iTunes stores runs on WebObjects, which heavily uses Java. So Apple must have already made arrangement with Oracle to supply the JVM.
1. WebObjects is even less updated and modernized than Apple's Java port. Apple hasn't released anything since 2008. It has been over 2 years since they released
any updates. It is stuck. It got
no upgrade with the Snow Leopard upgrade. Largely what is there now is what is there when Leopard shipped with some limited incremental updates that most likely originated as bug reports.
2. WebObjects was at one time portable. If Apple ports WebObjects to Linux or Solaris (i.e., the OS the other significant portions of their back-end store run on ) then still can run the iTunes store. Just not no Mac OS X.
You are assuming that Mac OS X is a requirement to run the iTunes store. It really isn't.
If Apple spent the last 2 years porting WebObjects to Linux or Solaris the don't have to make any arrangements for Oracle to port anything.
The UI elements of the WebObjects apps.... port those so they use HTML5/Javascript. It would be hard to believe that Apple would pay any heed to the browser/client side aspect of WebObjects leveraging any Java at all. In other words, they would turn WebObjects leveraging of Java into purely a server side implementation issue. At that point..... does it really make sense to only target Mac OS X as the primary development/deployment OS for WebObjects ?
3. Apple is soooooo committed to the server market.
a. dumped XServer RAID.
b. XServe sits there languishing even after the extended delay in getting updated Mac Pro out.
c. The Lion sneak peak is about 100% filled with userlevel programs, an Appstore , and making the user experience look more like an iPad. If there is where 98+% of the Lion focus is how likely is it going to make a great foundation for an updated Mac OS X Server ?
All that "backend" , "faceless" applications and infrastructure... Apple is spending time on that ? [ perhaps that is too boring stuff for a Jobs run circus show ... but there are lots of indications that Apple is drifting away with anything that isn't bright and shiny. ]
4. Herds more developers into using Mac OS X only APIs and tools.
[ Remember the "can't use any other language but ours" move this Spring? ]
Apple wants to maintain only their own software now onwards, and doesn't want to develop third party software.
Right. The can switch to a mode where only their use of WebObjects (on Linux or Solaris ... their choice) is all they will care about and anyone else who was using Java for something else can hit the road.
If Apple was showing concern for others and good stewardship they would have arranged to make an joint announcement with Oracle that they would be taking over Mac OS X Java work in the future. Not some backhanded "oh by the way this is now a dead end street" opaque announcement.
You guys should go back to old days where we had IE for Mac from Microsoft, which they stopped making and Apple had no option other than to come up
Rewritting history? Safari started before IE stopped. Sure Mac IE was lagging a bit behind Windows IE but that is different than canceled.
http://en.wikipedia.org/wiki/Apple_Safari#History_and_development
Apple and Microsoft had an agreement to make IE for Mac. At end of 5 year contract Apple decided to take control. ( cruft like ActiveX and gross proprietary cruft that Micrsoft was throwing in at that point certainly helped to make that prudent move. )
The biggest loser of this was obviously MS, IE's market share is all time low.
Cough... like Firefox and Mozilla had nothing to do with that decline in market share.
So Apple wants to support their platform, and if Oracle thinks OS X Platform is that important, they will maintain their software, Java, on Mac OS X Platform.
If Apple wants Oracle to support Mac OS X then they should have sat down with Oracle long ago and come up with a transition plan. There is zero evidence that they did this. If Apple "steve'd" the java port and didn't give Oracle/Sun a long enough head's up then this isn't going to fly.
What will end up happening is that the OpenJDK that using XWindows will be the "port". That is the least expensive thing for Oracle to do. Anyone who wants some backend, nonGUI apps (e.g., Apple if haven't finished porting WebObjects to a sever OS with a better long term future ) can use it.