No, there's no reason to assume it'll be slow, unless they just screw up the implementation.
If you're really worried about process creation overhead, then just create a pool of processes and reuse them.
More to the point, do have any idea how absurdly fast today's processors are compared to things like process creation. Exactly how long do you imagine it to take to create one? Long enough for you to notice?!!
No - it's almost certainly NOT a Linux distribution, because a Linux distribution is not just any old pile of crap that includes the Linux kernel.
It's meaningless to call something a Linux distribution unless it at least provides enough of the userspace stuff (glibc, pthreads, X, etc, etc) that you have a hope in hell of running a Linux app on it.
There's zero indication, certainly not from Google, that the Chrome OS will support Linux apps, so it's meaningless to call it Linux.
Given the copious hints Google provides about security, both x86 and ARM support, having some parts in common with Android - this all in addition to it's main feature of simply being able to boot fast and run a browser - the most reasonable guess is that the only form of native app support it'll provide, if any, will be some form of managed code.
So you think that Mono are more interested in interopability with EMCA apps than.NET ones?
Nothing will force Mono to implement.NET extensions that go beyond the EMCA standard... other than total irrelevance if they don't implement the moving target of what the bulk of.NET apps are using.
Nowadays you can certainly be a thin-client to web-based apps, and use web-based storage.
But that doesn't mean you don't need an OS.
An OS provides a high level abstraction and interface to the hardware. Things like:
- Processes/Threads - Memory management - Hardware access (device drivers: network, keyboard, mouse, screen...) / Device drivers - File system (minimally for the OS/drivers itself, even if you were mostly using online storage)
None of that is going away, even if all you want to do is be able to launch and run an app. like Chrome (the browser).
In this case the "OS" (hardware abstraction) may also also include a Java VM since they are targeting multiple CPU architectures (x86, ARM).
There's no mention of it being based on X (I'd assume it's not), Qt, or GTK, so there's no reason to assume that it'll support existing GUI based Linux apps, or Linux apps at all for that matter.
They may support some form of "native" apps as opposed to web-based ones using standard web-based standards that will also work in other browsers, but I'll bet that it's only in some highly restricted constrained environment given the new windowing system and security claims. Maybe you can write native apps using Java and some new Google GUI/etc APIs (or maybe they just ported whatever Java uses for GUI nowadays to sit on top of their own API). Given that they are targetting ARM as well as x86 I doubt there's any native code support.
What's the betting there'll be an app store for it too.
I don't think this is Google-branded Linux for the masses - it's sounds more like Android. Probably a managed code environment using a brand new (or Android?) windowing system, that happens to run on the Linux kernel. Heck, it may be JUST for web/browser-based apps, although I doubt it.
I wouldn't assume that middle management is going to offer better security, especially in an economy that it not doing well.
As a senior developer, assuming you are good, you will always be in demand, but as a junior manager you have little to offer. If you get laid off as an inexperienced manager, especially in the type of economy we're in (and look to be in for a while), then watch out.
If your contingency plan in the event of being laid off as a manager is to go back to being a developer, then why should anyone hire you over someone who hasn't flip-flopped and failed? It certainly won't help that you've demonstrated you don't really want to be a developer!
The data centers and fiber are only there as infrastructure for the web apps (Google, GMail, YouTube).
YouTube is losing money like crazy. GMail on the web makes none (although they could add advertising).
Android is free and open source, not a money maker. I'd guess the strategy is to keep the smart phone field open rather that let Microsoft, Nokia or anyone else control it, maybe as well as to encourage online usage for the same reason Intel spends money encouraging CPU usage.
This isn't meant to be helping Microsoft's competitors, such as Linux. Quite the opposite - it's meant to be helping Microsoft.
Proprietary lock-in the key to Microsoft's business model, whether we're talking about Office document formats, proprietary Internet Explorer extensions and incompatabilities or C#/CLI.
In this case it seems Microsoft accrues multiple benefits from open source developers who can't resist the free candy being offered:
1) It helps spread adoption of Microsoft's proprietary standards
2) It stops open source developers from pushing and developing alternative open source standards
3) It sets anyone (Mono) reimplementing these standards up for the future rug-pull when they change and/or extend the standard in the future, which will be done according to the needs of Microsofts business model
There's a reason the document format for Microsoft's office applications often changes from release to release, and it's not simply because new features are added. This is to force people to upgrade - which is the basis of Microsoft's business model. You can be sure that C#/CLI will be managed in the same way - don't expect Microsoft to keep these standards static now that they have "kindly" encouraged you to adopt/reimplement them. Quite the contrary, once there is significant open-source reliant upon them, then they have more incentive than ever to churn them.
This has got nothing to do with basing a company off free software, nor is it even limited to technology companies.
No company will stay in business forever - eventually they all seem to get made obsolete by some newer company that did a better job of predicting where the market was heading than they did.
It's possible that technology companies (in the broadest sense of technology) may be more liable to be made obsolete by "black swans" that seemingly come out of nowhere, due to the fast moving and semi-unpredictable nature of technology advances and fads, but again this is nothing to do with basing your business of free software.
Any company that fails to continually innovate and look over their shoulder, and instead just kicks back and milks the cash cow, is going to be made obsolete by a competitor with a better product, paradigm, or change in the market.
The best chance of surviving is business is if you can continually manage to make your existing products obsolete, before your competitors do.
The major danger to Google is that they are basically (at least in terms of generating revenue) a one-product company: web search based advertizing. It's not entirely obvious what their strategy is with GMail, Google Maps, etc, but I suspect that a major part of it is to help prop up their core business - to build the Google brand and customer (or rather fodder - the customers are the advertizers) retention.
Google should be worried though (and I'll bet they are) since there is so much room left to improve web search, and someone like Microsoft with the (certainly not free!) infrastructure in place to roll it out could render Google search obsolete overnight with the right software update. It's also possible that a lot of the adveritising market might switch to handhelds (presumably why Google is developing Android), or maybe to social networking sites like Facebook and Twitter where the kids nowadays all hang out.
Nothing with an e-ink screen is going to browse the internet "properly", but nothing without really fits the bill either.
Maybe not yet, but check out this dual-mode reflective/transmissive LCD screen from Pixel Qi.
It has the best of both worlds - in reflective mode it's like an e-ink display, readable in full sunlight, and in transmisssive mode it's a fast color display suitable for watching movies.
I always thought that testing and debugging were the lengthy manual steps
Not if you wrote the code well!;-)
Seriously, as someone who's been doing this a long time (since '78, professionally since '82), and who is still at the top of his game, I nowadays spend *very* little time on debugging since it works first time - even the complicated multi-threaded, mutex type of stuff which is what I primarily write nowadays. After a while you stop making mistakes!
But, anyways, it seems the main target for this adaptive optimization (from TFA) is embedded usage for novel targets. i.e the latest smart phone or maybe games console, where you don't have the typical "essentially infinite for the needs of the application" CPU resources of a desktop app, and you don't have the luxury of a fairly static target (PC architecture) that has had the benefit of years of code generator hand optimization. For this target usage, someone (either you or the compiler) may need to perform extensive low-level optimizations (in addition to the high level design and choice of algorithms), and it therefore helps if an off-the-shelf compiler is available that can do this, and does not rely on the code generator having to be hand-optimized for the hardware architecture you're targeting.
On most modern computers, shifts are slow. They are often even microcoded as multiplications.
You're right to say that recoding a multiplication as a combination of adds and shifts is likely to be a loss since multiplication is so fast, and since the extra instruction fetches (memory accesses, decode overhead) are going to kill it!
However, you're wrong about shifts being slow. Ever since the early days shifts are implemented by a "barrel shifter" can can shift by an arbitrary N bits in a single clock cycle. I very much doubt (although I don't actually know) that modern CPUs with their massive transistor counts have decided to save a few transistors by giving up this efficient implementation.
Will it find places when I am calculating something in a tight loop and move the code somewhere higher?
Quite likely, yes.
Even a dumb optimizer will love loop invariant code outside of a loop, and maybe partially unroll the loop to make the looping overhead less. The latest gcc will even automatically vectorize the loop for you to execute a number of iterations in parallel using SSE/etc instructions if it's a suitable candidate.
e.g.
You may write:
for (int i = 0; i < 10; i++) {
a[i] = (2 * b) + c; }
But the compiler may generate:
t = (2 * b) + c;// Loop invariant code moved outside of loop for (int i = 0; i < 10; i += 2) {
a[i] = t1;// Loop unrolling to halve number of iterations / loop overhead
a[i+1]= t1; }
That's just to illustrate the type of transformations that compilers already do - no exactly what they might generate, which is hopefully rather better!
While I respect him for wanting to try, it's rather clueless to concentrate on the tools needed for what is an artistic endeavor!
If you want to learn to design games then go ahead and start designing them... it doesn't matter what language you design them in. Game play and human factors isn't related to choice of tools. C/C++ and SDL would be as good a choice as any to give it a go on a PC.
OTOH if you want to be part of a commercial games company, then as others have noted it's evolved to become a highly structured business (with budgets comparable to movie production), and the game designer and coders are two different roles. I don't think he really wants to be just a schmuck coder!
Incidently, I remember c.1980 the company I worked for (Acorn, UK) had a games division that would sometimes publish games submitted by independent programmers. Just to illustrate the disconnect between design and tools, we once or twice had people come in with great playable games that had been developed in machine code - no, I don't mean assembler - people had large 16K games they had entirely coded directly in hex based on memorizing the instruction set! I recall one such individual being asked about a possible change to a game and on the spot jumping in an patching it to make work!
Yep - that's even better if you can get some exercise without even trying.
I think that's true of Europe in general (I'm originally from the UK) vs the US (where I now live)... People do still walk into and about town. The US is for the most part totally car-centric. Many residential neighborhoods don't even have sidewalks (= pavements).
However at this point, 3D VMs seem to be an experimental playtoy, not something that can be used for serious gaming.
It makes no sense to lump OpenGL and Direct3D together as "3D" when you're talking about VirtualBix, since they are implemented in very different ways.
VirtualBox OpenGL is basically just as pass-thru to the host driver. The guest box additions includes a virtual OpenGL driver that just passes the commands thru to the the host and the real driver. There must be some performance hit, but the approach seems simple enough.
VirtualBox Direct3D is implemented using the WINE driver that converts Direct3D calls into OpenGL which then get tunneled through to the host OpenGL driver as in the OpenGL case. VirtuaBox Direct3D should therefore be similar in functionality to that in WINE. One upside to the approach is that you don't need a Windows host to have D3D guest aceleration.
That guys usage does seem a bit strange, but each to their own!
However, there are different use cases for tabs that result in different numbers of tabs open:
e.g.
1) User wants to keep a number of sites open all day for quick switching - e.g. mail, news, documentation, etc. This is probably what you are thinking of, and it most likely is a relatively small number of sites - less than a dozen, say.
2) User wants to open a whole bunch of links off the sam page at once, since that's more convenient than flipping back and fro to open one read it, open the next, etc. Examples of this usage would be, for example opening search results (search engine, eBay, etc) or reading new posts on an online forum.
I routinely (many times a day - searching for collectables - made a $1500 profit on one yesterday) open 40-50 tabs at a time for eBay search results since it's so much faster to quickly run down a list of links doing open-in-new-tab (in background) then reading/discarding them, as opposed to doing it one at a time.
Without an IBM, Red Hat, or Mozilla bringing cash and discipline to an open-source project, including paying people to do the 'dirt work' that no one would otherwise do, can open source hope to thrive?"
Maybe not to the degree it has... Linux has certainly greatly benefited from the commercial distros and supporters (e.g. IBM) that have committed cash to adding polish and filling gaps. Unfortunately of course there's only so much they can do - they can commit manpower and technology to individual projects of strategic interest to themselves, but the overall degree of polish of a distro is necessarily heavily reliant on the individual pieces. Each appication/subsystem may be a gem on it's own, but without sufficient standards to guide them (or even desire to adhere to someone else's standards when it's your hobby project done on your time), the resulting pile of gems may be an incoherent and inconsistent mess.
But OTOH, open source thriving and the success of Linux on the desktop (which is the area where the inconsistency and lack of polish hits) are two different things. Linux is already thriving in other areas such as servers and embedded use, and many open source projects such as GNU are used on many platforms other than Linux.
Linux on the desktop is already, and has been for a while, plently good enough (much better than Windows, as is the nature of Unix) for developers who want to use it as a development environment, and seeing as these are the people who created it, that's good enough. If Linux on the desktop never becomes polished enough for a commercial distro to make big bucks off the back of open source developers, then why should we care?
But my guess is that this douche-bag NBIS company could hire enough lawyers to make it not worth anyones money to find out.
I doubt it, because it sounds like the owner of the installed system, the San Francisco Municipal Transit Agency, believes it owns the rights to the data it's generating (train arrival estimates).
From the article:
Muni spokesperson Judson True says otherwise. In fact, he says that, no, Muni owns the data in question and that the public is, of course, entitled to access it. In fact, he went even further: Muni isn't just giving us all permission to access the data, they're also committed to finding ways to make it easier to get to it.
Of course it could be that there's some fine print in the original procurement contract between the SFMTA and NextBus that the Muni spokesman isn't aware of, but given their commitment to making the data available for use it does at least seem as if that's a legal fight that'll be paid for by the SFMTA (i.e. the taxpayers who bought the system) rather than by any individual company wanting to use the data.
It would have to be where a network connection of any usable speed is impossible.
Having a network connection to a traditional room-filling air-conditioned supercomputer isn't an alternative unless they only want a couple of these, which I doubt is the case (on a side note I'd guess the input bandwidth they need isn't that great - probably just real-time video/voice).
I think it's more a matter of wanting to have this sort of computing power available in the field on a potentially widespread (not one-off) scale. They probably have in mind things like AI and advanced image/pattern recognition for robo-warrior type applications.
The SiCortex website mentions their top of the line machine using about 900mW per 1.4GFlop processor (1.55Glop/W), but the overall system uses 20KW for a 5832 processor system, giving a system figure of (5832 * 1.4) / 20,000 = 0.40GFlop/W.
Slashdot Mirror
No, there's no reason to assume it'll be slow, unless they just screw up the implementation.
If you're really worried about process creation overhead, then just create a pool of processes and reuse them.
More to the point, do have any idea how absurdly fast today's processors are compared to things like process creation. Exactly how long do you imagine it to take to create one? Long enough for you to notice?!!
No - it's almost certainly NOT a Linux distribution, because a Linux distribution is not just any old pile of crap that includes the Linux kernel.
It's meaningless to call something a Linux distribution unless it at least provides enough of the userspace stuff (glibc, pthreads, X, etc, etc) that you have a hope in hell of running a Linux app on it.
There's zero indication, certainly not from Google, that the Chrome OS will support Linux apps, so it's meaningless to call it Linux.
Given the copious hints Google provides about security, both x86 and ARM support, having some parts in common with Android - this all in addition to it's main feature of simply being able to boot fast and run a browser - the most reasonable guess is that the only form of native app support it'll provide, if any, will be some form of managed code.
So you think that Mono are more interested in interopability with EMCA apps than .NET ones?
Nothing will force Mono to implement .NET extensions that go beyond the EMCA standard... other than total irrelevance if they don't implement the moving target of what the bulk of .NET apps are using.
Nowadays you can certainly be a thin-client to web-based apps, and use web-based storage.
But that doesn't mean you don't need an OS.
An OS provides a high level abstraction and interface to the hardware. Things like:
- Processes/Threads
- Memory management
- Hardware access (device drivers: network, keyboard, mouse, screen...) / Device drivers
- File system (minimally for the OS/drivers itself, even if you were mostly using online storage)
None of that is going away, even if all you want to do is be able to launch and run an app. like Chrome (the browser).
In this case the "OS" (hardware abstraction) may also also include a Java VM since they are targeting multiple CPU architectures (x86, ARM).
This appears to have little to do with Linux other than the fact it uses the Linux kernel & drivers.
From the Google announcement it uses a "new windowing system" (sounds rather like Android, although they it's something different).
http://googleblog.blogspot.com/2009/07/introducing-google-chrome-os.html
There's no mention of it being based on X (I'd assume it's not), Qt, or GTK, so there's no reason to assume that it'll support existing GUI based Linux apps, or Linux apps at all for that matter.
They may support some form of "native" apps as opposed to web-based ones using standard web-based standards that will also work in other browsers, but I'll bet that it's only in some highly restricted constrained environment given the new windowing system and security claims. Maybe you can write native apps using Java and some new Google GUI/etc APIs (or maybe they just ported whatever Java uses for GUI nowadays to sit on top of their own API). Given that they are targetting ARM as well as x86 I doubt there's any native code support.
What's the betting there'll be an app store for it too.
I don't think this is Google-branded Linux for the masses - it's sounds more like Android. Probably a managed code environment using a brand new (or Android?) windowing system, that happens to run on the Linux kernel. Heck, it may be JUST for web/browser-based apps, although I doubt it.
Yeah, and that has nothing to do with who owns the intellectual property rights. Same for h.264.
Dumbass.
I wouldn't assume that middle management is going to offer better security, especially in an economy that it not doing well.
As a senior developer, assuming you are good, you will always be in demand, but as a junior manager you have little to offer. If you get laid off as an inexperienced manager, especially in the type of economy we're in (and look to be in for a while), then watch out.
If your contingency plan in the event of being laid off as a manager is to go back to being a developer, then why should anyone hire you over someone who hasn't flip-flopped and failed? It certainly won't help that you've demonstrated you don't really want to be a developer!
Yeah, but none of these are making any money.
The data centers and fiber are only there as infrastructure for the web apps (Google, GMail, YouTube).
YouTube is losing money like crazy. GMail on the web makes none (although they could add advertising).
Android is free and open source, not a money maker. I'd guess the strategy is to keep the smart phone field open rather that let Microsoft, Nokia or anyone else control it, maybe as well as to encourage online usage for the same reason Intel spends money encouraging CPU usage.
This isn't meant to be helping Microsoft's competitors, such as Linux. Quite the opposite - it's meant to be helping Microsoft.
Proprietary lock-in the key to Microsoft's business model, whether we're talking about Office document formats, proprietary Internet Explorer extensions and incompatabilities or C#/CLI.
In this case it seems Microsoft accrues multiple benefits from open source developers who can't resist the free candy being offered:
1) It helps spread adoption of Microsoft's proprietary standards
2) It stops open source developers from pushing and developing alternative open source standards
3) It sets anyone (Mono) reimplementing these standards up for the future rug-pull when they change and/or extend the standard in the future, which will be done according to the needs of Microsofts business model
There's a reason the document format for Microsoft's office applications often changes from release to release, and it's not simply because new features are added. This is to force people to upgrade - which is the basis of Microsoft's business model. You can be sure that C#/CLI will be managed in the same way - don't expect Microsoft to keep these standards static now that they have "kindly" encouraged you to adopt/reimplement them. Quite the contrary, once there is significant open-source reliant upon them, then they have more incentive than ever to churn them.
This has got nothing to do with basing a company off free software, nor is it even limited to technology companies.
No company will stay in business forever - eventually they all seem to get made obsolete by some newer company that did a better job of predicting where the market was heading than they did.
It's possible that technology companies (in the broadest sense of technology) may be more liable to be made obsolete by "black swans" that seemingly come out of nowhere, due to the fast moving and semi-unpredictable nature of technology advances and fads, but again this is nothing to do with basing your business of free software.
Any company that fails to continually innovate and look over their shoulder, and instead just kicks back and milks the cash cow, is going to be made obsolete by a competitor with a better product, paradigm, or change in the market.
The best chance of surviving is business is if you can continually manage to make your existing products obsolete, before your competitors do.
The major danger to Google is that they are basically (at least in terms of generating revenue) a one-product company: web search based advertizing. It's not entirely obvious what their strategy is with GMail, Google Maps, etc, but I suspect that a major part of it is to help prop up their core business - to build the Google brand and customer (or rather fodder - the customers are the advertizers) retention.
Google should be worried though (and I'll bet they are) since there is so much room left to improve web search, and someone like Microsoft with the (certainly not free!) infrastructure in place to roll it out could render Google search obsolete overnight with the right software update. It's also possible that a lot of the adveritising market might switch to handhelds (presumably why Google is developing Android), or maybe to social networking sites like Facebook and Twitter where the kids nowadays all hang out.
Nothing with an e-ink screen is going to browse the internet "properly", but nothing without really fits the bill either.
Maybe not yet, but check out this dual-mode reflective/transmissive LCD screen from Pixel Qi.
It has the best of both worlds - in reflective mode it's like an e-ink display, readable in full sunlight, and in transmisssive mode it's a fast color display suitable for watching movies.
http://www.youtube.com/watch?v=Mm8WoItVRn0
http://www.youtube.com/watch?v=oawX3wenxNc
There's an article here discussing a new electric car battery that Chrysler will be using that can take an 80% charge in 5 minutes!
http://www.gizmag.com/chrysler-a123-batteries-electric-vehicles/11497/
I always thought that testing and debugging were the lengthy manual steps
Not if you wrote the code well! ;-)
Seriously, as someone who's been doing this a long time (since '78, professionally since '82), and who is still at the top of his game, I nowadays spend *very* little time on debugging since it works first time - even the complicated multi-threaded, mutex type of stuff which is what I primarily write nowadays. After a while you stop making mistakes!
But, anyways, it seems the main target for this adaptive optimization (from TFA) is embedded usage for novel targets. i.e the latest smart phone or maybe games console, where you don't have the typical "essentially infinite for the needs of the application" CPU resources of a desktop app, and you don't have the luxury of a fairly static target (PC architecture) that has had the benefit of years of code generator hand optimization. For this target usage, someone (either you or the compiler) may need to perform extensive low-level optimizations (in addition to the high level design and choice of algorithms), and it therefore helps if an off-the-shelf compiler is available that can do this, and does not rely on the code generator having to be hand-optimized for the hardware architecture you're targeting.
On most modern computers, shifts are slow. They are often even microcoded as multiplications.
You're right to say that recoding a multiplication as a combination of adds and shifts is likely to be a loss since multiplication is so fast, and since the extra instruction fetches (memory accesses, decode overhead) are going to kill it!
However, you're wrong about shifts being slow. Ever since the early days shifts are implemented by a "barrel shifter" can can shift by an arbitrary N bits in a single clock cycle. I very much doubt (although I don't actually know) that modern CPUs with their massive transistor counts have decided to save a few transistors by giving up this efficient implementation.
Will it find places when I am calculating something in a tight loop and move the code somewhere higher?
Quite likely, yes.
Even a dumb optimizer will love loop invariant code outside of a loop, and maybe partially unroll the loop to make the looping overhead less. The latest gcc will even automatically vectorize the loop for you to execute a number of iterations in parallel using SSE/etc instructions if it's a suitable candidate.
e.g.
You may write:
That's just to illustrate the type of transformations that compilers already do - no exactly what they might generate, which is hopefully rather better!
While I respect him for wanting to try, it's rather clueless to concentrate on the tools needed for what is an artistic endeavor!
If you want to learn to design games then go ahead and start designing them... it doesn't matter what language you design them in. Game play and human factors isn't related to choice of tools. C/C++ and SDL would be as good a choice as any to give it a go on a PC.
OTOH if you want to be part of a commercial games company, then as others have noted it's evolved to become a highly structured business (with budgets comparable to movie production), and the game designer and coders are two different roles. I don't think he really wants to be just a schmuck coder!
Incidently, I remember c.1980 the company I worked for (Acorn, UK) had a games division that would sometimes publish games submitted by independent programmers. Just to illustrate the disconnect between design and tools, we once or twice had people come in with great playable games that had been developed in machine code - no, I don't mean assembler - people had large 16K games they had entirely coded directly in hex based on memorizing the instruction set! I recall one such individual being asked about a possible change to a game and on the spot jumping in an patching it to make work!
What paintbrush should I use?
I also like reading books and want to write one. What kinda typewriter do the pros use?
As a backup plan I'm pretty sure I could make a career as a great Chef since I sure like eating!
Can anyone recommend a frying pan?
Yep - that's even better if you can get some exercise without even trying.
I think that's true of Europe in general (I'm originally from the UK) vs the US (where I now live)... People do still walk into and about town. The US is for the most part totally car-centric. Many residential neighborhoods don't even have sidewalks (= pavements).
However at this point, 3D VMs seem to be an experimental playtoy, not something that can be used for serious gaming.
It makes no sense to lump OpenGL and Direct3D together as "3D" when you're talking about VirtualBix, since they are implemented in very different ways.
VirtualBox OpenGL is basically just as pass-thru to the host driver. The guest box additions includes a virtual OpenGL driver that just passes the commands thru to the the host and the real driver. There must be some performance hit, but the approach seems simple enough.
VirtualBox Direct3D is implemented using the WINE driver that converts Direct3D calls into OpenGL which then get tunneled through to the host OpenGL driver as in the OpenGL case. VirtuaBox Direct3D should therefore be similar in functionality to that in WINE. One upside to the approach is that you don't need a Windows host to have D3D guest aceleration.
During your "lunchtime", take 30 minutes for a brisk walk (not stroll) every day. Around the parking lot or wherever is convenient.
You'd be surprised how much weight you can lose, and quickly, this way. I lost 25lb (210->185lb) in less than 6months.
You can help buy cutting out any snacks, but no need to change your diet. Just the walking will do it.
If you can't find the time to walk for 30 minutes a day, then you don't want to.
That guys usage does seem a bit strange, but each to their own!
However, there are different use cases for tabs that result in different numbers of tabs open:
e.g.
1) User wants to keep a number of sites open all day for quick switching - e.g. mail, news, documentation, etc. This is probably what you are thinking of, and it most likely is a relatively small number of sites - less than a dozen, say.
2) User wants to open a whole bunch of links off the sam page at once, since that's more convenient than flipping back and fro to open one read it, open the next, etc. Examples of this usage would be, for example opening search results (search engine, eBay, etc) or reading new posts on an online forum.
I routinely (many times a day - searching for collectables - made a $1500 profit on one yesterday) open 40-50 tabs at a time for eBay search results since it's so much faster to quickly run down a list of links doing open-in-new-tab (in background) then reading/discarding them, as opposed to doing it one at a time.
Without an IBM, Red Hat, or Mozilla bringing cash and discipline to an open-source project, including paying people to do the 'dirt work' that no one would otherwise do, can open source hope to thrive?"
Maybe not to the degree it has... Linux has certainly greatly benefited from the commercial distros and supporters (e.g. IBM) that have committed cash to adding polish and filling gaps. Unfortunately of course there's only so much they can do - they can commit manpower and technology to individual projects of strategic interest to themselves, but the overall degree of polish of a distro is necessarily heavily reliant on the individual pieces. Each appication/subsystem may be a gem on it's own, but without sufficient standards to guide them (or even desire to adhere to someone else's standards when it's your hobby project done on your time), the resulting pile of gems may be an incoherent and inconsistent mess.
But OTOH, open source thriving and the success of Linux on the desktop (which is the area where the inconsistency and lack of polish hits) are two different things. Linux is already thriving in other areas such as servers and embedded use, and many open source projects such as GNU are used on many platforms other than Linux.
Linux on the desktop is already, and has been for a while, plently good enough (much better than Windows, as is the nature of Unix) for developers who want to use it as a development environment, and seeing as these are the people who created it, that's good enough. If Linux on the desktop never becomes polished enough for a commercial distro to make big bucks off the back of open source developers, then why should we care?
Ben
But my guess is that this douche-bag NBIS company could hire enough lawyers to make it not worth anyones money to find out.
I doubt it, because it sounds like the owner of the installed system, the San Francisco Municipal Transit Agency, believes it owns the rights to the data it's generating (train arrival estimates).
From the article:
Muni spokesperson Judson True says otherwise. In fact, he says that, no, Muni owns the data in question and that the public is, of course, entitled to access it. In fact, he went even further: Muni isn't just giving us all permission to access the data, they're also committed to finding ways to make it easier to get to it.
Of course it could be that there's some fine print in the original procurement contract between the SFMTA and NextBus that the Muni spokesman isn't aware of, but given their commitment to making the data available for use it does at least seem as if that's a legal fight that'll be paid for by the SFMTA (i.e. the taxpayers who bought the system) rather than by any individual company wanting to use the data.
It would have to be where a network connection of any usable speed is impossible.
Having a network connection to a traditional room-filling air-conditioned supercomputer isn't an alternative unless they only want a couple of these, which I doubt is the case (on a side note I'd guess the input bandwidth they need isn't that great - probably just real-time video/voice).
I think it's more a matter of wanting to have this sort of computing power available in the field on a potentially widespread (not one-off) scale. They probably have in mind things like AI and advanced image/pattern recognition for robo-warrior type applications.
The SiCortex website mentions their top of the line machine using about 900mW per 1.4GFlop processor (1.55Glop/W), but the overall system uses 20KW for a 5832 processor system, giving a system figure of (5832 * 1.4) / 20,000 = 0.40GFlop/W.
http://www.pathscale.com/products/high_capability_system_sc5832
Now, consider that DARPA wants 50GFlop/W - a factor of 100x improvement over the SiCortex number.
So, I'd hardly classify SiCortex's products, cool as they may be, as "just this sort" of energy-efficient supercomputer.
In fact the IBM Roadrunner supercomputer TFA refers to does better thsn SiCortex's system, and achieves 0.44GFlop/W.
http://en.wikipedia.org/wiki/IBM_Roadrunner
Just keepin' it real!