They didn't get into enterprise server rooms very much. They had niches:
1) People who wanted the g4/g5 for supercomputing applications. Xserve were much cheaper than IBM p-series. Those people bought tons of XServes. When they switched to Intel though...
2) Small business or schools with a small number of racks.
3) Rendering farms for certain specialized applications that had crossed over.
When Apple got in the server room it was with the G4/G5. XServers were way way cheaper than IBM p-series machines. Once they switched to Intel it is hard to see what specific advantages Apple has.
IT is a market that Apple is doing well in. Huge numbers of programmers use Apple and have since the 10.2 days.
Now if you mean enterprise... Apple has little interest in meeting the criteria for enterprise. They might make some minor concessions but really to satisfy enterprise requires a separate product line and then that product line drives your business. All their competitors are focused primarily on enterprise.
The pro multimedia market isn't a safety net for a company the size of Apple. Especially with Apple's margins. If Apple goes out of style they find new markets or they end up in dozens of niches not just 10- and have to rethink the business entirely to fit those niches.
Well written post. It is absolutely true that Apple doesn't allow for layering on their advanced stuff. Being a developer for Apple requires constant upgrading and for a chain of products that is impossible.
Ironically since Chen left Microsoft they've gotten less interested in legacy support as well. They are becoming more like Apple (though the difference is still fast). No question Windows is going to become the main platform for that sort of layering but it might end up being impossible in Windows as well. Interesting possible niche for Linux which is excellent on the complex layering and about maintaining old APIs.
OSX supports keyboards with more keys, that's a hardware choice.
As for the only menu being on top.... Yeah that's a switch. The advantage is you can see what application is tied to your window manager as active with a glance. As an aside in Display item of system preference you can change behavior.
They took it out. The same way they dropped classic after a few years. It is part of apple culture to make developers keep up and not have 20 year old software running around.
Programmers and SE's? Apple early on pretty much stole the Linux development crowd around the OSX 10.2 days. Seems to me DarwinPorts, Fink, Terminal and Quartz-wm still exist and offer a very good environment. As far as OSX development, Developer tools are more popular than ever, approaching visual studio levels of use (though this is driven mostly by iPhone apps).
The issue isn't total sales its percentage of profits. Even when apple was about 8% of US sales it represented about 90% of the profits in the market. Quite simply Apple is one of the few vendors that isn't competing mostly on price and selling systems for approximately cost so as to get other business (like servers).
You can use flash fine on OSX. The flash issue is on the iphone, who does creative work on any cell phone?
As for dumbing down the OS interfaces.... you can alter you.xinitrc and replace quartz-wm with whatever window manager you like. On the Aqua side there are always hacks to make the system more sophisticated that come out. I'll admit Lion doesn't have many yet, but Snow Leopard got plenty.
Quite simply if you've been using Apple for years you should be easily able to customize past these problems.
1) That Apple hardware is to expensive and pros are drawn towards PCs to keep costs down 2) That Lion's move towards consumer (i.e. less expensive) devices is bad for pros.
The reality is that 10.4-10.6 did a lot for pros. Core animation, core video, core audio, core image, Quartz Composer.... The pros have plenty of technology they need to take advantage of.
If you look at their todo list they could release it constantly it won't matter. What they are doing is essentially creating extensions for their commercial products that work with Hadoop. I think they have every intention of trying to get the small parts that need to be in Hadoop back into the main tree.
This analysis was done in the days of video stores but..
While the vast majority (like 90%+) of rentals are new releases and porn people choose their video stores based on a wide range and selection. People consider access to foreign movies, classics, indys.... very important even though they don't choose to watch them very much.
-- there is no other way to get content on the device without going through Apple's store.
There are of course 2 ways to get content on the device without going through the Apple store sold by Apple:
a) Use the developer SDK and create provisioning files for yourself and friends b) Repoint your device to any set of servers running the enterprise SDK and then they play the role of Apple in terms of securing software.
-- All well and good, but how would any of the above be threatened by allowing you to download third-party apps?
You are allowed to download third-party apps. What you can't do is run an application without a provisioning file. And all that requires is that you have communicated with the developer and let him know what iPhone / iPod/ iPad you intend to install it on.
The issue is not one of not being able to install applications. The issue is that developers can't use a model where they compile software and distribute it themselves to a bunch of people they don't know freely without anyone checking it out first.
This isn't high school philosophy. There are going to be a limited number of systems, perhaps one. The freedom to choose is a collective responsibility which as a society will make it much more difficult for other people to choose in the future.
One can argue the advantages of a train system system over a highway system. But once the vast majority of the housing in your country is in suburbs which one you should have chosen is a moot point.
Stallman did not say he was happy about Jobs death, but that he was happy to see Jobs influence diminished. Most people don't understand the politics in the computer world, but easily understand that fringe public interest groups can often have disagreements with big corporations.
No question. The mid 1990s response of Microsoft to the web was to create a platform specific version of the internet. ActiveX wasn't the web, but it was the internet. I was doing stuff by '98 like embedding ActiveX controls that would do address correction on address boxes in applications. The solution mixed internet, LAN and local information in a way that 15 years later I'd be nervous to try. Without getting into details think about a simple table join to create a view where one table is local, one is on a lan server (not internet accessible) and the 3rd was on the internet. Do the pieces exist today, yeah. Is it the sort of thing you could whip up in Visual Basic and deploy in a few days, hell no. Oracle has solutions that allow this.
I'm happy the web stayed open and I was arguing the importance of an open internet in the 1990s and early 2000s. That being said though, those were some amazing technologies. As an aside, Shockwave did a good deal of what ActiveX used to in a much safer way. It was Macromedia/Adobe so it wasn't as data centric, but Oracle has done a nice job on the data side. I agree that NaCl sounds cool. I hope it fills the void.
As an aside. (4) was the worst in terms of proprietary because it caused people to go proprietary without understanding what they were doing. The people authoring intranet applications in VB that only ran against IIS often didn't understand how married they were going to make their company to Microsoft technology and the companies themselves didn't understand it.
(2) I don't see how that has to be platform specific. There is no reason there couldn't have been generic push protocols. Pointcast made their money from advertising, implementing clients for Windows was just a cost.
I'm not sure why that matters, the Mozilla Suite was the form of the "Firefox" in the early years. The Mozilla suite was supposed to be the open sourced layers under Netscape Communicator Suite. Phoenix (which became Firebird which became Firefox) was at that time a minority project reacting to AOL control, bloat being an issue. When AOL shutdown Netscape many of the employees moved over to the Mozilla foundation and Firebird became a much higher priority project.
In terms of dates Aug 2002 there was a source version fork. Sep 2002 were the first Pheonix binaries. Safari was beta on Jan 2003 and released June, 2003.
I quite explicitly though talked about Gecko. Gecko was started in 1997 and released in 1998.
Netscape and IE battle for dominance. I.E. wins in a landslide and websites design around I.E. (trident's) quirks. Netscape creates a whole new rendering engine (Gecko) and stars building browsers based on it. Gecko is standards compliant and somewhat compliant with the older Netscape renderings. Website designers essentially consider trident not W3 to be the effective standard.
In that environment getting rendering to work right was a major focus. But the problem was solved. 1) FF became better enough that early adopters saw the advantages and market share rose above the 1% level. 2) Microsoft discontinued I.E. for Mac and Apple released Safari (kHTML) 3) Smart phones and their browsers (like Opera mini) started to become a factor, and Windows 7 hadn't been successful in that platform.
In that environment web designers faced a choice of: design a FF/Safari site, an I.E. site and a mobile site of be standards compliant. The move towards standards compliance required Microsoft to become standards compliant. So rendering correctly is no big deal (that is reliability was satisfied).
As for obsession with performance being tied to Chrome that makes sense. The reason Google got in the browser market was that JavaScript performance was bad on all the competitive browsers. Rendering wasn't an issue since Chrome used Safari's rendering engine.
youtube videos take a long time to load. My normal usage pattern is to click on a bunch of videos and let them load in the background. 2-5 videos at once is common.
And I might get distracted and want to look at something else. I'd love my system to be reliable with 10 video streams.
Slashdot Mirror
They didn't get into enterprise server rooms very much. They had niches:
1) People who wanted the g4/g5 for supercomputing applications. Xserve were much cheaper than IBM p-series. Those people bought tons of XServes. When they switched to Intel though...
2) Small business or schools with a small number of racks.
3) Rendering farms for certain specialized applications that had crossed over.
When Apple got in the server room it was with the G4/G5. XServers were way way cheaper than IBM p-series machines. Once they switched to Intel it is hard to see what specific advantages Apple has.
IT is a market that Apple is doing well in. Huge numbers of programmers use Apple and have since the 10.2 days.
Now if you mean enterprise... Apple has little interest in meeting the criteria for enterprise. They might make some minor concessions but really to satisfy enterprise requires a separate product line and then that product line drives your business. All their competitors are focused primarily on enterprise.
The pro multimedia market isn't a safety net for a company the size of Apple. Especially with Apple's margins. If Apple goes out of style they find new markets or they end up in dozens of niches not just 10- and have to rethink the business entirely to fit those niches.
Well written post. It is absolutely true that Apple doesn't allow for layering on their advanced stuff. Being a developer for Apple requires constant upgrading and for a chain of products that is impossible.
Ironically since Chen left Microsoft they've gotten less interested in legacy support as well. They are becoming more like Apple (though the difference is still fast). No question Windows is going to become the main platform for that sort of layering but it might end up being impossible in Windows as well. Interesting possible niche for Linux which is excellent on the complex layering and about maintaining old APIs.
Huh? Many of my applications still take a long time to start while starting new windows in them is much faster.
The problem isn't fixed.
OSX supports keyboards with more keys, that's a hardware choice.
As for the only menu being on top.... Yeah that's a switch. The advantage is you can see what application is tied to your window manager as active with a glance. As an aside in Display item of system preference you can change behavior.
They took it out. The same way they dropped classic after a few years. It is part of apple culture to make developers keep up and not have 20 year old software running around.
Programmers and SE's? Apple early on pretty much stole the Linux development crowd around the OSX 10.2 days. Seems to me DarwinPorts, Fink, Terminal and Quartz-wm still exist and offer a very good environment. As far as OSX development, Developer tools are more popular than ever, approaching visual studio levels of use (though this is driven mostly by iPhone apps).
The issue isn't total sales its percentage of profits. Even when apple was about 8% of US sales it represented about 90% of the profits in the market. Quite simply Apple is one of the few vendors that isn't competing mostly on price and selling systems for approximately cost so as to get other business (like servers).
Well argued comment on Final Cut Pro. But of course by your own admission that is a small piece of the pro-market.
You can use flash fine on OSX. The flash issue is on the iphone, who does creative work on any cell phone?
As for dumbing down the OS interfaces.... you can alter you .xinitrc and replace quartz-wm with whatever window manager you like. On the Aqua side there are always hacks to make the system more sophisticated that come out. I'll admit Lion doesn't have many yet, but Snow Leopard got plenty.
Quite simply if you've been using Apple for years you should be easily able to customize past these problems.
The article is essentially 2 contradictory points
1) That Apple hardware is to expensive and pros are drawn towards PCs to keep costs down
2) That Lion's move towards consumer (i.e. less expensive) devices is bad for pros.
The reality is that 10.4-10.6 did a lot for pros. Core animation, core video, core audio, core image, Quartz Composer.... The pros have plenty of technology they need to take advantage of.
If you look at their todo list they could release it constantly it won't matter. What they are doing is essentially creating extensions for their commercial products that work with Hadoop. I think they have every intention of trying to get the small parts that need to be in Hadoop back into the main tree.
I think they are telling the truth about their goals:
1) Get Hadoop to work on Windows servers
2) Create a Windows server management interface for Hadoop
3) Create SQL Server extension to manage Hadoop.
And the motive is:
a) Sell server licenses
b) Sell SQL Server licenses
They are quite publicly indicating their intention is to embrace and extend:
1) Get Hadoop to work on Windows servers
2) Create a Windows server management interface for Hadoop
3) Create SQL Server extension to manage Hadoop.
So we don't have to speculate, that's what they say they are doing. That being the case all that stuff might be useful for Hadoop.
This analysis was done in the days of video stores but..
While the vast majority (like 90%+) of rentals are new releases and porn people choose their video stores based on a wide range and selection. People consider access to foreign movies, classics, indys.... very important even though they don't choose to watch them very much.
-- there is no other way to get content on the device without going through Apple's store.
There are of course 2 ways to get content on the device without going through the Apple store sold by Apple:
a) Use the developer SDK and create provisioning files for yourself and friends
b) Repoint your device to any set of servers running the enterprise SDK and then they play the role of Apple in terms of securing software.
-- All well and good, but how would any of the above be threatened by allowing you to download third-party apps?
You are allowed to download third-party apps. What you can't do is run an application without a provisioning file. And all that requires is that you have communicated with the developer and let him know what iPhone / iPod/ iPad you intend to install it on.
The issue is not one of not being able to install applications. The issue is that developers can't use a model where they compile software and distribute it themselves to a bunch of people they don't know freely without anyone checking it out first.
This isn't high school philosophy. There are going to be a limited number of systems, perhaps one. The freedom to choose is a collective responsibility which as a society will make it much more difficult for other people to choose in the future.
One can argue the advantages of a train system system over a highway system. But once the vast majority of the housing in your country is in suburbs which one you should have chosen is a moot point.
Stallman did not say he was happy about Jobs death, but that he was happy to see Jobs influence diminished. Most people don't understand the politics in the computer world, but easily understand that fringe public interest groups can often have disagreements with big corporations.
No question. The mid 1990s response of Microsoft to the web was to create a platform specific version of the internet. ActiveX wasn't the web, but it was the internet. I was doing stuff by '98 like embedding ActiveX controls that would do address correction on address boxes in applications. The solution mixed internet, LAN and local information in a way that 15 years later I'd be nervous to try. Without getting into details think about a simple table join to create a view where one table is local, one is on a lan server (not internet accessible) and the 3rd was on the internet. Do the pieces exist today, yeah. Is it the sort of thing you could whip up in Visual Basic and deploy in a few days, hell no. Oracle has solutions that allow this.
I'm happy the web stayed open and I was arguing the importance of an open internet in the 1990s and early 2000s. That being said though, those were some amazing technologies. As an aside, Shockwave did a good deal of what ActiveX used to in a much safer way. It was Macromedia/Adobe so it wasn't as data centric, but Oracle has done a nice job on the data side. I agree that NaCl sounds cool. I hope it fills the void.
As an aside. (4) was the worst in terms of proprietary because it caused people to go proprietary without understanding what they were doing. The people authoring intranet applications in VB that only ran against IIS often didn't understand how married they were going to make their company to Microsoft technology and the companies themselves didn't understand it.
(2) I don't see how that has to be platform specific. There is no reason there couldn't have been generic push protocols. Pointcast made their money from advertising, implementing clients for Windows was just a cost.
I'm not sure why that matters, the Mozilla Suite was the form of the "Firefox" in the early years. The Mozilla suite was supposed to be the open sourced layers under Netscape Communicator Suite. Phoenix (which became Firebird which became Firefox) was at that time a minority project reacting to AOL control, bloat being an issue. When AOL shutdown Netscape many of the employees moved over to the Mozilla foundation and Firebird became a much higher priority project.
In terms of dates Aug 2002 there was a source version fork. Sep 2002 were the first Pheonix binaries. Safari was beta on Jan 2003 and released June, 2003.
I quite explicitly though talked about Gecko. Gecko was started in 1997 and released in 1998.
Think about it this way.
Netscape and IE battle for dominance. I.E. wins in a landslide and websites design around I.E. (trident's) quirks. Netscape creates a whole new rendering engine (Gecko) and stars building browsers based on it. Gecko is standards compliant and somewhat compliant with the older Netscape renderings. Website designers essentially consider trident not W3 to be the effective standard.
In that environment getting rendering to work right was a major focus. But the problem was solved.
1) FF became better enough that early adopters saw the advantages and market share rose above the 1% level.
2) Microsoft discontinued I.E. for Mac and Apple released Safari (kHTML)
3) Smart phones and their browsers (like Opera mini) started to become a factor, and Windows 7 hadn't been successful in that platform.
In that environment web designers faced a choice of: design a FF/Safari site, an I.E. site and a mobile site of be standards compliant. The move towards standards compliance required Microsoft to become standards compliant. So rendering correctly is no big deal (that is reliability was satisfied).
As for obsession with performance being tied to Chrome that makes sense. The reason Google got in the browser market was that JavaScript performance was bad on all the competitive browsers. Rendering wasn't an issue since Chrome used Safari's rendering engine.
youtube videos take a long time to load. My normal usage pattern is to click on a bunch of videos and let them load in the background. 2-5 videos at once is common.
And I might get distracted and want to look at something else. I'd love my system to be reliable with 10 video streams.