Lots of medium companies buy Microsoft software indirectly to leverage up or get consolidated pricing. Large all have relationship with Microsoft but the "sales weasel" might have the relationship with the guys in SQL Server and Dynamics and have no relationship at all with the desktop team.
They can't do that. The heavy duty applications don't exist for Android and won't be ported to Android anytime soon. Dell's entire innovation is customization at a low cost. They can't beat Samsung, LG, Motorolla at making 100m of one model. Android is the worst of all words for Dell. Same with Lenovo. Acer I don't know enough about.
The IBM PC was the one strange thing in that you could install any OS on it.
No it was just as locked down. Microsoft, Intel and Western Digital created the standard that allowed IBM compatibles to get compatible enough that this became a multi vendor platform with Intel and Microsoft in the driver's seat.
Non-Apple consumer x86 sales are down 10% year over year. That tablet and cell phone replacement is already happening in consumer.
It was great when there were generous margins and rapidly growing sales. It was good when margins tightened but sales still surged. It was getting kinda sucky when margin collapsed but at least sales kept surging. It got worse when sales flattened out. Now we are facing the real possibility of sales collapsing.
Microsoft is doing what they can to fend that off. If they are going to fend that off they need to change the platform ecosystem: OS, hardware, development libraries, software... and they need to do this fast. They need to do this in a platform that moribund. Hopefully confronted with Windows 8, consumers start buying the devices they will work on. Dell, Asus, HP... now how to make those devices they just don't know how to make them for $500.
The moment this happens we'll know that Microsoft has failed and not even they can escape the black hole that is their own legacy.
I agree with you. Microsoft is going to have to take a page from Apple's book in killing off legacy cruft quickly but with lots of steps to make it look gradual so people don't freak.
Gamers used to be very willing to boot to a gaming OS. It was standard to have a "gaming" configuration of DOS and a "windows" configuration in the Windows 3.0/3.1/95 days. I see no reason if Linux had strong gaming support that their wouldn't be gamer machines sold with Linuxes tweaked to optimize gaming. That honestly is a pretty good niche for Linux, a place where custom kernel could really matter. And heck the hardware OEMs would probably love a product where they could do a real value ad and make some margin.
I have no idea how big the gaming market is, but it seems to be carrying several midsized computer companies. So why not?
I love the DOS 4.0 reference! Yes that was a lousy OS but you can blame IBM for that one, PCDOS 4.0 was what sucked Microsoft was just following their lead on the DOS front for the last time.
I think Windows 8 might work out. I see Windows 8 as potentially being a lot like IE 2. By IE 2 Microsoft had realized this internet thing was real and a real threat, and was aggressively pulling technologies into their browser. IE 2 was a a real browser while IE1 had been a research project. People would still buy Netscape in the IE 2 days but within 2 years after IE 2 they'd be up to IE 4, which was way more advanced than anything Netscape could do and frankly had a lot of really cool features that most browsers even today don't.
Now while will open systems cost as much as a car in this new world?
As for DRM I'm not sure that Apple isn't happy to play DRM either way. When media companies don't play along they side with consumers. When media companies do play along they side with publishers. But if you look at the history of iTunes, every song sold now is DRM free. We have today the market people always asked for, individual songs sold at a reasonable cost DRM free from a variety of vendors in open formats. Things don't always turn out so bad.
I'm not a photo guy but my understanding is that when Aperture came out it was designed to compete with Photoshop, there was no Lightroom. Other than HDR what doesn't Aperture do? I'd assume Photoshop has more features but I'm looking for big stuff. Aperture seems much more full featured than Lightroom.
I own Pixelmator. I'm not sure I'd compare to it Photoshop.
What is the difference between a VM and an emulator?
The intent of the software running on it.
Program A is designed to run with hardware B. Program C is designed to run against VM D.
On B I can run A or C. On some other hardware I can run VM D, but when I try and A I have to emulate.
Abstractly an emulator is just a VM. In practice an emulator is a large, complex, inconsistent VM that alters the behavior of code in unpredictable ways and thwarts far too many of the code's original designers. As the saying goes: in theory there is no difference between theory and practice, in practice there is.
Not anymore. Modern web browsers JIT compile it like Java nowadays
I doubt that. Most likely they JIT it more like Perl, where code can (indirectly) call the evaluator on data. Dynamic languages are much, much harder to write compilers for. I doubt anyone has reduces Javascript, which has full blown lambdas to a compile.
Safari since iOS 4.3 or so even uses a special permission not available to third-party applications to circumvent the platform's compulsory DEP so that it can JIT compile JavaScript code.
I'm aware of that, two versions of webkit when it comes to Javascript. I still use Dolphin as my browser even knowing that though.
So? Did I say the iPhone was a clone of Palms? Apple copied liberally from all major vendors: Nokia, Windows Mobile, and Palm
Let me just stop you here. If Apple produced a unique combination of existing technologies in a way that would work better, that's insight and... that's patentable. I don't have to have invented either the transistor or the radio to be able to patent the transistor radio. The only question would be whether everyone in the radio would have obviously seen the applicability of the transistor of if that applicability was a matter of unique insight.
You can't have it both ways here. Either they copied from Palm or they did something new.
I just don't think it has practical importance in most cases, since there are so many ways to hack into a computer once the OS is up and running (and unlike installing a rootkit at boot, many of them can be executed remotely).
Trusted computing also gets rid of most of those problems too. UEFI doesn't do anything there, but the technology needed to take UEFI to the next level would help quite a bit with hacking exploits. Mainframes don't get hacked nearly as easily because the security model is very different.
I don't doubt the Linux developers will find a way to get around the technical problems. But Redhat apparently thought something as simple as disabling SecureBoot manually, would be enough to scare away some of their customers.
And I think RedHat's right. There is a difference between:
a) Will scare some customers. b) Will make something completely impossible.
(a) and (b) aren't the same thing.
Well... isn't that easier than getting around hardware locks? Content can only be protected by encrypting it, and if legitimate users are to be able to access it, the decryption keys need to be stored in cleartext on their computers.
that's now how it is done. Here is an example. A long message is broken into a string of smaller messages (like like 1mb each). Each small message gets an AES encryption with the AES key stored RSA encrypted along with the block. The RSA private key is on the TCPA chip which can decrypt but will only unlock that set of keys based on the checksum of the kernel performed by the UEFI. Outside the TCPA chip the RSA private key is unknown. And the TCPA chip will only perform decryptions to an authorized kernel. The AES keys are known to the system but they are one time use.
Note that if I release the document to say 30 people, those 30 people correspond to me signing each small message with 30 different AES keys. Or one time a group key can be loaded on the TCPA chip and everyone can share a group key. Private keys are never stored on the computer because the computer as a whole cannot be trusted.
For example, when Microsoft saw Netscape as a threat, they changed the licensing terms for Windows NT Workstation to make it impossible to run web servers on it. You could still run web servers on Windows NT Server, but when you bought NT Server you got a web server for free, so you didn't need Netscape's. Nobody was actually prevented from running Netscape's web server; Microsoft just made sure it didn't make financial sense for most users. (This was one of the practices the DoJ found to be anti-competitive.)
I used to run Netscape's webserver. I have to tell you IIS was a lot better. It wasn't licensing that got me to switch but Active-X, the speed of configuration, the ease of management.... But I do get your point that's a good example. And I agree that Microsoft will nudge not compel. And that was my point, a total lockout is compel.
In reality if a program crashed the Amiga CPU simply jumped to the next program because the CPU maintained control of which task executed.
The Amiga CPU, the 68000, had no idea what a task was or what a program was. It doesn't maintain control of anything. That's the point there is no hardware here that can act independently of the current task. There was no protected memory. If Eudora went down lightly things might be salvageable, for example a scheduled interrupt. But if Eudora has a bad pointer it is corrupting all of memory including the kernel or even the interrupts table.
Just think for a second about what you are claiming here. Figure out how this could work, in detail.
. it's very easy to build a competing product, for say a HTTP server.
I'm putting that down to complexity. An HTTP server extension is usually under a few man years of code. Its easier than an office suite just because it is less complex not because of open standards. I don't think it is any easier to write extensions for Apache than for IIS.
If the real world would be like the end-user software world, you would have 5 different sockets for lamps, 10 different power plugs. And each TV would come with a custom proprietary power and cable plug.
That's what electrical devices were like in the late 19th century. Also if there were a standard 10 years it would have been the Microsoft standard. Microsoft was the defacto standard. The free software community rejected the Microsoft standard and successfully shattered it first via LAMP and then via. Firefox.
The issue is that computers grew like almost no other new technology and with little government help; and government are frequently the ones that impose and enforce standards.Partially that was because of the American mood, we had a new conservatism which created amounts of private investment capital while stripping public investment capital. Partially it was because the productivity savings from computers were so massive and so obvious government investment was not required. But regardless there is a lack of enforced standards.
Nobody would fund you if you start a competing application for, for example, MS Office or Photoshop, with the goal to compete open in the market.
I don't know about that. Apple in 2004 funded a competitor to PhotoShop when Adobe was threatening to drop the platform. And they were successful. Sun did fund a competitor to MS Office for years and took StarOffice from a kinda crudy office suite to something in the same ballpark as MS Office. And let me point out, AOL funded a competitor to IE for 6 years. The last 2 were both open source efforts.
OO.org, LibreOffice, GIMP goal is not to compete commercially, so they survive as a not-for-profit organization.
Actually the goal of OO.org / Star Office was to compete commercially. Sun wanted to push something like the cloud SaaS model ten years earlier. They absolutely wanted to compete with Microsoft commercially. If you mean selling boxes of software Office isn't really that either, though it is a huge money maker it is also the lynchpin of Microsoft's entire server based suite and their OS monopoly.
Because nobody can compete with Adobe in the marketplace.
Of course they can. Heck Adobe just lost to Javascript. Adobe's rip technology lost to HP's (essentially free) PCL. Adobe Premiere got its but kicked by Apple with iMovie and Final Cut Pro/X. GIMP can't beat Adobe doesn't mean no one can.
No I actually meant a virtual PC included in Microsoft office. There are still some links to version of the product on the web for example: http://www.microsoft.com/australia/office/mac/virtualpc7/default.aspx So for example Microsoft Virtual PC 6 came with Office 2004 for Mac (this was after Microsoft bought Connectix).
Anyway I've never heard people consider the experience of a PC via. Parallels to be near native. I'm not sure what that is owing to, but it doesn't seem in practice to be much different than using VirtualPC.
email accounts, often act as a proxy for a member identifier / account identifier. They aren't perfectly unique in either direction. Sometimes multiple people share an email but then they are sharing an account; sometimes the same person has multiple emails but then effectively that person is acting like multiple people.
For most companies the majority of their middleware are desktop productivity applications like Access combined with a semi skilled office worker. A file gets pulled from one server, manipulated by hand, and then sent to another server.
The failure is that.NET on iOS or Android costs a developer hundreds of dollars per year payable to Xamarin
Hold on a second..NET is a platform specific compiler, not a platform independent one. You shouldn't be using.NET for platform independent work, no one has ever said it was cross platform not intended to be. Nor has Microsoft been a fan of cross platform development, in general. They themselves fork their own code base when they support multiple platforms rather than use conversion tools. What you are saying is like saying I have a bad air conditioner because it doesn't sweep the driveway.
The job of a compiler is to take code written in a higher level language and reduce it to hardware / platform executable code. The.NET compiler does a fantastic job on the criteria by which compilers are evaluated, things like optimization efficiency. C# is allowing for some interesting ideas from functional programming to be imported into high speed structural languages without loss of efficiency.
You have an objective that Microsoft doesn't agree with, no support lots of platforms without rewriting your code..NET isn't bad because it doesn't help you do stuff that it was never meant to help you do.
That's what Google Native Client is supposed to be.
Thanks! Don't know how I missed that technology when it was announced. Yes that is exactly the kind of thing I was thinking of.
Java is a VM. Javascript is an interpreted language..NET is virtualized runtime reduction engine that compiles code to a machine executable platform specific form (CLI).
I agree with you. Those packages to develop good versions can involve hundreds of man years of work. The Linux kernel, Firefox, Open Office... there are only a few products with anything like 100 man years of of intensity / funding. Conversely in the server area a) products were less mature. b) where they weren't less mature they were so expensive
So free software could get a foothold more easily. I think (b) is the approach for the low end of the market. The people who buy $350 computers don't want to buy a $300-2k software package. There is money to be made.
Legislation should exist to benefit society, not to maximize profits for a select few corporate entities.
The problem here is not the legislation so much as the regulation. Different parts of the American economy need different patent policies and quite a bit of judgement needs to be applied to achieve that. That is we need an expensive, well staffed powerful patent office with strong congressional support.
if emulation has never been popular, why has Nintendo introduced Virtual Console to compete with PC-based emulators that run infringing ROM images of games for discontinued video gaming platforms?
Because it is easy money for not much work. There is a huge difference between emulators as secondary support, like Parallels or VMWare on OSX to run a few Windows apps; and emulators as primary support to be in regular use.
There are two issues here: 1) Is.NET a terrific compiler: i.e. feature rich, fast,.... 2) Is.NET compatible with other compilers.
Arguably (1) and (2) are opposites..NET is more advanced than the standards. Type safety is a good example of this. It is a feature of the compiler. By making it a requirement you are absolutely correct that makes porting from a low level language complex, maybe even a full on rewrite. But I don't see how that's a failure of.NET.
It seems to me with neither Microsoft, nor Apple interested in focusing on cross platform, mobile developers are going to end up supporting different code bases. And if Android vs. Apple are any indication the customer basis of these different platforms are likely to fragment in terms of taste so it might not be so bad. On the desktop Apple customers want different stuff than Windows customers, and it appears that on the phone that is true as well. So I'm not sure that the tradition of mobile developers deploying the same app to multiple devices isn't really an artifact of the JavaVM days where phones weren't supposed to have rich eco systems.
That's because browser developers have deliberately crippled the stack for the web, offering no API to gain the user's consent for camera and phone access or reading the user's contacts. This pushes developers to things like PhoneGap.
I gotcha. Back in the IE 4 days Active-X was amazing in terms of what you could do. A secure version of Active-X would be incredible.
Slashdot Mirror
How is there any more lockin on Metro than on GDI/.NET? Metro if anything seems somewhat more open.
Lots of medium companies buy Microsoft software indirectly to leverage up or get consolidated pricing. Large all have relationship with Microsoft but the "sales weasel" might have the relationship with the guys in SQL Server and Dynamics and have no relationship at all with the desktop team.
They can't do that. The heavy duty applications don't exist for Android and won't be ported to Android anytime soon. Dell's entire innovation is customization at a low cost. They can't beat Samsung, LG, Motorolla at making 100m of one model. Android is the worst of all words for Dell. Same with Lenovo. Acer I don't know enough about.
The IBM PC was the one strange thing in that you could install any OS on it.
No it was just as locked down. Microsoft, Intel and Western Digital created the standard that allowed IBM compatibles to get compatible enough that this became a multi vendor platform with Intel and Microsoft in the driver's seat.
Non-Apple consumer x86 sales are down 10% year over year. That tablet and cell phone replacement is already happening in consumer.
It was great when there were generous margins and rapidly growing sales.
It was good when margins tightened but sales still surged.
It was getting kinda sucky when margin collapsed but at least sales kept surging.
It got worse when sales flattened out.
Now we are facing the real possibility of sales collapsing.
Microsoft is doing what they can to fend that off. If they are going to fend that off they need to change the platform ecosystem: OS, hardware, development libraries, software... and they need to do this fast. They need to do this in a platform that moribund. Hopefully confronted with Windows 8, consumers start buying the devices they will work on. Dell, Asus, HP... now how to make those devices they just don't know how to make them for $500.
The moment this happens we'll know that Microsoft has failed and not even they can escape the black hole that is their own legacy.
I agree with you. Microsoft is going to have to take a page from Apple's book in killing off legacy cruft quickly but with lots of steps to make it look gradual so people don't freak.
Gamers used to be very willing to boot to a gaming OS. It was standard to have a "gaming" configuration of DOS and a "windows" configuration in the Windows 3.0/3.1/95 days. I see no reason if Linux had strong gaming support that their wouldn't be gamer machines sold with Linuxes tweaked to optimize gaming. That honestly is a pretty good niche for Linux, a place where custom kernel could really matter. And heck the hardware OEMs would probably love a product where they could do a real value ad and make some margin.
I have no idea how big the gaming market is, but it seems to be carrying several midsized computer companies. So why not?
I love the DOS 4.0 reference! Yes that was a lousy OS but you can blame IBM for that one, PCDOS 4.0 was what sucked Microsoft was just following their lead on the DOS front for the last time.
I think Windows 8 might work out. I see Windows 8 as potentially being a lot like IE 2. By IE 2 Microsoft had realized this internet thing was real and a real threat, and was aggressively pulling technologies into their browser. IE 2 was a a real browser while IE1 had been a research project. People would still buy Netscape in the IE 2 days but within 2 years after IE 2 they'd be up to IE 4, which was way more advanced than anything Netscape could do and frankly had a lot of really cool features that most browsers even today don't.
Now while will open systems cost as much as a car in this new world?
As for DRM I'm not sure that Apple isn't happy to play DRM either way. When media companies don't play along they side with consumers. When media companies do play along they side with publishers. But if you look at the history of iTunes, every song sold now is DRM free. We have today the market people always asked for, individual songs sold at a reasonable cost DRM free from a variety of vendors in open formats. Things don't always turn out so bad.
I'm not a photo guy but my understanding is that when Aperture came out it was designed to compete with Photoshop, there was no Lightroom. Other than HDR what doesn't Aperture do? I'd assume Photoshop has more features but I'm looking for big stuff. Aperture seems much more full featured than Lightroom.
I own Pixelmator. I'm not sure I'd compare to it Photoshop.
That is pretty funny. And the sort of thing that infuriates customers.
What is the difference between a VM and an emulator?
The intent of the software running on it.
Program A is designed to run with hardware B.
Program C is designed to run against VM D.
On B I can run A or C.
On some other hardware I can run VM D, but when I try and A I have to emulate.
Abstractly an emulator is just a VM. In practice an emulator is a large, complex, inconsistent VM that alters the behavior of code in unpredictable ways and thwarts far too many of the code's original designers. As the saying goes: in theory there is no difference between theory and practice, in practice there is.
Not anymore. Modern web browsers JIT compile it like Java nowadays
I doubt that. Most likely they JIT it more like Perl, where code can (indirectly) call the evaluator on data. Dynamic languages are much, much harder to write compilers for. I doubt anyone has reduces Javascript, which has full blown lambdas to a compile.
Safari since iOS 4.3 or so even uses a special permission not available to third-party applications to circumvent the platform's compulsory DEP so that it can JIT compile JavaScript code.
I'm aware of that, two versions of webkit when it comes to Javascript. I still use Dolphin as my browser even knowing that though.
There is no way to buy OSX legally. Apple doesn't sell OS X. You can only buy:
a) A computer which includes installation media (or now virtual installation media)
b) An upgrade.
So? Did I say the iPhone was a clone of Palms? Apple copied liberally from all major vendors: Nokia, Windows Mobile, and Palm
Let me just stop you here. If Apple produced a unique combination of existing technologies in a way that would work better, that's insight and... that's patentable. I don't have to have invented either the transistor or the radio to be able to patent the transistor radio. The only question would be whether everyone in the radio would have obviously seen the applicability of the transistor of if that applicability was a matter of unique insight.
You can't have it both ways here. Either they copied from Palm or they did something new.
I just don't think it has practical importance in most cases, since there are so many ways to hack into a computer once the OS is up and running (and unlike installing a rootkit at boot, many of them can be executed remotely).
Trusted computing also gets rid of most of those problems too. UEFI doesn't do anything there, but the technology needed to take UEFI to the next level would help quite a bit with hacking exploits. Mainframes don't get hacked nearly as easily because the security model is very different.
I don't doubt the Linux developers will find a way to get around the technical problems. But Redhat apparently thought something as simple as disabling SecureBoot manually, would be enough to scare away some of their customers.
And I think RedHat's right. There is a difference between:
a) Will scare some customers.
b) Will make something completely impossible.
(a) and (b) aren't the same thing.
Well... isn't that easier than getting around hardware locks? Content can only be protected by encrypting it, and if legitimate users are to be able to access it, the decryption keys need to be stored in cleartext on their computers.
that's now how it is done. Here is an example. A long message is broken into a string of smaller messages (like like 1mb each). Each small message gets an AES encryption with the AES key stored RSA encrypted along with the block. The RSA private key is on the TCPA chip which can decrypt but will only unlock that set of keys based on the checksum of the kernel performed by the UEFI. Outside the TCPA chip the RSA private key is unknown. And the TCPA chip will only perform decryptions to an authorized kernel. The AES keys are known to the system but they are one time use.
Note that if I release the document to say 30 people, those 30 people correspond to me signing each small message with 30 different AES keys. Or one time a group key can be loaded on the TCPA chip and everyone can share a group key. Private keys are never stored on the computer because the computer as a whole cannot be trusted.
For example, when Microsoft saw Netscape as a threat, they changed the licensing terms for Windows NT Workstation to make it impossible to run web servers on it. You could still run web servers on Windows NT Server, but when you bought NT Server you got a web server for free, so you didn't need Netscape's. Nobody was actually prevented from running Netscape's web server; Microsoft just made sure it didn't make financial sense for most users. (This was one of the practices the DoJ found to be anti-competitive.)
I used to run Netscape's webserver. I have to tell you IIS was a lot better. It wasn't licensing that got me to switch but Active-X, the speed of configuration, the ease of management.... But I do get your point that's a good example. And I agree that Microsoft will nudge not compel. And that was my point, a total lockout is compel.
In reality if a program crashed the Amiga CPU simply jumped to the next program because the CPU maintained control of which task executed.
The Amiga CPU, the 68000, had no idea what a task was or what a program was. It doesn't maintain control of anything. That's the point there is no hardware here that can act independently of the current task. There was no protected memory. If Eudora went down lightly things might be salvageable, for example a scheduled interrupt. But if Eudora has a bad pointer it is corrupting all of memory including the kernel or even the interrupts table.
Just think for a second about what you are claiming here. Figure out how this could work, in detail.
. it's very easy to build a competing product, for say a HTTP server.
I'm putting that down to complexity. An HTTP server extension is usually under a few man years of code. Its easier than an office suite just because it is less complex not because of open standards. I don't think it is any easier to write extensions for Apache than for IIS.
If the real world would be like the end-user software world, you would have 5 different sockets for lamps, 10 different power plugs. And each TV would come with a custom proprietary power and cable plug.
That's what electrical devices were like in the late 19th century. Also if there were a standard 10 years it would have been the Microsoft standard. Microsoft was the defacto standard. The free software community rejected the Microsoft standard and successfully shattered it first via LAMP and then via. Firefox.
The issue is that computers grew like almost no other new technology and with little government help; and government are frequently the ones that impose and enforce standards.Partially that was because of the American mood, we had a new conservatism which created amounts of private investment capital while stripping public investment capital. Partially it was because the productivity savings from computers were so massive and so obvious government investment was not required. But regardless there is a lack of enforced standards.
Nobody would fund you if you start a competing application for, for example, MS Office or Photoshop, with the goal to compete open in the market.
I don't know about that. Apple in 2004 funded a competitor to PhotoShop when Adobe was threatening to drop the platform. And they were successful. Sun did fund a competitor to MS Office for years and took StarOffice from a kinda crudy office suite to something in the same ballpark as MS Office. And let me point out, AOL funded a competitor to IE for 6 years. The last 2 were both open source efforts.
OO.org, LibreOffice, GIMP goal is not to compete commercially, so they survive as a not-for-profit organization.
Actually the goal of OO.org / Star Office was to compete commercially. Sun wanted to push something like the cloud SaaS model ten years earlier. They absolutely wanted to compete with Microsoft commercially. If you mean selling boxes of software Office isn't really that either, though it is a huge money maker it is also the lynchpin of Microsoft's entire server based suite and their OS monopoly.
Because nobody can compete with Adobe in the marketplace.
Of course they can. Heck Adobe just lost to Javascript. Adobe's rip technology lost to HP's (essentially free) PCL. Adobe Premiere got its but kicked by Apple with iMovie and Final Cut Pro/X. GIMP can't beat Adobe doesn't mean no one can.
No I actually meant a virtual PC included in Microsoft office. There are still some links to version of the product on the web for example: http://www.microsoft.com/australia/office/mac/virtualpc7/default.aspx So for example Microsoft Virtual PC 6 came with Office 2004 for Mac (this was after Microsoft bought Connectix).
Anyway I've never heard people consider the experience of a PC via. Parallels to be near native. I'm not sure what that is owing to, but it doesn't seem in practice to be much different than using VirtualPC.
email accounts, often act as a proxy for a member identifier / account identifier. They aren't perfectly unique in either direction. Sometimes multiple people share an
email but then they are sharing an account; sometimes the same person has multiple emails but then effectively that person is acting like multiple people.
For most companies the majority of their middleware are desktop productivity applications like Access combined with a semi skilled office worker. A file gets pulled from one server, manipulated by hand, and then sent to another server.
The failure is that .NET on iOS or Android costs a developer hundreds of dollars per year payable to Xamarin
Hold on a second. .NET is a platform specific compiler, not a platform independent one. You shouldn't be using .NET for platform independent work, no one has ever said it was cross platform not intended to be. Nor has Microsoft been a fan of cross platform development, in general. They themselves fork their own code base when they support multiple platforms rather than use conversion tools. What you are saying is like saying I have a bad air conditioner because it doesn't sweep the driveway.
The job of a compiler is to take code written in a higher level language and reduce it to hardware / platform executable code. The .NET compiler does a fantastic job on the criteria by which compilers are evaluated, things like optimization efficiency. C# is allowing for some interesting ideas from functional programming to be imported into high speed structural languages without loss of efficiency.
You have an objective that Microsoft doesn't agree with, no support lots of platforms without rewriting your code. .NET isn't bad because it doesn't help you do stuff that it was never meant to help you do.
That's what Google Native Client is supposed to be.
Thanks! Don't know how I missed that technology when it was announced. Yes that is exactly the kind of thing I was thinking of.
Java is a VM. Javascript is an interpreted language. .NET is virtualized runtime reduction engine that compiles code to a machine executable platform specific form (CLI).
None of those are emulators.
I agree with you. Those packages to develop good versions can involve hundreds of man years of work. The Linux kernel, Firefox, Open Office... there are only a few products with anything like 100 man years of of intensity / funding. Conversely in the server area
a) products were less mature.
b) where they weren't less mature they were so expensive
So free software could get a foothold more easily. I think (b) is the approach for the low end of the market. The people who buy $350 computers don't want to buy a $300-2k software package. There is money to be made.
Legislation should exist to benefit society, not to maximize profits for a select few corporate entities.
The problem here is not the legislation so much as the regulation. Different parts of the American economy need different patent policies and quite a bit of judgement needs to be applied to achieve that. That is we need an expensive, well staffed powerful patent office with strong congressional support.
if emulation has never been popular, why has Nintendo introduced Virtual Console to compete with PC-based emulators that run infringing ROM images of games for discontinued video gaming platforms?
Because it is easy money for not much work. There is a huge difference between emulators as secondary support, like Parallels or VMWare on OSX to run a few Windows apps; and emulators as primary support to be in regular use.
and now it looks like version 8 will break compatibility yet again.
Its been officially announced and repeated by both Nokia and Microsoft. I think we can just say it as fact, no need to hedge with "looks".
There are two issues here: .NET a terrific compiler: i.e. feature rich, fast, .... .NET compatible with other compilers.
1) Is
2) Is
Arguably (1) and (2) are opposites. .NET is more advanced than the standards. Type safety is a good example of this. It is a feature of the compiler. By making it a requirement you are absolutely correct that makes porting from a low level language complex, maybe even a full on rewrite. But I don't see how that's a failure of .NET.
It seems to me with neither Microsoft, nor Apple interested in focusing on cross platform, mobile developers are going to end up supporting different code bases. And if Android vs. Apple are any indication the customer basis of these different platforms are likely to fragment in terms of taste so it might not be so bad. On the desktop Apple customers want different stuff than Windows customers, and it appears that on the phone that is true as well. So I'm not sure that the tradition of mobile developers deploying the same app to multiple devices isn't really an artifact of the JavaVM days where phones weren't supposed to have rich eco systems.
That's because browser developers have deliberately crippled the stack for the web, offering no API to gain the user's consent for camera and phone access or reading the user's contacts. This pushes developers to things like PhoneGap.
I gotcha. Back in the IE 4 days Active-X was amazing in terms of what you could do. A secure version of Active-X would be incredible.