I disagree about the ceremony of C++, C++ is a lot lower level and library code is pretty direct. C# on the other hand does like you to create classes all over the place (I'm thinking especially of XAML where you can only bind a gadget to a variable via an object which has properties - the equivalent in C++ would be 2 lines of code, and one of those would be the variable declaration).
I wouldn't like to resolve those namespaces or the horrendously long method names that most style guides expect you to use without an IDE -- think of all the event / delegate hookups you've written and tell me they fit on an 80 characters display:)
Java, of course, is in the middle - and not as good as either for the different areas C# or C++ excels at. a good reason not to use it as the start point.
just great... Java took at the 'best' ideas of C++ and then mangled it into something pretty nasty - a memory and resource hog with poor performance, especially for GUIs, and added non-deterministic finalisation (eventually - the original didn't even have that!) with verbose OO code everywhere. The only good thing it gives is a huge set of libraries.
C# took that and did even more to it, so much so that you require an IDE to write code.
So what does Google do? Take these languages and builds on top of them even more! Great, just great.
Why don't they go back to the source, learn the mistakes that Java made (yes, some were just design decisions, others were due to the limitations of computing back then) and create a language initially based on C++ with the good bits of that included. Modern C++ is pretty good, if Google could update it by removing the 'backward compatible with C' stuff, and some of its poor design decisions (eg auto_ptr for example) then there's a good chance we'd get something pretty damn good.
As it is, I imagine Dart will be just Java in all but name with a javascript compiler. I doubt anyone will be too enthused about that if they can write javascript in the first place.
no, but a majority are what we used to call 'VB devs', only now they have curly brackets, woot!
Try to develop some C# code outside VS and you'll see it is a very verbose language, you have to do a lot of stuff that used to be a single line or two of C++ - I don't mean access the APIs, but it seems to me that everything you do requires a new object being created. Try writing the code in notepad and you won't like it.
Now a lot of people say 'but I never will write C# in notepad so its a moot point' not really, it shows how dependant on the environment you are, how much you rely on the IDE to write code for you. I know some people who can't write C# code without resharper being installed!
So its not the language that gives you productivity, but the IDE. I wish they'd sent all that effort on making the IDE write C++ code for you instead and not bothering with C# at all.
do you have a link to that? I know MS is seeing a renaissance in C++ dev, and that can only be a good thing:) but I thought you could use C#/VB or C++ to write WinRT apps (or HTML/JS based ones).
it is, if you look at the Build slideshows you'll see that the primary interface for metro dev is HTML+JS. However, you can write XAML code using C#/VB or C++ (they made XAML 'native'). I'm not sure how well the interfaces are - the XAML stuff compiles into special Metro interface code so it's not 100% native. At least that's how I read it all.
The applications need to use a different API though - WinRT..NET framework is not supported as some of it accesses outside the sandboxes metro apps run in, so that's now really a legacy system too.
hmm. but that suggests that we have 2 "fully" supported APIs - the new WinRT and the old (legacy).Net Framework. (plus the real legacy ones, like Win32). And I didn't think MS was into that, WinRT was supposed to be the new, API to rule them all.
Experience says that this is stage 1, and WinRT is going to be the one API in the future, just that it isn't complete yet and so you have to continue to develop 'legacy' apps. Its like Silverlight - the message was 'its a first class dev environment' until it wasn't.
One thing is certain, this week the focus at Microsoft is back to C++ development. Maybe next week it'll be something else, but I think it makes a lot of sense to develop C++ back ends with HTML/js front ends, and then wire them together, you might have better luck when the music stops and the next cool thing comes along.
don't forget that we currently have that situation today, its the motherboard-based main DRAM that is the slow memory, the fast 'main' memory is the L1, L2 and sometimes L3 caches on the CPU.
But still, my point is not about having lots of RAM, it's about having everything you own stored in RAM, never being pulled back and forth from the HDD, never having to load files, everything you want is.. just there ready for you.
the biggest problem with modern computers is the bottleneck of RAM - when you want something you have to push it from A to B, in some cases that's from the HDD into RAM, but other times its pushing it from RAM to CPU cache. This is why your ultra-fast computer always appears to be so slow, it's taking ages pushing huge amounts of data back and forth all the time. Putting everything in RAM would stop that happening. We'd still have the CPU cache problem of course, but it would be somewhat mitigated.
And yes, I'm sure HDDs would be cheaper so people would buy them for long-term storage. but I'm dreaming of the future:)
it won't mean the end of the reboot, stupid editor. This is sSlashdot, don't you know you have a Microsoft-Windows-BSOD-Daily-reboot meme to maintain??:)
What I think it could mean the end of is the HDD, or rather the distinction between memory and storage. If all your apps and long-term-storage data could be placed into RAM, then you'd do it wouldn't you. (this assume a few things, like reliability and long-term unpowered persistence) but imagine having 500Gb of RAM that just happened to hold all your data, rather than keeping it separate and shuffling it between the two. That could be quite a change for the way we see computers compared to the ways we've been using them for the last 40 odd years.
oh I agree there. But not with the insane markers - they're too vague. They're like comments you'd apply to a youtube video of someone falling off a skateboard.
"-1 Inaccurate" might be useful, but only if it wasn't abused. "-1 Noise" might also be useful for the rubbish, say-nothing, posts that we get with no benefit to the discussion. "-1 Astroturf" might be a useful one too, for those obviously not-quite-fanboi but obviously promotion posts. Thinking about it, this could fit nicely under the Fanboi mod.
"+1 Well Played, Sir." excellent, truly excellent.
totally agree with that - the previous metamod system was fine. the new one is confusing. Metamod is as important a part of the system as moderating in the first place (if not more).
I might like to see a system where the metamod options are placed inside the article itself, and let people metamod while they read the discussions, so every meta would also be shown in context.
This might be beneficial as the metamod link nearly never appears at the top of the front page, and so I find myself metamodding far, far less often then I used to.
I'm not sure there's much in it from a service provider PoV. iPhone purchasers tend to spend more on apps, so it sounds fair they're more likely to want the device/plan and not care as much about the cost of it.
I know one thing that they'll be thinking - no WP7 or Android phones for sale in their stores, or at least, tucked away in the far corner and the salesmen given instructions (and bonuses) to shift iPhone stocks.
You thought it was going to be hard for Windows Phone to get some market share before, it just got a lot harder.
crashed less than my android phones did (once every month or two),.... . Still, that kind of "windows BSODs" Crap really went out with the 90s unless you buy crap hardware, in which case, no OS will save you.
So you first blame the crashes on the OS, and then say how it's hardware that causes crashes. You probably had a dodgy phone - my Android has crashed twice since I had it, and I think they were battery/hardware issues anyway. So think about your criticisms of something that you already admit are incorrect.
LEO is not the same as Lunar orbit. When the private companies get there, then we can talk. (and I'm sure they will, eventually, cheaper and more efficiently than any government agency could do it)
I've learned a fair few languages - which reinforces my point that we don't need loads more. It seems that every week there's a new language (or framework or development paradigm) that's supposed to be a silver bullet. Guess what, they never are. What they do achieve, is to draw you away from getting better at the 'old boring' stuff while you learn how to do the new - and then you ave to spend longer than you think mastering the new anyway.
There are better ways to achieve what you need, by adding functionality to the existing libraries. What you could do in Ruby, you could do in C/C++ given the library support - even stuff like ActiveRecord (which I like loads) is available in a true 'old fashioned' language like C++ and you'd get to keep all your old code, which is a pragmatic bonus when you get paid to deliver products rather than write code.
Rubbish attitude - this is why software is regarded as a 'hobby' for inexperienced and generally poor developers - you're too busy 'learning something new' all the time and not focussing on getting things done.
The software industry will never become as established and professional until this attitude disappears, we need masters of things, not continual change to something else.
Now, I don't have too much problem with a new language, but the bar for adoption really needs to be set very high to avoid the "change for change's sake" problem I just described. I think js is relatively poor, performance could be a lot better, and there could be a web-client language better geared towards IO, text parsing and GUI development. If this is it, then we'll see - but at the moment, its just hype and vapourware. If you really need something new, go for Google's native client and get that better supported across browsers and platforms.
That's mid-to-long term though. in the short term, nothing portable is powerful enough to replace a real desktop for real computing work. sending an email or reading a pdf is not the kind of work I'm talking about either.
although I agree with you about the dockable tablet/phones coming to take over all our bases, I disagree when you say "sending an email or reading a pdf" is not real computing work. It is, it's what 99% of work usually consists of (if you all writing Word/Excel documents in there too).
Sure, there's always going to be space for niche markets like your 'real' computing work, but that will simply shift to the server where real computing power can be applied. Chances are you'll still be doing that serious work, but on a thin client app which will happily run on your tablet. So if you ever wanted your PC to compile your apps instantly or a heavy-processing game to run, it'll happen, but not quite in the way you imagine.
Come to think of it, it'll be ironic if IBM's original estimate of 6 computers in the world comes true - the 'computers' will be called 'Google's datacentre platform' or Amazon's and we'll all connect to them via mobile clients over the "cloud".
the problem is cost of verification. Currently this is pretty much how it works, you ask the CA for a cert and they issue it to you after verifying your identity in some way.
Trouble is, the 'verification' often consists of little more than an automated email with a link in it, so its no wonder the issued certs end up in the hands of others.
There are a lot of certs issued, so making verification foolproof would ramp the cost of them up to prohibitive prices.
On the other hand - maybe this is exactly what is needed, alongside certs that are issued for encryption only (for the 80% of time when it really doesn't matter who you are) so if only the banks, Microsoft and Google can afford verified certs, then you will have a guarantee that sites with these certs really are them.
seriously, the reason for not airing bad laundry is the interview for the next job - you do not want to turn up to an interview with the guy hiring you reading all about the nasty bitchy things you said about your previous company in the newspaper.
The board makes decisions you don't like, then quit. The CEO is part of the board don't forget, its not some personal fiefdom where you and only you are the guy who has to do everything. Its not much different from being one of the peons whose boss and/or peers and/or upper management make a decision that you disagree with.
Payments for leaving despite making an almighty f***up are just plain wrong.
How does Erlang solve problems where 2 threads want to write to the same memory at the same time without using locks?
the reason Erlang appears easier to do parallel code is that it is designed to work things in parallel - ie completely separate blocks that work without much interoperation. You can get the same with Apple's Grand Central, or Intels Thread Building Blocks, or to a lesser extent OpenMP. These solve the problems you're talking about quite nicely.
I think the difference is that this is a Microsoft application-level technology, not one particularly suited to building an OS. Like your towerblock-from-sugar example, it would fall apart at the fist sign of rain.
Now, we see articles similar to this where something more practical has been achieved, eg Arduino or Rasperry Pi or a thousand XBox/PS3 cluster and that gets more favourable responses, possibly because they are more practical and the geeks can see a reason of sorts for doing it.
I think if something came out of this 'research' then we'd get positive responses. Unfortunately the summary neglected to mention the C# -> native x86 compiler that makes it actually work.
ah, now suddenly the article becomes more interesting - who cares they've written an OS in "C#", what does sound interesting is the whole 'bypass the.NET CLR and everything that makes.NET work and compile C# code to native". A lot of people were interested in that kind of thing a while back, I wonder if they still are.
Slashdot Mirror
I disagree about the ceremony of C++, C++ is a lot lower level and library code is pretty direct. C# on the other hand does like you to create classes all over the place (I'm thinking especially of XAML where you can only bind a gadget to a variable via an object which has properties - the equivalent in C++ would be 2 lines of code, and one of those would be the variable declaration).
I wouldn't like to resolve those namespaces or the horrendously long method names that most style guides expect you to use without an IDE -- think of all the event / delegate hookups you've written and tell me they fit on an 80 characters display :)
Java, of course, is in the middle - and not as good as either for the different areas C# or C++ excels at. a good reason not to use it as the start point.
just great... Java took at the 'best' ideas of C++ and then mangled it into something pretty nasty - a memory and resource hog with poor performance, especially for GUIs, and added non-deterministic finalisation (eventually - the original didn't even have that!) with verbose OO code everywhere. The only good thing it gives is a huge set of libraries.
C# took that and did even more to it, so much so that you require an IDE to write code.
So what does Google do? Take these languages and builds on top of them even more! Great, just great.
Why don't they go back to the source, learn the mistakes that Java made (yes, some were just design decisions, others were due to the limitations of computing back then) and create a language initially based on C++ with the good bits of that included. Modern C++ is pretty good, if Google could update it by removing the 'backward compatible with C' stuff, and some of its poor design decisions (eg auto_ptr for example) then there's a good chance we'd get something pretty damn good.
As it is, I imagine Dart will be just Java in all but name with a javascript compiler. I doubt anyone will be too enthused about that if they can write javascript in the first place.
no, but a majority are what we used to call 'VB devs', only now they have curly brackets, woot!
Try to develop some C# code outside VS and you'll see it is a very verbose language, you have to do a lot of stuff that used to be a single line or two of C++ - I don't mean access the APIs, but it seems to me that everything you do requires a new object being created. Try writing the code in notepad and you won't like it.
Now a lot of people say 'but I never will write C# in notepad so its a moot point' not really, it shows how dependant on the environment you are, how much you rely on the IDE to write code for you. I know some people who can't write C# code without resharper being installed!
So its not the language that gives you productivity, but the IDE. I wish they'd sent all that effort on making the IDE write C++ code for you instead and not bothering with C# at all.
do you have a link to that? I know MS is seeing a renaissance in C++ dev, and that can only be a good thing :) but I thought you could use C#/VB or C++ to write WinRT apps (or HTML/JS based ones).
it is, if you look at the Build slideshows you'll see that the primary interface for metro dev is HTML+JS. However, you can write XAML code using C#/VB or C++ (they made XAML 'native'). I'm not sure how well the interfaces are - the XAML stuff compiles into special Metro interface code so it's not 100% native. At least that's how I read it all.
The applications need to use a different API though - WinRT. .NET framework is not supported as some of it accesses outside the sandboxes metro apps run in, so that's now really a legacy system too.
hmm. but that suggests that we have 2 "fully" supported APIs - the new WinRT and the old (legacy) .Net Framework. (plus the real legacy ones, like Win32). And I didn't think MS was into that, WinRT was supposed to be the new, API to rule them all.
Experience says that this is stage 1, and WinRT is going to be the one API in the future, just that it isn't complete yet and so you have to continue to develop 'legacy' apps. Its like Silverlight - the message was 'its a first class dev environment' until it wasn't.
One thing is certain, this week the focus at Microsoft is back to C++ development. Maybe next week it'll be something else, but I think it makes a lot of sense to develop C++ back ends with HTML/js front ends, and then wire them together, you might have better luck when the music stops and the next cool thing comes along.
don't forget that we currently have that situation today, its the motherboard-based main DRAM that is the slow memory, the fast 'main' memory is the L1, L2 and sometimes L3 caches on the CPU.
But still, my point is not about having lots of RAM, it's about having everything you own stored in RAM, never being pulled back and forth from the HDD, never having to load files, everything you want is .. just there ready for you.
the biggest problem with modern computers is the bottleneck of RAM - when you want something you have to push it from A to B, in some cases that's from the HDD into RAM, but other times its pushing it from RAM to CPU cache. This is why your ultra-fast computer always appears to be so slow, it's taking ages pushing huge amounts of data back and forth all the time. Putting everything in RAM would stop that happening. We'd still have the CPU cache problem of course, but it would be somewhat mitigated.
And yes, I'm sure HDDs would be cheaper so people would buy them for long-term storage. but I'm dreaming of the future :)
it won't mean the end of the reboot, stupid editor. This is sSlashdot, don't you know you have a Microsoft-Windows-BSOD-Daily-reboot meme to maintain?? :)
What I think it could mean the end of is the HDD, or rather the distinction between memory and storage. If all your apps and long-term-storage data could be placed into RAM, then you'd do it wouldn't you. (this assume a few things, like reliability and long-term unpowered persistence) but imagine having 500Gb of RAM that just happened to hold all your data, rather than keeping it separate and shuffling it between the two. That could be quite a change for the way we see computers compared to the ways we've been using them for the last 40 odd years.
oh I agree there. But not with the insane markers - they're too vague. They're like comments you'd apply to a youtube video of someone falling off a skateboard.
"-1 Inaccurate" might be useful, but only if it wasn't abused.
"-1 Noise" might also be useful for the rubbish, say-nothing, posts that we get with no benefit to the discussion.
"-1 Astroturf" might be a useful one too, for those obviously not-quite-fanboi but obviously promotion posts. Thinking about it, this could fit nicely under the Fanboi mod.
"+1 Well Played, Sir." excellent, truly excellent.
totally agree with that - the previous metamod system was fine. the new one is confusing. Metamod is as important a part of the system as moderating in the first place (if not more).
I might like to see a system where the metamod options are placed inside the article itself, and let people metamod while they read the discussions, so every meta would also be shown in context.
This might be beneficial as the metamod link nearly never appears at the top of the front page, and so I find myself metamodding far, far less often then I used to.
I'm not sure there's much in it from a service provider PoV. iPhone purchasers tend to spend more on apps, so it sounds fair they're more likely to want the device/plan and not care as much about the cost of it.
I know one thing that they'll be thinking - no WP7 or Android phones for sale in their stores, or at least, tucked away in the far corner and the salesmen given instructions (and bonuses) to shift iPhone stocks.
You thought it was going to be hard for Windows Phone to get some market share before, it just got a lot harder.
crashed less than my android phones did (once every month or two), .... . Still, that kind of "windows BSODs" Crap really went out with the 90s unless you buy crap hardware, in which case, no OS will save you.
So you first blame the crashes on the OS, and then say how it's hardware that causes crashes. You probably had a dodgy phone - my Android has crashed twice since I had it, and I think they were battery/hardware issues anyway. So think about your criticisms of something that you already admit are incorrect.
LEO is not the same as Lunar orbit. When the private companies get there, then we can talk. (and I'm sure they will, eventually, cheaper and more efficiently than any government agency could do it)
.NET would be two major examples of functionality which will never be emulated perfectly
you mean they can't add Mono to it, or that Mono doesn't implement 100% functionality or .NET (I'm sure the Mono guys said it did) :)
I've learned a fair few languages - which reinforces my point that we don't need loads more. It seems that every week there's a new language (or framework or development paradigm) that's supposed to be a silver bullet. Guess what, they never are. What they do achieve, is to draw you away from getting better at the 'old boring' stuff while you learn how to do the new - and then you ave to spend longer than you think mastering the new anyway.
There are better ways to achieve what you need, by adding functionality to the existing libraries. What you could do in Ruby, you could do in C/C++ given the library support - even stuff like ActiveRecord (which I like loads) is available in a true 'old fashioned' language like C++ and you'd get to keep all your old code, which is a pragmatic bonus when you get paid to deliver products rather than write code.
Rubbish attitude - this is why software is regarded as a 'hobby' for inexperienced and generally poor developers - you're too busy 'learning something new' all the time and not focussing on getting things done.
The software industry will never become as established and professional until this attitude disappears, we need masters of things, not continual change to something else.
Now, I don't have too much problem with a new language, but the bar for adoption really needs to be set very high to avoid the "change for change's sake" problem I just described. I think js is relatively poor, performance could be a lot better, and there could be a web-client language better geared towards IO, text parsing and GUI development. If this is it, then we'll see - but at the moment, its just hype and vapourware. If you really need something new, go for Google's native client and get that better supported across browsers and platforms.
That's mid-to-long term though. in the short term, nothing portable is powerful enough to replace a real desktop for real computing work. sending an email or reading a pdf is not the kind of work I'm talking about either.
although I agree with you about the dockable tablet/phones coming to take over all our bases, I disagree when you say "sending an email or reading a pdf" is not real computing work. It is, it's what 99% of work usually consists of (if you all writing Word/Excel documents in there too).
Sure, there's always going to be space for niche markets like your 'real' computing work, but that will simply shift to the server where real computing power can be applied. Chances are you'll still be doing that serious work, but on a thin client app which will happily run on your tablet. So if you ever wanted your PC to compile your apps instantly or a heavy-processing game to run, it'll happen, but not quite in the way you imagine.
Come to think of it, it'll be ironic if IBM's original estimate of 6 computers in the world comes true - the 'computers' will be called 'Google's datacentre platform' or Amazon's and we'll all connect to them via mobile clients over the "cloud".
the problem is cost of verification. Currently this is pretty much how it works, you ask the CA for a cert and they issue it to you after verifying your identity in some way.
Trouble is, the 'verification' often consists of little more than an automated email with a link in it, so its no wonder the issued certs end up in the hands of others.
There are a lot of certs issued, so making verification foolproof would ramp the cost of them up to prohibitive prices.
On the other hand - maybe this is exactly what is needed, alongside certs that are issued for encryption only (for the 80% of time when it really doesn't matter who you are) so if only the banks, Microsoft and Google can afford verified certs, then you will have a guarantee that sites with these certs really are them.
ah yeah, of course - I was confusing them as most CEOs tend to be the corporate psychos who are chairman too.
I wonder if banning that practice would make corporate business more healthy?
They definitely should have done so. It made the top ten tech industry executive disasters.
The offer was for more like $47bn, and was not in the interests of Jerry Yang, the f***wit.
seriously, the reason for not airing bad laundry is the interview for the next job - you do not want to turn up to an interview with the guy hiring you reading all about the nasty bitchy things you said about your previous company in the newspaper.
The board makes decisions you don't like, then quit. The CEO is part of the board don't forget, its not some personal fiefdom where you and only you are the guy who has to do everything. Its not much different from being one of the peons whose boss and/or peers and/or upper management make a decision that you disagree with.
Payments for leaving despite making an almighty f***up are just plain wrong.
How does Erlang solve problems where 2 threads want to write to the same memory at the same time without using locks?
the reason Erlang appears easier to do parallel code is that it is designed to work things in parallel - ie completely separate blocks that work without much interoperation. You can get the same with Apple's Grand Central, or Intels Thread Building Blocks, or to a lesser extent OpenMP. These solve the problems you're talking about quite nicely.
I think the difference is that this is a Microsoft application-level technology, not one particularly suited to building an OS. Like your towerblock-from-sugar example, it would fall apart at the fist sign of rain.
Now, we see articles similar to this where something more practical has been achieved, eg Arduino or Rasperry Pi or a thousand XBox/PS3 cluster and that gets more favourable responses, possibly because they are more practical and the geeks can see a reason of sorts for doing it.
I think if something came out of this 'research' then we'd get positive responses. Unfortunately the summary neglected to mention the C# -> native x86 compiler that makes it actually work.
ah, now suddenly the article becomes more interesting - who cares they've written an OS in "C#", what does sound interesting is the whole 'bypass the .NET CLR and everything that makes .NET work and compile C# code to native". A lot of people were interested in that kind of thing a while back, I wonder if they still are.