a carrier.
Have you actually done a survey on this, or is it just an assumption? In my experience, computer voice falls into the uncanny valley very quickly - people find computers that try to sound like humans to be creepy.
I had exactly that experience with Friends Reunited 10 years ago. If I actually like someone then it's worth spending the token half hour every six months sending them an email and staying in touch. If I don't even care about them enough to do that, then there's no reason for using a social networking system to stay in touch with them. We're not friends, and we don't gain anything from pretending to be. Remembering that experience, I opted out of the current social networking bubble.
It's okay to lose touch with people. You change, they change. You end up with different sets of interests. You will still be close to some of the people who are your friends now in 20 years. You will find you have almost nothing in common with others much sooner. That's life. You make new friends, you lose contact with some of the others. Trying to hang on to every single friend you've ever met just seems a bit desperate.
Sure. Don't use it. Facebook's only value to people comes from the fact that other people use it. If you say 'well, I just use it because other people will only communicate with me using it' then you are adding to the problem. Stop using it. Delete your account (in as much as you can), and use other mechanisms for communicating. Set up mailing lists for your friends to use. Help them set up their own web sites. And if people say that they will only use Facebook, then tell them you're not on it, you have no intention of being on it, and that if they want to communicate with you then they can use something else.
That's a UI problem. Knowing how to use it should mean 'not clicking on the enable/disable encryption checkbox that defaults to enable'. Unfortunately, we don't have any mechanisms for key distribution that can make it this simple.
If you're using Google Mail via IMAP / POP (as I do for a client who insists I have an @theircompany.com to use for stuff where I might be perceived as representing them) then you can use S/MIME or PGP to encrypt the message. Then no one who has access to Google's servers can decrypt the message unless they have the decryption key.
The problem, as always, is key distribution. If I have someone's public PGP key, I can encrypt messages addressed to them. If I don't, then I can't send them an encrypted message. How do I get this key? If enough people use PGP then there is probably a version of their public key online somewhere signed by one or more people I trust. If not...
I can only speak for the ones where I've spent a reasonable amount of time, which basically means New York City, Salt Lake City, and Pittsburgh, but all of those did. Pittsburgh had a long street entirely filled with interesting food shops, but it was a little bit out of the centre so it's easy to miss. The main problem with US cities is that they tend to sprawl and they tend to put all similar shops together, so there can be a dozen of them within a mile of you but you won't notice unless you've gone through that part of the city for another reason. In the UK, small shops like this tend to be scattered through residential and commerical areas.
What about small butchers shops? I'm vegetarian, so I haven't actually tried, but I'm pretty sure that there are four butchers near me that will sell parts of animals from local farms. This also solves the problem of 'WTF do I do with a dead cow?' - someone else has already chopped out the bits that you don't want to eat and cut the bits you do into sensible-size portions. For vegetables, the local market sells things from several local farms.
I use a drive's warranty period as the time after which I expect it to fail. If your drive comes with a 1-year warranty, then I won't expect it to last more than one year. If your competitor's drive comes with a 5-year warranty, I'll expect to have to replace your drive five times before I replace theirs and I'll take that into account when deciding which to buy.
Slashdot Mirror
When one thinks of NYC, the same stirrings of exciting bleeding edge technological progress does not come to mind as say, Palo Alto
Remind me again where Google Research's main US site is?
I'd say it depends on where they were educated more than where they're from. The Indian education system seems to be designed to actively prohibit thinking, so if it's an Indian programmer they're more likely to be competent the quicker they were removed from the Indian education system. If they got away before university then they may be okay. If they escaped as children then they're a lot more likely to be competent.
I'd also add Russia to the original list. Some of the most competent developers I've worked with recently are Russian.
Yes, it's true (looks like the GNU fanboys have mod points today). Clang can build in Visual Studio, but there are still some rough corners to it. A few of the ReactOS developers and some Sony people are now actively working on LLVM / Clang in Windows, so it's likely to improve a lot in the 3.1 timeframe. Currently the big missing things are support for Microsoft's non-standard template instantiation (which means it can't quite parse all of the Windows C++ headers, only the C ones) and support for exceptions. Both of these are in progress.
C90 does not contain a standard type that has a 64-bit range. C99 defines long long, which must have a range greater than or equal to a 64-bit binary number. It also defines int64_t and uint64_t, which must have exactly the range of a 64-bit binary number. More usefully, it also defines [u]intptr_t, which must have the same size as void*. There is no C90 integer type that is guaranteed to be the same size as a pointer, and the fact that a lot of people assume that sizeof(long) == sizeof(void*) is one of the things most likely to be responsible for code not being portable.
I wouldn't hire anyone who wrote C90 these days. There's simply no excuse for it.
Some of the not-so-nice features include threads.h, which is equivalent to pthreads but with a different function names (and ones that seem quite likely to cause conflicts with existing code).
GCC? People still use that? Clang can now parse a lot of the standard windows headers. 3.1 should have finished implementing the required quirks to understand the Windows templates. There's also work underway to support the Win64 exception model, which will hopefully be done by the 3.1 release.
In this case it's entirely justified, although those functions have been an optional part of C as TR 24731-1 for a few years. They are entirely pointless (with the possible exception of gets_s) - they only address bugs that result from spectacularly bad code, and the sort of person who writes code that bad won't use the _s variants.
The last draft and the errata are always free downloads. That's what I've been using to implement the atomics stuff in clang / FreeBSD.
The study goes on to say that failed attempts at a first post can harm your karma.
To be fair to the grandparent, glowing peas would be pretty awesome and if someone is not trying to splice firefly and pea DNA to achieve this then I think we should be looking hard at the genetics community and asking 'why not?'
That's it. There may be phones that meet these requirements, but I've not yet found one.
You can use heuristics like that. Or you can try the really novel approach of actually testing. On this machine, -j4 to -j8 are about the same speed (although use more RAM, so with C++ stuff that takes 512MB+ per compiler process it can start swapping, which completely kills performance). Any higher numbers and it gets slower. -j4 is the optimum.
Does this mean Xerox has a chance?
Xerox PARC hasn't existed for a long time. PARC was sold off. It is not affiliated with Xerox and lacks most of what made Xerox PARC cool in the first place.
Oh, and it's a mistake to say, as TFA claims, that PARC was famous for failing to commercialise its inventions. I can think of at four things off the top of my head that made Xerox more money than the total operating costs of PARC for the entire time that Xerox ran it.
Overclocking my old Cyrix CPUs made a noticeable difference in a lot of games. These days? I just don't care. My laptop has a quad-core 2.2GHz i7. It is insanely fast. Even big compile jobs can run with -j4 and it's still responsive. My tablet has a 1.2GHz dual-core processor, and it's faster than the desktop that my mother uses - why would I bother overclocking it?
Add to that, a lot of work on mobile devices is bursty and the operating systems tend to be very aggressive about sleep states. If you make your CPU faster, it takes less time to do a small burst of work (e.g. rendering a web page) and then goes back to sleep. The faster processor - drawing more power when under load - can therefore use less power overall.
The test with HIV+ people is to check for side effects. They won't get any benefit from it, but they will experience all of the side effects that come with it (hopefully none, but the point of testing is to try to turn 'hopefully' into 'definitely'). The next round of testing (with the HIV- people) is to see if it actually works. These people are likely to be selected from demographics deemed to have a high risk of HIV infection, and if none of them become infected then it worked. If some of them, but a lower population than would be expected, become HIV+ then the vaccine is partially successful and may be used anyway if it has no side effects: reducing the probability of infection by 50% goes a long way towards eliminating the disease, because now there's a much higher chance that no one who is not immune will come into contact with a carrier.
Have you actually done a survey on this, or is it just an assumption? In my experience, computer voice falls into the uncanny valley very quickly - people find computers that try to sound like humans to be creepy.
I had exactly that experience with Friends Reunited 10 years ago. If I actually like someone then it's worth spending the token half hour every six months sending them an email and staying in touch. If I don't even care about them enough to do that, then there's no reason for using a social networking system to stay in touch with them. We're not friends, and we don't gain anything from pretending to be. Remembering that experience, I opted out of the current social networking bubble.
It's okay to lose touch with people. You change, they change. You end up with different sets of interests. You will still be close to some of the people who are your friends now in 20 years. You will find you have almost nothing in common with others much sooner. That's life. You make new friends, you lose contact with some of the others. Trying to hang on to every single friend you've ever met just seems a bit desperate.
Sure. Don't use it. Facebook's only value to people comes from the fact that other people use it. If you say 'well, I just use it because other people will only communicate with me using it' then you are adding to the problem. Stop using it. Delete your account (in as much as you can), and use other mechanisms for communicating. Set up mailing lists for your friends to use. Help them set up their own web sites. And if people say that they will only use Facebook, then tell them you're not on it, you have no intention of being on it, and that if they want to communicate with you then they can use something else.
The down side, presumably, being that those women turned out to have good taste in men?
That's a UI problem. Knowing how to use it should mean 'not clicking on the enable/disable encryption checkbox that defaults to enable'. Unfortunately, we don't have any mechanisms for key distribution that can make it this simple.
If you're using Google Mail via IMAP / POP (as I do for a client who insists I have an @theircompany.com to use for stuff where I might be perceived as representing them) then you can use S/MIME or PGP to encrypt the message. Then no one who has access to Google's servers can decrypt the message unless they have the decryption key.
The problem, as always, is key distribution. If I have someone's public PGP key, I can encrypt messages addressed to them. If I don't, then I can't send them an encrypted message. How do I get this key? If enough people use PGP then there is probably a version of their public key online somewhere signed by one or more people I trust. If not...
Do US cities/towns not have things like that ?
I can only speak for the ones where I've spent a reasonable amount of time, which basically means New York City, Salt Lake City, and Pittsburgh, but all of those did. Pittsburgh had a long street entirely filled with interesting food shops, but it was a little bit out of the centre so it's easy to miss. The main problem with US cities is that they tend to sprawl and they tend to put all similar shops together, so there can be a dozen of them within a mile of you but you won't notice unless you've gone through that part of the city for another reason. In the UK, small shops like this tend to be scattered through residential and commerical areas.
What about small butchers shops? I'm vegetarian, so I haven't actually tried, but I'm pretty sure that there are four butchers near me that will sell parts of animals from local farms. This also solves the problem of 'WTF do I do with a dead cow?' - someone else has already chopped out the bits that you don't want to eat and cut the bits you do into sensible-size portions. For vegetables, the local market sells things from several local farms.
I use a drive's warranty period as the time after which I expect it to fail. If your drive comes with a 1-year warranty, then I won't expect it to last more than one year. If your competitor's drive comes with a 5-year warranty, I'll expect to have to replace your drive five times before I replace theirs and I'll take that into account when deciding which to buy.