The problem with that is that OSX also provides a start-menu of sorts, it's just got nowhere near as much power.
What are you talking about? The OSX folders on doc is at least as powerful as the start menu. I can can create either a physical or smart folder (the result of query). Inside that folder I can have applications shortcuts, actual executables (applescript, shell script...), other actual folders or smart folders. The system will make reasonable choices about icons based on the contents. And I have 4 browsing schemes I can use with any folder or smart folder anywhere in the hierarchy.
So I so / so 3 1/2 year old computer. In terms of browsers: Safari, Firefox, Chrome are the man ones. All of them get pushed too far and I have to start closing tabs, and sometimes reboot the browser. There is no problem with my computer, I have the same problems on other computers. The web is taxing.
In fact my web based problems are one of the reasons I'm upgrading to the new retina (400m/sec ssd, 16g, quad core). It should buy me at least a few good years of being able to push the browser without taxing the system.
As far as I can tell my system does pretty much fetch all the comments at once on/. That's not the issue.
The issue is doing ad hoc data analysis on the client with those comments. Like "show comments by x in this thread" or "show last ten comments by x" (has to go back to server) or "show my previous responses" (has to run a mini version of the server side app).
Is it doable in HTML5. Yeah probably in theory Javascript is a very rich language, but I don't see it. And I don't see it because things that are basic in desktop apps are complex are often quite complex in web based applications to implement easily. I can certainly use the.sort method in javascript for sorts, but I'd have to have fairly complex data structures and output routines.... Web apps, with some exceptions like google docs, mostly are very low feature.
Evernote for example is fundamentally a web app, yet they keep implementing desktop and browser based clients to be able to access system state and write permanent records.
Well yes if you are going to assume the client doesn't need performance then Metro doesn't give you any advantages over web.
But a 300x multiplier is a big deal. A very big deal. I was giving you an example of where/. doesn't do something it should for performance reasons. Most web apps are very low feature compared to their desktop counterparts because of this problem. This BTW is precisely how Microsoft saved Java. Under Netscape Java was dreadful. When I.E. came out with an optimized Java that included lower level calls, Java sped up and became usable. Sun freaked out of course because this broke the whole "write once, run anywhere" paradigm but I think without Microsoft's speed enhancements Java applets never catch on and Java never ends up becoming the web applications language.
Microsoft still sells a license if people are using VMs. Possibly more than one per user. Further its unclear to me how running one or more Windows VMs under Linux is going to be less confusing than running one VM under Windows 8. I don't see how VMs matter.
I agree that if most applications were running in the VM and not native then yes, it does allow you to run anything on the client and windows VMs over it. The same way that when Windows moved in lots of businesses used terminal emulators to essentially VM their old terminal systems clients.
Window 8 isn't released yet. And no there isn't a guarantee but even with Vista there was rather rapid adoption well beyond the user base needed in the iPhone analogy.
As for business and the sarcasm.... there will likely be far more business users than there were iPhone users 2 years ago within a year of release even if Windows 8 bombs. You can't claim the iPhone population is large enough and at the same time the Windows 8 population won't be.
I'm already getting the entire thread. They are already sending all the data it to me. There is no reason with a thicker client I couldn't be doing computations on that data. Or even just passing the results of those computations along and allowing for selecting / sorting / searching operations.
The easiest thing is to only semi pre chew the data and let the client do most of the work. But that requires a more complex client.
Most businesses today have so much and such deep Microsoft integration that even if they wanted to getting off would be far to expensive. Just look at the low success ratios of the Unix switching movement from a decade ago: IBM, Oracle, Sun all failed. The only people who were successful were the companies that never developed a windows culture in the first place.
Sure pissed off corporate customers could get mildly more Mac friendly. They could move off of exchange. They could start having open office also be installed on their desktop image. And they could have a Linux pilot project. But no they can't quit. Lockin is real.
The user impression -correct or not- is that stuff changed in Office just for kicks. And if they got their hands on 8, where stuff is "just changed" we would expect the same complaints.
They do expect the same complaints. They know its coming. On the other hand, objectively the ribbon interface allows for far more controls and as people get used to ribbon they can start offering context sensitive to a much greater degree and move towards ribbon menus with tens of thousands of controls for all sorts of specialized work.
In the same way, Metro if it is successful, will allow for a level of ubiquitous complex applications that will surpass anything their competitors can offer. A full generation of Windows as the default OS for the planet. And that is worth some complaints.
No it wouldn't be huge. It would become a standard way of people operating. It means that applications vendors wouldn't change to offering touch friendly installation. It would undermine the whole point of the hard line Microsoft is taking.
What is behind it is you aren't taking phones seriously enough. A high end smartphone today is roughly comparable with a desktop of a decade ago in terms of CPU, RAM and HD space. Currently about 1/4b sell per year, heading towards 1/3b, much faster than desktops. Because phones have margins for parts manufacturers and that market is adopting aggressively, the distance is likely to decrease. So by 2022 they might only be 5 years behind.
Phone OSes have already moved upmarket to tablets which while still only doing tens of millions of units should be at 100m by around 2015 and growing rapidly. Tablets unlike phones demand lots of good software so they are already creating a vibrant and growing software market which is innovative in a way that desktops have not been.
Microsoft has already lost the over $1k laptop market to OSX. They could potentially be wiped out of consumer / small business by the end of the decade. Which would put them in the same position Unisys, IBM, DEC... were when they replaced those guys in the enterprise. Microsoft does not want history to repeat.
If they get blown out of consumer and become an enterprise desktop solution then defending their more limited territory becomes easier. They move towards much better integrations with their server based products: dynamics, exchange, universal communicator and the OS is loaded to the gills with medium+ sized business friendly features. Apple and Google own the Home/SB market and neither has any real intention of moving up market.
Innovation though is happening at the consumer market because of size and slowly what they did to IBM/DEC/Unisys... is done to them.
I think Microsoft is about to be surprised by how many are similar in this regard.
No they won't. Microsoft fully expects this. They feel they don't have a choice with this one. They understand full well business like yours will be on Windows 7 for quite a while. They just are doing this work so that in 2015-2020 when you move off Windows 7 it is to another Windows OS and not iOS/Android/OSX....
Personally, I don't really see any obvious cases where a Metro app would give any considerable advantage over a web app in a business environment
Anytime you want the person to have a complex workflow on the screen. Native code is around 300x faster which allows for a richer GUI. You are on/.. How much better an experience could these dialogues be if/. were able to process complex commands like compute in the background all shutdown -p now's comments and show them to me when I read a comment. so I don't end up responding and then 1 minute later respond again to the next comment in the thread with a more detailed version.
What is the percentage in coding to a single, specialized environment when everyone else in the world is coding using mature cross-platform web-based solutions.
Except they aren't. See the Apple Store for examples of recent coding to single specialized environments. And not only that in a C based language which more or less quintuples developer effort.
Microsoft's whole point here is not to be careful. They want to shake up the windows market. They understand their user base is conservative and that's a serious problem to their long term prospects.
Is that a profit problem for MS? How? They're collecting license fees on every new machine.
Yes they are but upgrade cycles are getting longer. More importantly in terms of eco system though the hardware partners aren't making much money and so aren't doing R&D. They also aren't investing in the lastest hardware. Meanwhile a quarter billion smart phones many with high profit margins and all of them with innovative hardware sell each year. An entire tablet eco system is forming and getting into the enterprise. Apple owns 90% of the over $1k laptop market.
Yes, it is a problem a big problem. They are facing exactly the sorts of pressures that IBM, DEC, Unisys... were facing in the late 80s on that led to Windows PCs coming in at the department level. They are starting to see low level non Windows support (exchange support for iPhones) just like good terminal emulators and Windows for Workgroups started to allow Windows machines to be the standard.
They could easily be blown out of consumer over the next 10 years and be in a position in enterprise where people only want to purchase Windows licenses to run on a VM on say Android computers.
Microsoft is at this point the most conservative major OS vendor.
Apple has a 4 year support window more or less. Linux desktops are deeply fragmented and based on radical choices like KDE vs. Gnome vs. XFCE vs. LXDE vs. window manager only Android has a short support window and is fragmented.
Moving from Microsoft for stability is jumping from the frying pan to the fire.
Windows 8 = beginning of the end of the Wintel ecosystem
We are at most a few years away from a world of 300m smart phones sold per year. Apple already owns 90% of the over $1k laptop market. And the tablets are coming on strong all non Windows. the Wintel ecosystem is in serious trouble, that's why Balmer is doing Windows 8. He's taking aggressive moves while there still is something to defend.
Microsoft gets it fine. They understand that's their customer's attitude in enterprise. They just have to navigate that attitude while trying to use their existing leverage to adopt to the next generation of technology. They don't want Android/iOS to do to them what they did to IBM/Unisys/DEC...
I know you are being sarcastic, but having used Database filesystems on things like OS/400 and OpenVMS... they are fantastic. It would be a huge improvement. I wish Microsoft hadn't chickened out on that one.
I have to tell you I run all kinds of very hardcore applications on my laptops like database servers (development) and large compiles.
One of the most stressing applications I run is web browsers.
Video: tens of hundreds of megabytes of data that needs real time performance to resize images multiple times per second running through a virtual machine in an application that runs against the OS
Webpages with ads: often megabytes of poorly written inefficient interpreted code that needs to finish executing in less than a second
I just don't see today's computers as remotely fast enough to browse the web.
You have to remember that in the 1990s Microsoft was a rogue system coming in from below. Departments were introducing Windows for Workgroups and Windows 95 to run applications that they couldn't get support for on their Minis and Mainframes. Departments saw no reason to pay $100k for an application from Unisys when they could get 80% of the functionality they cared about $299 with a DOS/Windows solution. Windows was a consumer OS there weren't the long support contracts nor any attempt to get it compatible with changing hardware standards.
Once Microsoft had the opportunity to replace the enterprise desktop and not just be a secondary system they took it. And that meant:
-- management features locking those very early adopters down -- long support cycles etc..
And that introduced an exciting time in technology where enterprise ITs were getting features they cared deeply about in the new Windows versions. Windows NT 4.0 was a huge upgrade for them from Windows for Workgroups. 2000 smoothed out Windows NT and there was no legacy to support. This led to a major change, Windows owned the desktop and it wasn't just Windows machines running small business / consumer software and terminal emulators for the real stuff but rather they had native applications. And suddenly enterprise IT had real expenses to any kind of change.
Slashdot Mirror
When we had automatic transmission and "park" became a transmission setting distinct from neutral.
The problem with that is that OSX also provides a start-menu of sorts, it's just got nowhere near as much power.
What are you talking about? The OSX folders on doc is at least as powerful as the start menu. I can can create either a physical or smart folder (the result of query). Inside that folder I can have applications shortcuts, actual executables (applescript, shell script...), other actual folders or smart folders. The system will make reasonable choices about icons based on the contents. And I have 4 browsing schemes I can use with any folder or smart folder anywhere in the hierarchy.
How is that less powerful?
The coolness of HTML5 video (h.264), was one of the many reasons. And that works great on the iPhone.
CPU = Dual core 2.4ghz 3m of L3.
RAM = 4g
So I so / so 3 1/2 year old computer. In terms of browsers: Safari, Firefox, Chrome are the man ones. All of them get pushed too far and I have to start closing tabs, and sometimes reboot the browser. There is no problem with my computer, I have the same problems on other computers. The web is taxing.
In fact my web based problems are one of the reasons I'm upgrading to the new retina (400m/sec ssd, 16g, quad core). It should buy me at least a few good years of being able to push the browser without taxing the system.
As far as I can tell my system does pretty much fetch all the comments at once on /. That's not the issue.
The issue is doing ad hoc data analysis on the client with those comments. Like "show comments by x in this thread" or "show last ten comments by x" (has to go back to server) or "show my previous responses" (has to run a mini version of the server side app).
Is it doable in HTML5. Yeah probably in theory Javascript is a very rich language, but I don't see it. And I don't see it because things that are basic in desktop apps are complex are often quite complex in web based applications to implement easily. I can certainly use the .sort method in javascript for sorts, but I'd have to have fairly complex data structures and output routines.... Web apps, with some exceptions like google docs, mostly are very low feature.
Evernote for example is fundamentally a web app, yet they keep implementing desktop and browser based clients to be able to access system state and write permanent records.
Well yes if you are going to assume the client doesn't need performance then Metro doesn't give you any advantages over web.
But a 300x multiplier is a big deal. A very big deal. I was giving you an example of where /. doesn't do something it should for performance reasons. Most web apps are very low feature compared to their desktop counterparts because of this problem. This BTW is precisely how Microsoft saved Java. Under Netscape Java was dreadful. When I.E. came out with an optimized Java that included lower level calls, Java sped up and became usable. Sun freaked out of course because this broke the whole "write once, run anywhere" paradigm but I think without Microsoft's speed enhancements Java applets never catch on and Java never ends up becoming the web applications language.
Microsoft still sells a license if people are using VMs. Possibly more than one per user. Further its unclear to me how running one or more Windows VMs under Linux is going to be less confusing than running one VM under Windows 8. I don't see how VMs matter.
I agree that if most applications were running in the VM and not native then yes, it does allow you to run anything on the client and windows VMs over it. The same way that when Windows moved in lots of businesses used terminal emulators to essentially VM their old terminal systems clients.
Window 8 isn't released yet. And no there isn't a guarantee but even with Vista there was rather rapid adoption well beyond the user base needed in the iPhone analogy.
As for business and the sarcasm.... there will likely be far more business users than there were iPhone users 2 years ago within a year of release even if Windows 8 bombs. You can't claim the iPhone population is large enough and at the same time the Windows 8 population won't be.
I'm already getting the entire thread. They are already sending all the data it to me. There is no reason with a thicker client I couldn't be doing computations on that data. Or even just passing the results of those computations along and allowing for selecting / sorting / searching operations.
The easiest thing is to only semi pre chew the data and let the client do most of the work. But that requires a more complex client.
Most businesses today have so much and such deep Microsoft integration that even if they wanted to getting off would be far to expensive. Just look at the low success ratios of the Unix switching movement from a decade ago: IBM, Oracle, Sun all failed. The only people who were successful were the companies that never developed a windows culture in the first place.
Sure pissed off corporate customers could get mildly more Mac friendly. They could move off of exchange. They could start having open office also be installed on their desktop image. And they could have a Linux pilot project. But no they can't quit. Lockin is real.
Just using the internet:
There are 2.2b desktop internet users about 90% windows
There are 400m smart phone users about 1/3rd iPhone.
10% of windows users exceeds iPhone. Microsoft will get past whatever you think the cutoff is quickly.
The user impression -correct or not- is that stuff changed in Office just for kicks. And if they got their hands on 8, where stuff is "just changed" we would expect the same complaints.
They do expect the same complaints. They know its coming. On the other hand, objectively the ribbon interface allows for far more controls and as people get used to ribbon they can start offering context sensitive to a much greater degree and move towards ribbon menus with tens of thousands of controls for all sorts of specialized work.
In the same way, Metro if it is successful, will allow for a level of ubiquitous complex applications that will surpass anything their competitors can offer. A full generation of Windows as the default OS for the planet. And that is worth some complaints.
No it wouldn't be huge. It would become a standard way of people operating. It means that applications vendors wouldn't change to offering touch friendly installation. It would undermine the whole point of the hard line Microsoft is taking.
They don't want the current paradigms.
What is behind it is you aren't taking phones seriously enough. A high end smartphone today is roughly comparable with a desktop of a decade ago in terms of CPU, RAM and HD space. Currently about 1/4b sell per year, heading towards 1/3b, much faster than desktops. Because phones have margins for parts manufacturers and that market is adopting aggressively, the distance is likely to decrease. So by 2022 they might only be 5 years behind.
Phone OSes have already moved upmarket to tablets which while still only doing tens of millions of units should be at 100m by around 2015 and growing rapidly. Tablets unlike phones demand lots of good software so they are already creating a vibrant and growing software market which is innovative in a way that desktops have not been.
Microsoft has already lost the over $1k laptop market to OSX. They could potentially be wiped out of consumer / small business by the end of the decade. Which would put them in the same position Unisys, IBM, DEC... were when they replaced those guys in the enterprise. Microsoft does not want history to repeat.
If they get blown out of consumer and become an enterprise desktop solution then defending their more limited territory becomes easier. They move towards much better integrations with their server based products: dynamics, exchange, universal communicator and the OS is loaded to the gills with medium+ sized business friendly features. Apple and Google own the Home/SB market and neither has any real intention of moving up market.
Innovation though is happening at the consumer market because of size and slowly what they did to IBM/DEC/Unisys... is done to them.
I think Microsoft is about to be surprised by how many are similar in this regard.
No they won't. Microsoft fully expects this. They feel they don't have a choice with this one. They understand full well business like yours will be on Windows 7 for quite a while. They just are doing this work so that in 2015-2020 when you move off Windows 7 it is to another Windows OS and not iOS/Android/OSX....
Personally, I don't really see any obvious cases where a Metro app would give any considerable advantage over a web app in a business environment
Anytime you want the person to have a complex workflow on the screen. Native code is around 300x faster which allows for a richer GUI. You are on /.. How much better an experience could these dialogues be if /. were able to process complex commands like compute in the background all shutdown -p now's comments and show them to me when I read a comment. so I don't end up responding and then 1 minute later respond again to the next comment in the thread with a more detailed version.
What is the percentage in coding to a single, specialized environment when everyone else in the world is coding using mature cross-platform web-based solutions.
Except they aren't. See the Apple Store for examples of recent coding to single specialized environments. And not only that in a C based language which more or less quintuples developer effort.
Microsoft's whole point here is not to be careful. They want to shake up the windows market. They understand their user base is conservative and that's a serious problem to their long term prospects.
Is that a profit problem for MS? How? They're collecting license fees on every new machine.
Yes they are but upgrade cycles are getting longer. More importantly in terms of eco system though the hardware partners aren't making much money and so aren't doing R&D. They also aren't investing in the lastest hardware. Meanwhile a quarter billion smart phones many with high profit margins and all of them with innovative hardware sell each year. An entire tablet eco system is forming and getting into the enterprise. Apple owns 90% of the over $1k laptop market.
Yes, it is a problem a big problem. They are facing exactly the sorts of pressures that IBM, DEC, Unisys... were facing in the late 80s on that led to Windows PCs coming in at the department level. They are starting to see low level non Windows support (exchange support for iPhones) just like good terminal emulators and Windows for Workgroups started to allow Windows machines to be the standard.
They could easily be blown out of consumer over the next 10 years and be in a position in enterprise where people only want to purchase Windows licenses to run on a VM on say Android computers.
Microsoft is at this point the most conservative major OS vendor.
Apple has a 4 year support window more or less.
Linux desktops are deeply fragmented and based on radical choices like KDE vs. Gnome vs. XFCE vs. LXDE vs. window manager only
Android has a short support window and is fragmented.
Moving from Microsoft for stability is jumping from the frying pan to the fire.
Windows 8 = beginning of the end of the Wintel ecosystem
We are at most a few years away from a world of 300m smart phones sold per year. Apple already owns 90% of the over $1k laptop market. And the tablets are coming on strong all non Windows. the Wintel ecosystem is in serious trouble, that's why Balmer is doing Windows 8. He's taking aggressive moves while there still is something to defend.
Microsoft gets it fine. They understand that's their customer's attitude in enterprise. They just have to navigate that attitude while trying to use their existing leverage to adopt to the next generation of technology. They don't want Android/iOS to do to them what they did to IBM/Unisys/DEC...
I know you are being sarcastic, but having used Database filesystems on things like OS/400 and OpenVMS... they are fantastic. It would be a huge improvement. I wish Microsoft hadn't chickened out on that one.
I have to tell you I run all kinds of very hardcore applications on my laptops like database servers (development) and large compiles.
One of the most stressing applications I run is web browsers.
Video: tens of hundreds of megabytes of data that needs real time performance to resize images multiple times per second running through a virtual machine in an application that runs against the OS
Webpages with ads: often megabytes of poorly written inefficient interpreted code that needs to finish executing in less than a second
I just don't see today's computers as remotely fast enough to browse the web.
You have to remember that in the 1990s Microsoft was a rogue system coming in from below. Departments were introducing Windows for Workgroups and Windows 95 to run applications that they couldn't get support for on their Minis and Mainframes. Departments saw no reason to pay $100k for an application from Unisys when they could get 80% of the functionality they cared about $299 with a DOS/Windows solution. Windows was a consumer OS there weren't the long support contracts nor any attempt to get it compatible with changing hardware standards.
Once Microsoft had the opportunity to replace the enterprise desktop and not just be a secondary system they took it. And that meant:
-- management features locking those very early adopters down
-- long support cycles
etc..
And that introduced an exciting time in technology where enterprise ITs were getting features they cared deeply about in the new Windows versions. Windows NT 4.0 was a huge upgrade for them from Windows for Workgroups. 2000 smoothed out Windows NT and there was no legacy to support. This led to a major change, Windows owned the desktop and it wasn't just Windows machines running small business / consumer software and terminal emulators for the real stuff but rather they had native applications. And suddenly enterprise IT had real expenses to any kind of change.
That's what happened.