My personal taste is in the middle. I appreciate minimalistic lightning fast websites, and these are the ones that use minimal CSS - fonts, sizes, few if ever boxes, colors, and background colors. That is it. The website ends up looking very decent, modern, and loads in an instant. Sometimes one can do buttons too. Otherwise it is all more than enough, and like you said if it is functional - people come, download what they need or read a written piece, and leave. Mission accomplished. But for many out there, the WWW is a road sign, and they use all the real estate they can to blast the visitor with information and what is much worse - bad style. Not everybody is apparently a minimalist, either that or their motivation with a website differs from mine.
And, by the way, your email address is easily harvestable, they (those who care) read and decode it from an image just fine. It has been done for at least a year now, and was the precursor to the breaking of many CAPTCHAs out there.
You can output the email with javascript. It will go on the page as text, but will be decoded first from an encoded string with a key, and inserted in the document on the fly. I am not aware of any robots that do Javascript, so they will not operate the script, although they will sweep through it, looking for addresses.
Point taken, although I am pretty sure youre pulling a joke on me. I dont do IE6 websites, you may. People seek information, and they do not strictly need a three column page layout for that. If they do, a better browser is just a download away for most of the cases. Clients do not pay for two websites after all.
Ugh, I remember the days when IE6 was all the rage, and how difficult it was to explain to management that the future of web is W3C standards and how important it was to design websites with future in mind, according to standards, and how easy it was (even at the time) to switch to Firefox or Opera. Nobody important enough listened, and we were payed to heed and implement those awful CSS hacks for that horrible, horrible commercial abomination of a web browser that is IE6. I had had so much of it already then, even now I can't bring myself to use any version of it to read web pages. This is Microsoft caring for its users at its 'best'. Talk about reputation. Stubborn as they were, keeping to their bad code, excusing themselves for their corporate stock holders.
But then again, there were the days when we used to slice images in Photoshop and Dreamweaver and cooked up monstrous HTML-spaghetti webpages out of such ingridients as tables, javascript browser detection of the if(useragent.indexOf("MSIE"))... kind, and a really bad mix of inline DHTML events and the worst part of it all - document.write - IEs best tool for the DHTML job. Those websites were so fragile, they would collapse on browser switch twice a day for free, but I remember they were payed for in hundreds of thousands if not millions of norwegian crowns. Yuck. The only thing that matters is the looks - if it looks like it works, it must be working.
Just do what I used to do, when I was doing web development (and they payed me for it) - Disable CSS linking for IE6 altogether by not sending the <link type="text/css" rel="stylesheet".../>. If you use PHP, just read the HTTP user agent header and if it is IE 6, do not output the LINK element. If you are as good as you seem to be, catering for webstandards and all, chances are your webpage is readable WITHOUT stylesheets, and NOBODY has complained to me yet about bad looking black on white webpage. It is when things stop working they complain, but when there is no style at all they see, there is nothing to complain about. Webpages are free, since visitors seldom pay to see them, I do not feel guilty discriminating against a web browser, since it cannot display stylesheet properly anyway. The rest of CSS quirks that work differently in Firefox and Opera can be worked out, but IE6 is just too alien for my web-dev tastes. I used to ask for extra money to do IE6 web-dev before, but of course nobody wanted to see that part of the budget, so instead they get a no-style (X)HTML page which works. Even in Lynx, with proper mime type and headers. If your boss or a client threatens to break your kneecaps for leaving out IE6 support brutally like that, make a simple stylesheet from scratch just for IE, small one, with fonts, colors and backgrounds, no fancy box model usage and selectors it has not even heard about. It might end up looking decent, ant it only took you a quarter.
Yes it is a must, and Intel HD4500 does it. There is an undeserved stigmata on integrated graphics solutions, even if it is Intel, when these perform adequately for most of CAD people out there.
I used to design an Audi TT 3d model from NURBS surfaces with an ATI Rage 128 graphics card, and I dont remember any piece of hardware stopping my productivity. Point is, even medium 3D scenes can be designed with an OpenGL compatible integrated graphics solution.
You are trying to justify NVidias solutions with all the wrong arguments. What you wanted to say is probably "We need it for hardcore gaming, and hardcore gaming pushes the industry forwards, which is why we need it as well".
I don't know what he was TRYING to say, but saying "the CPU has run its course" is nonsense, which is what I was trying to say. Especially given the fact that Tegra line, to which APX-2500 (not ATX but APX) belongs, is a computer-on-chip system with its own CPU indeed.
Are you saying we NEED NVidia GPUs to do text, vectors and anti-aliasing?
The ONLY reason one needs an NVidia GPU is specialized scientific applications that have no perceptual needs of limits of computation, and games (which are a subset of those scientific applications in a way they are designed).
Intel graphics chips can do hundreds of millions of antialiased primitives with OpenGL, so even if NVidias GPUs do hundreds of millions at four times the wattage, nobody is impressed. And that is a fact.
I can honestly say I am pretty happy with the GMA X3100 that sits in my Thinkpad T61, and that is not even the latest generation, which is the GMA HD4500, which does H.264 decoding in hardware, consumes even less power while doing more per watt. Compiz on my Ubuntu flies just like I want it to, and while sometimes I hope it would run latest Windows games better (I dualboot into Windows when playing Windows games) somehow it runs them still.
I do not think integrated graphics will be dead, as it is anyway. The lessons learned perfecting it, lead to energy-efficient hardware, tighter pipelines and execution units and also paves way for trully parallel systems. Imagine they shrink the GMA chip down to 25nm, make it use about 1cm of real estate space (i am looking a couple of years down the future road) - with the modest power it delivers per watt, they can scale it up hundreds of times for a truly parallel system of 100 chips, all the while it consuming just about as much power (~1350W) as 8 top-of-the-line Nvidia cards today do, and even though the latter is vastly more powerful, even 8 NVidia chips are no match for a massively parallel 100 GMA chip system. Something like that, anyway. By then, granted, NVidia will have something vastly more powerful than todays offerings, but I doubt the energy consumption will go down.
It is true that systems of all kinds get energy efficient, but my laptop still consumes 8 watts when idle, which is about 10 times less than an average desktop system at its best. So, in a way, we are not really accomplishing much. That is why I am inclined to believe Intel should continue with its integrated graphics, Larrabee and the superscalar generic CPUs it does best. NVidia seems to want their share of a broader market, but they are not exactly doing anything to get there, except of bashing Intel, which is what I critisized in another commend here.
The integrated hardware you do not seem to be fond of, provides me with the much needed laptop battery time that goes directly into productive work. I do not need a half-megawatt machine that weighs 50lbs and produces 50dB of noise to write and compile code, which is what I am doing. So, for me, it makes sense that they would concentrate on integrated stuff actually, at least on hardware that consumes less power and accomplishes a set of goals for a specific need. We have different 'right direction's, it seems.
It is in fact NVidia that opens mouth where it should not. License valid or not, using every opportunity to do their usual "CPU is dead, long live (Nvidia's) GPU!" line won't fool anyone but the dumbest. Here they go again - "We are confident that our license, as negotiated, applies. At the heart of this issue is that the CPU has run its course and the soul of the PC is shifting quickly to the GPU. This is clearly an attempt to stifle innovation to protect a decaying CPU business." Of course you are confident, you have to defend yourself, what else would you be saying? But watch as the CEO hijacks the opportunity again and bash Intel and CPUs in general. Well guess what mr. CEO, you HAVE to have a central processor (oh-oh, that's what the CP in CPU stands for, did you know?), that's not just one doing 3D graphics, which is what you have been doing all these years, but actually centralizes resource management in the system, and even facilities that do not strictly need the CPU still use it for coordination - memory controller, FPU, DMA, interrupt controllers and more. Without the CPU, where do we send orders on bootstrapping the system to a usable and programmable and serviceable state? Tell you what, mr. CEO, - When you have a chip that not only does teraflops with SIMD but can actually CONTROL a motherboard, let us know. Until then be humble and let Intel and others provide. Nobody said you cannot compete, but a GPU alone does not make a system, even less so one that requires a special compiler to make it run generic code. NVidia should learn the meaning of humble, they think slapping together an SLI chip that consumes 200W and only works for two years is a good job done, they think wrong. I'll stick with a 25W x86 CPU for a while longer thank you. And no, NVidia "south"- and "north"-bridge chipsets are not good enough.
I understand what you are trying to say, but the factor is insignificant in comparison to the performance bottlenecks and the realm of problems Firefox and Linux currently have. A slow page refresh, or slow user response dont have much to do with either paging or type of linking, granted too much paging does irritate most of us. Nevertheless, as long as Firefox enjoys its share of RAM, it seldom pages, and like I said, whole different things matter. Rule of optimization - run profiler. Don't start improving Firefox with improving dynamic linker and system paging mechanism. If these were really slow, we would have noticed elsewhere, but as these are good enough to go unnoticed for other more serious problems, I say tidy own backyard first, then start cleaning up the street. That said and rest assured, improving on CPU scheduler, system paging and perhaps even dynamic linker - system wide facilities - is just about the most important kind of optimization there is because it benefits WHOLE system, not just Firefox or X. But that does not mean allowing sloppy code and shove off responsibility on kernel. My point is, improving Firefox' code will make it faster sooner than improving kernel will. When FF is good enough, then waiting for a better kernel never hurts:-)
I dont see how that factor plays any significant role where things like optimizing Firefox code, X drivers, and system libraries are the real issue. I am talking about importances. Cost of paging and dynamic linking is not a factor in the grand scheme of things Firefox and Linux today. When Firefox squeezes whatever it can out of its own code, and when system and user libraries do the same, and after performance is asserted, then things like paging and dynamic linking costs can be re-evaluated. I am not saying kernel paging mechanism is not important as a system, but the way it is now, it is out of the way of those seeking performance, and so is dynamic linking.
It may be fast enough for you, but on my laptop its inneficiency shows itself in subpar battery life and thermal characteristics - inefficient program code wakes the CPU up, controller starts the system fan, it gets loud and warm, i turn off and use pen and paper instead. This is DIRECT consequence of lazyness. I am lazy, you are lazy, Mozilla is lazy. Inching along, proud of our own mediocrity. But I understand Mozillas reasons - this is as good as it gets, or pay more to devs. Windows is prima target, or else shut up, they may say. And they are right. We live in the real world and have to do it without loosing grip on reality. That said, situation with Linux is improving...
Yes I know, and you are absolutely right. But then it does not make much sense for Wine folks to tout that Wine is portable. Because once it is ported to another platform, and has to perform its function where it runs win32-x86 applications, there has to be some sort of translation obviously. If it would be called emulating, I do not know, but by that time the acronym has lost its original meaning and value.
And, again, I am a programmer and am pretty well aware of the subtle differences between JITting, compiling and emulating on the fly. Are you a programmer? I can try to explain how x86 Windows apps can be run under ARM _WITHOUT_ emulation. But am too lazy to bother just for the sake of explanation. And it is not like Wine folks are themselves unaware of ways in which it can be done. I also mentioned bits and pieces of the issues in other comments to this same article.
Enshortened personal opinionating: 1. People understand computers just fine, it is the constructs like Windows (and Linux too) they cannot get their minds around. Habits help, but re-learning sucks dont it? Why is it Windows habits are better than Linux habits? I honestly do not see why my mom should use Windows? she just does three things, and all those three work fine. It is not like Windows install is a walk in the park, with all the drivers etc. When it works, yes, but then again it is the same story with Linux installers. I say go Linux, and let us wait and see. Things are changing. 2. Instead of 'supporting' hardware, give it wings. Dont have it crawling on a leash, let it out in the big world together with programming reference manual, and you dont have to think about it anymore! Of course greed and control are by many considered healthy human qualities. And I agree, but I see potential in free hardware, where others see loss without doubt.
Ok, i had something a bit different in mind - put one little (they are tiny, these ARMs, most of them anyway) ARM as the CPU together with one x86 unit and optionally a GPU, all in one normal laptop, work out logic where the latter two PUs can by completely controlled by the ARM (it being the CPU of the bunch), which is able to also coldstart and shutdown the other two as programmer sees fit? That sounds like a good thing to me, but I do not own a hardware company.
What you are proposing has a slightly different application. It makes sense at home, and is doable even now - you bring a tiny laptop to your home, plug it in into a dock or just use it as a remote control (keyboard, mouse, screen) for the much more powerful desktop you have in the corner. But the software (that is quality) for that is just not there.
Ok. So it does. I know nobody who runs it. My point was meanwhile, as Microsoft releases niche Windows versions that may be portable and are indeed sort of ported to other platforms, the mainstream Windows line of products that majority of us has used, uses and will continue to use is not ported to ARM but remains the only option of running Windows applications. Besides WINE, which if ported itself, will sort of open the port (pun intended) for many Windows products (the list is expanding) to reach a broader 'ecosystem'. Speaking besides point, I hope Microsoft starts re-writing Windows in a portable way, since they seem so concerned with open source suddenly. If that is how they will survive the open source, fine by me. Windows GDI is said to be written in assembly (despite being layered on top of Windows HAL), but Microsoft likes to mention these days it is 'deprecated'. I have not been paying attention to Windows environment in a good while, since I stopped developing for it. But I am pretty sure the chances of Windows being ported succesfully are just as slim as with porting Wine (in a comparable time period).
What, your stupid 'Hello World' code does not sell? Try lowering the price and offer source code, you might get suggestions on market value and code quality instead.
Linux does not work on freeloading. Freeloading is just a sideeffect, and not necessarily unfortunate at that. Usually those who get free code produce code too. Logical thinking is part of human mind, it is the elite industry who is trying to have you believe only they can give you the true quality product and the reasons there are so many zombie programmers producing zombie code (Hello World application, Visual Studio 2009 style!) is exactly because code has a price tag, and time is money, no time to REALLY think (logically), time to produce stuff!
With freely available source code, more people can participate in improving quality of product. Also, after having bought a blackbox commercial product, they are not denied changes and are threatened with lawsuit if they just want to make it work for them - the product they have legally bought. Open source makes people interested in it independent, it divorces them from the partner they did not want to have - the vendor, who only thinks with their wallet, so obviously it is natural there are people who want to just buy something that works, and there are people who want to buy something that they can adapt their needs. I belong to the latter group (i can actually code), and it looks as if that group is miraculously growing.
Want a job, go work in Africa, there is a good chance their ancestors were freeloaded by your ancestors just over a hundred years ago, you owe them that. Why are you complaining on behalf people who do not want to be complained for? Did I raise the price on your house loan or something by not paying for my software? In that case, why should I care? Wolf against wolf, just like you seem to like it.
There is always market for a good pair of hands, and I am not talking about keyboard typing. When I choose to 'waste' my life in 'my mothers basement' at least I do not ask of others to pay me for that. Because I work somewhere else entirely.
I produce code and leave it open, but I also use code that is left open.
Stop putting a price on everything, you have your head too deep in that market economy improve-yourself-and-get-rich self-help book you keep around. Want to be MONEY rich, do something people WANT to pay MONEY for. Evidently the times when people did want to pay for code are gone.
You may read the WINE's Myth.10 on their webpage, they explicitly tell of some of the problems associated with porting WINE to other architectures. Just because it has never been done does not mean it cannot be done, or is monumentally difficult. It took long enough to set WINE up and running as it is, so it is not a priority to get it ported, but i do not see a reason it cannot be used for running legacy Win32 apps on alien platforms. Even as Windows itself has not been ported to anything else than x86.
As to JIT vs compilation, i did imply these are not very diffrent. The whatever difference these have does almost result in a world of difference literally though. JITting usually favours speed of compilation, since it has to do it just in time (hehehe), so it leaves out some of the optimizations and reaction to condition checks standalone compilers include, because it has to run the program today, right now, not tomorrow. There are other things too, but in total these two techniques that have the same conceptual root bring about two different classes of programs, which is why I differentiate between the two. But something tells me you know how all that works yourself:-)
Well, yes it would be the biggest problem, but I do not think it would be a BIG problem in itself. After all one x86 CPU is able and does run two applications in parallel.
I was commenting on Windows being a limiting factor, and that is the kind of x86 binary I was referring to. Since WINE itself is open source, it should not be too hard at all to let it run x86-win32 applications in arm-linux environments. This is what I was trying to say. Rewriting is also too generic a word for the discussion. There is JIT and there is compiling, and the two, although based on exactly the same routine, deliver two different results, performance- and battery-life- wise.
Slashdot Mirror
My personal taste is in the middle. I appreciate minimalistic lightning fast websites, and these are the ones that use minimal CSS - fonts, sizes, few if ever boxes, colors, and background colors. That is it. The website ends up looking very decent, modern, and loads in an instant. Sometimes one can do buttons too. Otherwise it is all more than enough, and like you said if it is functional - people come, download what they need or read a written piece, and leave. Mission accomplished. But for many out there, the WWW is a road sign, and they use all the real estate they can to blast the visitor with information and what is much worse - bad style. Not everybody is apparently a minimalist, either that or their motivation with a website differs from mine.
And, by the way, your email address is easily harvestable, they (those who care) read and decode it from an image just fine. It has been done for at least a year now, and was the precursor to the breaking of many CAPTCHAs out there.
You can output the email with javascript. It will go on the page as text, but will be decoded first from an encoded string with a key, and inserted in the document on the fly. I am not aware of any robots that do Javascript, so they will not operate the script, although they will sweep through it, looking for addresses.
Point taken, although I am pretty sure youre pulling a joke on me. I dont do IE6 websites, you may. People seek information, and they do not strictly need a three column page layout for that. If they do, a better browser is just a download away for most of the cases. Clients do not pay for two websites after all.
Ugh, I remember the days when IE6 was all the rage, and how difficult it was to explain to management that the future of web is W3C standards and how important it was to design websites with future in mind, according to standards, and how easy it was (even at the time) to switch to Firefox or Opera. Nobody important enough listened, and we were payed to heed and implement those awful CSS hacks for that horrible, horrible commercial abomination of a web browser that is IE6. I had had so much of it already then, even now I can't bring myself to use any version of it to read web pages. This is Microsoft caring for its users at its 'best'. Talk about reputation. Stubborn as they were, keeping to their bad code, excusing themselves for their corporate stock holders.
But then again, there were the days when we used to slice images in Photoshop and Dreamweaver and cooked up monstrous HTML-spaghetti webpages out of such ingridients as tables, javascript browser detection of the if(useragent.indexOf("MSIE"))... kind, and a really bad mix of inline DHTML events and the worst part of it all - document.write - IEs best tool for the DHTML job. Those websites were so fragile, they would collapse on browser switch twice a day for free, but I remember they were payed for in hundreds of thousands if not millions of norwegian crowns. Yuck. The only thing that matters is the looks - if it looks like it works, it must be working.
Just do what I used to do, when I was doing web development (and they payed me for it) - Disable CSS linking for IE6 altogether by not sending the <link type="text/css" rel="stylesheet" ... />. If you use PHP, just read the HTTP user agent header and if it is IE 6, do not output the LINK element. If you are as good as you seem to be, catering for webstandards and all, chances are your webpage is readable WITHOUT stylesheets, and NOBODY has complained to me yet about bad looking black on white webpage. It is when things stop working they complain, but when there is no style at all they see, there is nothing to complain about. Webpages are free, since visitors seldom pay to see them, I do not feel guilty discriminating against a web browser, since it cannot display stylesheet properly anyway. The rest of CSS quirks that work differently in Firefox and Opera can be worked out, but IE6 is just too alien for my web-dev tastes. I used to ask for extra money to do IE6 web-dev before, but of course nobody wanted to see that part of the budget, so instead they get a no-style (X)HTML page which works. Even in Lynx, with proper mime type and headers. If your boss or a client threatens to break your kneecaps for leaving out IE6 support brutally like that, make a simple stylesheet from scratch just for IE, small one, with fonts, colors and backgrounds, no fancy box model usage and selectors it has not even heard about. It might end up looking decent, ant it only took you a quarter.
Yes it is a must, and Intel HD4500 does it. There is an undeserved stigmata on integrated graphics solutions, even if it is Intel, when these perform adequately for most of CAD people out there.
I used to design an Audi TT 3d model from NURBS surfaces with an ATI Rage 128 graphics card, and I dont remember any piece of hardware stopping my productivity. Point is, even medium 3D scenes can be designed with an OpenGL compatible integrated graphics solution.
You are trying to justify NVidias solutions with all the wrong arguments. What you wanted to say is probably "We need it for hardcore gaming, and hardcore gaming pushes the industry forwards, which is why we need it as well".
I don't know what he was TRYING to say, but saying "the CPU has run its course" is nonsense, which is what I was trying to say. Especially given the fact that Tegra line, to which APX-2500 (not ATX but APX) belongs, is a computer-on-chip system with its own CPU indeed.
Are you saying we NEED NVidia GPUs to do text, vectors and anti-aliasing?
The ONLY reason one needs an NVidia GPU is specialized scientific applications that have no perceptual needs of limits of computation, and games (which are a subset of those scientific applications in a way they are designed).
Intel graphics chips can do hundreds of millions of antialiased primitives with OpenGL, so even if NVidias GPUs do hundreds of millions at four times the wattage, nobody is impressed. And that is a fact.
I can honestly say I am pretty happy with the GMA X3100 that sits in my Thinkpad T61, and that is not even the latest generation, which is the GMA HD4500, which does H.264 decoding in hardware, consumes even less power while doing more per watt. Compiz on my Ubuntu flies just like I want it to, and while sometimes I hope it would run latest Windows games better (I dualboot into Windows when playing Windows games) somehow it runs them still.
I do not think integrated graphics will be dead, as it is anyway. The lessons learned perfecting it, lead to energy-efficient hardware, tighter pipelines and execution units and also paves way for trully parallel systems. Imagine they shrink the GMA chip down to 25nm, make it use about 1cm of real estate space (i am looking a couple of years down the future road) - with the modest power it delivers per watt, they can scale it up hundreds of times for a truly parallel system of 100 chips, all the while it consuming just about as much power (~1350W) as 8 top-of-the-line Nvidia cards today do, and even though the latter is vastly more powerful, even 8 NVidia chips are no match for a massively parallel 100 GMA chip system. Something like that, anyway. By then, granted, NVidia will have something vastly more powerful than todays offerings, but I doubt the energy consumption will go down.
It is true that systems of all kinds get energy efficient, but my laptop still consumes 8 watts when idle, which is about 10 times less than an average desktop system at its best. So, in a way, we are not really accomplishing much. That is why I am inclined to believe Intel should continue with its integrated graphics, Larrabee and the superscalar generic CPUs it does best. NVidia seems to want their share of a broader market, but they are not exactly doing anything to get there, except of bashing Intel, which is what I critisized in another commend here.
The integrated hardware you do not seem to be fond of, provides me with the much needed laptop battery time that goes directly into productive work. I do not need a half-megawatt machine that weighs 50lbs and produces 50dB of noise to write and compile code, which is what I am doing. So, for me, it makes sense that they would concentrate on integrated stuff actually, at least on hardware that consumes less power and accomplishes a set of goals for a specific need. We have different 'right direction's, it seems.
It is in fact NVidia that opens mouth where it should not. License valid or not, using every opportunity to do their usual "CPU is dead, long live (Nvidia's) GPU!" line won't fool anyone but the dumbest. Here they go again - "We are confident that our license, as negotiated, applies. At the heart of this issue is that the CPU has run its course and the soul of the PC is shifting quickly to the GPU. This is clearly an attempt to stifle innovation to protect a decaying CPU business." Of course you are confident, you have to defend yourself, what else would you be saying? But watch as the CEO hijacks the opportunity again and bash Intel and CPUs in general. Well guess what mr. CEO, you HAVE to have a central processor (oh-oh, that's what the CP in CPU stands for, did you know?), that's not just one doing 3D graphics, which is what you have been doing all these years, but actually centralizes resource management in the system, and even facilities that do not strictly need the CPU still use it for coordination - memory controller, FPU, DMA, interrupt controllers and more. Without the CPU, where do we send orders on bootstrapping the system to a usable and programmable and serviceable state? Tell you what, mr. CEO, - When you have a chip that not only does teraflops with SIMD but can actually CONTROL a motherboard, let us know. Until then be humble and let Intel and others provide. Nobody said you cannot compete, but a GPU alone does not make a system, even less so one that requires a special compiler to make it run generic code. NVidia should learn the meaning of humble, they think slapping together an SLI chip that consumes 200W and only works for two years is a good job done, they think wrong. I'll stick with a 25W x86 CPU for a while longer thank you. And no, NVidia "south"- and "north"-bridge chipsets are not good enough.
I understand what you are trying to say, but the factor is insignificant in comparison to the performance bottlenecks and the realm of problems Firefox and Linux currently have. A slow page refresh, or slow user response dont have much to do with either paging or type of linking, granted too much paging does irritate most of us. Nevertheless, as long as Firefox enjoys its share of RAM, it seldom pages, and like I said, whole different things matter. Rule of optimization - run profiler. Don't start improving Firefox with improving dynamic linker and system paging mechanism. If these were really slow, we would have noticed elsewhere, but as these are good enough to go unnoticed for other more serious problems, I say tidy own backyard first, then start cleaning up the street. That said and rest assured, improving on CPU scheduler, system paging and perhaps even dynamic linker - system wide facilities - is just about the most important kind of optimization there is because it benefits WHOLE system, not just Firefox or X. But that does not mean allowing sloppy code and shove off responsibility on kernel. My point is, improving Firefox' code will make it faster sooner than improving kernel will. When FF is good enough, then waiting for a better kernel never hurts :-)
I dont see how that factor plays any significant role where things like optimizing Firefox code, X drivers, and system libraries are the real issue. I am talking about importances. Cost of paging and dynamic linking is not a factor in the grand scheme of things Firefox and Linux today. When Firefox squeezes whatever it can out of its own code, and when system and user libraries do the same, and after performance is asserted, then things like paging and dynamic linking costs can be re-evaluated. I am not saying kernel paging mechanism is not important as a system, but the way it is now, it is out of the way of those seeking performance, and so is dynamic linking.
Looking indefinitely for that which really does not exist has always been the curse (or blessing) of man. Some call it faith.
It may be fast enough for you, but on my laptop its inneficiency shows itself in subpar battery life and thermal characteristics - inefficient program code wakes the CPU up, controller starts the system fan, it gets loud and warm, i turn off and use pen and paper instead. This is DIRECT consequence of lazyness. I am lazy, you are lazy, Mozilla is lazy. Inching along, proud of our own mediocrity. But I understand Mozillas reasons - this is as good as it gets, or pay more to devs. Windows is prima target, or else shut up, they may say. And they are right. We live in the real world and have to do it without loosing grip on reality. That said, situation with Linux is improving...
Yes I know, and you are absolutely right. But then it does not make much sense for Wine folks to tout that Wine is portable. Because once it is ported to another platform, and has to perform its function where it runs win32-x86 applications, there has to be some sort of translation obviously. If it would be called emulating, I do not know, but by that time the acronym has lost its original meaning and value.
And, again, I am a programmer and am pretty well aware of the subtle differences between JITting, compiling and emulating on the fly. Are you a programmer? I can try to explain how x86 Windows apps can be run under ARM _WITHOUT_ emulation. But am too lazy to bother just for the sake of explanation. And it is not like Wine folks are themselves unaware of ways in which it can be done. I also mentioned bits and pieces of the issues in other comments to this same article.
Enshortened personal opinionating:
1. People understand computers just fine, it is the constructs like Windows (and Linux too) they cannot get their minds around. Habits help, but re-learning sucks dont it? Why is it Windows habits are better than Linux habits? I honestly do not see why my mom should use Windows? she just does three things, and all those three work fine. It is not like Windows install is a walk in the park, with all the drivers etc. When it works, yes, but then again it is the same story with Linux installers. I say go Linux, and let us wait and see. Things are changing.
2. Instead of 'supporting' hardware, give it wings. Dont have it crawling on a leash, let it out in the big world together with programming reference manual, and you dont have to think about it anymore! Of course greed and control are by many considered healthy human qualities. And I agree, but I see potential in free hardware, where others see loss without doubt.
I liked the story in Doom 3. It made all the killing more intense [fun]. Simple as the story may seem, it was all in the details.
Ok, i had something a bit different in mind - put one little (they are tiny, these ARMs, most of them anyway) ARM as the CPU together with one x86 unit and optionally a GPU, all in one normal laptop, work out logic where the latter two PUs can by completely controlled by the ARM (it being the CPU of the bunch), which is able to also coldstart and shutdown the other two as programmer sees fit? That sounds like a good thing to me, but I do not own a hardware company.
What you are proposing has a slightly different application. It makes sense at home, and is doable even now - you bring a tiny laptop to your home, plug it in into a dock or just use it as a remote control (keyboard, mouse, screen) for the much more powerful desktop you have in the corner. But the software (that is quality) for that is just not there.
Ok. So it does. I know nobody who runs it. My point was meanwhile, as Microsoft releases niche Windows versions that may be portable and are indeed sort of ported to other platforms, the mainstream Windows line of products that majority of us has used, uses and will continue to use is not ported to ARM but remains the only option of running Windows applications. Besides WINE, which if ported itself, will sort of open the port (pun intended) for many Windows products (the list is expanding) to reach a broader 'ecosystem'. Speaking besides point, I hope Microsoft starts re-writing Windows in a portable way, since they seem so concerned with open source suddenly. If that is how they will survive the open source, fine by me. Windows GDI is said to be written in assembly (despite being layered on top of Windows HAL), but Microsoft likes to mention these days it is 'deprecated'. I have not been paying attention to Windows environment in a good while, since I stopped developing for it. But I am pretty sure the chances of Windows being ported succesfully are just as slim as with porting Wine (in a comparable time period).
The NT 4.0 operating system ran on four different architectures -- x86, PowerPC, MIPS, and Alpha.
(: I KNEW THAT!
Seriously though, yes I know NT was, but my concerns lie in the stuff M$ has today - mainly Vista platform.
What, your stupid 'Hello World' code does not sell? Try lowering the price and offer source code, you might get suggestions on market value and code quality instead.
Linux does not work on freeloading. Freeloading is just a sideeffect, and not necessarily unfortunate at that. Usually those who get free code produce code too. Logical thinking is part of human mind, it is the elite industry who is trying to have you believe only they can give you the true quality product and the reasons there are so many zombie programmers producing zombie code (Hello World application, Visual Studio 2009 style!) is exactly because code has a price tag, and time is money, no time to REALLY think (logically), time to produce stuff!
With freely available source code, more people can participate in improving quality of product.
Also, after having bought a blackbox commercial product, they are not denied changes and are threatened with lawsuit if they just want to make it work for them - the product they have legally bought. Open source makes people interested in it independent, it divorces them from the partner they did not want to have - the vendor, who only thinks with their wallet, so obviously it is natural there are people who want to just buy something that works, and there are people who want to buy something that they can adapt their needs. I belong to the latter group (i can actually code), and it looks as if that group is miraculously growing.
Want a job, go work in Africa, there is a good chance their ancestors were freeloaded by your ancestors just over a hundred years ago, you owe them that. Why are you complaining on behalf people who do not want to be complained for? Did I raise the price on your house loan or something by not paying for my software? In that case, why should I care? Wolf against wolf, just like you seem to like it.
There is always market for a good pair of hands, and I am not talking about keyboard typing. When I choose to 'waste' my life in 'my mothers basement' at least I do not ask of others to pay me for that. Because I work somewhere else entirely.
I produce code and leave it open, but I also use code that is left open.
Stop putting a price on everything, you have your head too deep in that market economy improve-yourself-and-get-rich self-help book you keep around. Want to be MONEY rich, do something people WANT to pay MONEY for. Evidently the times when people did want to pay for code are gone.
You may read the WINE's Myth.10 on their webpage, they explicitly tell of some of the problems associated with porting WINE to other architectures. Just because it has never been done does not mean it cannot be done, or is monumentally difficult. It took long enough to set WINE up and running as it is, so it is not a priority to get it ported, but i do not see a reason it cannot be used for running legacy Win32 apps on alien platforms. Even as Windows itself has not been ported to anything else than x86.
As to JIT vs compilation, i did imply these are not very diffrent. The whatever difference these have does almost result in a world of difference literally though. JITting usually favours speed of compilation, since it has to do it just in time (hehehe), so it leaves out some of the optimizations and reaction to condition checks standalone compilers include, because it has to run the program today, right now, not tomorrow. There are other things too, but in total these two techniques that have the same conceptual root bring about two different classes of programs, which is why I differentiate between the two. But something tells me you know how all that works yourself :-)
Well, yes it would be the biggest problem, but I do not think it would be a BIG problem in itself. After all one x86 CPU is able and does run two applications in parallel.
Right now I can drive up the street ...
Do you drive on gas?
I was commenting on Windows being a limiting factor, and that is the kind of x86 binary I was referring to. Since WINE itself is open source, it should not be too hard at all to let it run x86-win32 applications in arm-linux environments. This is what I was trying to say. Rewriting is also too generic a word for the discussion. There is JIT and there is compiling, and the two, although based on exactly the same routine, deliver two different results, performance- and battery-life- wise.