Snow Leopard to Get Cocoa Finder and ImageBoot

[AidenShaw]

No Core Duo systems....

I hope Snow Leopard is completely, fully 64-bit Cocoa

Good idea, even though it will make the Core Duo Apples obsolete...

You'll probably want to allow support for 32-bit applications, though.

 

[AidenShaw]

It has to be in order to maintain backwards compatibility. For instance the win32 function WriteProcessMemory() is still available under Vista.

BS on the backwards compatibility argument.

Citing the example of a privileged debugger API is hardly evidence of what you're trying to imply - that NT is in the league of systems like Win 3.1 and OS 9 which lack memory protection between processes.

You may also want to look at how OSX allows debuggers to write into other processes' memory (see vmmap and other OSX APIs).

However, the maximum permissions allow writing so that the debugger can request write access to a page to insert breakpoints.

Permissions for executables appear as "r-x/rwx" to indicate these permissions

http://developer.apple.com/documentation/Darwin/Reference/Manpages/man1/vmmap.1.html


Pot. Kettle. Black.

 

[CallsignBaron]

I just highlight the portion I want to print and then print. Only the highlighted portion prints then. Have you tried this?

Tried this but it didn't work. Highlighting text then clicking print or doing P still selects the entire document to be printed. If I am doing it wrong then please let me know how this works for you. BTW, using Leopard 10.5.5.

"Print Selection" is very useful and saves paper and ink by only printing what you want out of a document, email, webpage or whatever. I don't think it is a feature that could be too hard to implement and it goes well with Apple's desire to go green.

 

[digitalbiker]

Tried this but it didn't work. Highlighting text then clicking print or doing P still selects the entire document to be printed. If I am doing it wrong then please let me know how this works for you. BTW, using Leopard 10.5.5.

"Print Selection" is very useful and saves paper and ink by only printing what you want out of a document, email, webpage or whatever. I don't think it is a feature that could be too hard to implement and it goes well with Apple's desire to go green.

One thing you can do that is very similar to print selection.

1) Select text to be printed.
2) Drag the selected text to desktop. This creates a text clipping.
3) Right click (cntrl-click) on Desktop clipping and select print.

This print the selected text. You might be able to drag the text clipping to an icon of the print queue and it would work as well however I can't test it at the moment because I am on "The Road" and don't have a printer.

 

[Cromulent]

You may also want to look at how OSX allows debuggers to write into other processes' memory (see vmmap and other OSX APIs).



http://developer.apple.com/documentation/Darwin/Reference/Manpages/man1/vmmap.1.html


Pot. Kettle. Black.

You really can't be serious? That OS X API has strict permissions to set whether something is writable in memory.

WriteProcessMemory from the MSDN documentation has no such restrictions, requiring you to just make a call to OpenProcess and then to use the handle returned in WriteProcessMemory to write an entire buffer to arbitary memory points. It is not even limited to debugging either (the Windows functions).

How you can compare the two is beyond me.

 

[Trajectory]

I hope they offer Finder in more than just the Cocoa flavor. Why not add Finder Mint? Finder Cherry? Finder Kiwi-Coconut?

 

[Tosser]

I hope they offer Finder in more than just the Cocoa flavor. Why not add Finder Mint? Finder Cherry? Finder Kiwi-Coconut?

I didn't know you sucked on your computer!?

(yes, connotations intended) …

 

[Xtremehkr]

I'm really looking forward to Snow Leopard, new features are important for adding utility, but what makes an OS great is when the reliability, performance, and compatibility are excellent. The evolution of OSX continues to make it the best operating system available.

 

[MattInOz]

Heh, that is complete rubbish. VMs have nothing to do with breaking an encryption scheme or not.

Indeed Wouldn't VM be a good way to protect the DRM.
It could lock control of the chain to itself the the DRM'ed info would never have to leave the VM.

 

[Krevnik]

Indeed Wouldn't VM be a good way to protect the DRM.
It could lock control of the chain to itself the the DRM'ed info would never have to leave the VM.

Not really. The VM actually makes it easier to figure out what is going on. An OS can put up walls around its DRM code so you can't poke at it... but put the OS in a VM, and suddenly you can again, since the OS is now inside your control.

It is actually very problematic to secure data when the same people who you don't want to know the keys are the same ones who need to know the keys to use the data. Someone who wants the key really just needs to find where they can make a copy of the key by eavesdropping, snooping, etc on their computer via code. VMs make these techniques easier, not harder.

 

[Cromulent]

It is actually very problematic to secure data when the same people who you don't want to know the keys are the same ones who need to know the keys to use the data. Someone who wants the key really just needs to find where they can make a copy of the key by eavesdropping, snooping, etc on their computer via code. VMs make these techniques easier, not harder.

Those that want to find out that information are unlikely to be put off by the hurdles an OS puts in their way. Just look at those who broke the iTunes DRM scheme even though Apple put up quite a few defences to stop it. As far as I am aware they did it in the native OS and not in a VM although I do see your point.

 

[pmpknetr21]

• Did you consider that maybe he likes the smaller dimensions of the MacBook?
• Did you also consider that maybe he is unable or unwilling to pay $1,999 base price for a MacBook Pro?
• Windows has had plenty of innovation. Including innovation that Apple eventually "borrowed."
• Windows is also not crap. If it was, it wouldn't have a 90%+ market share.
• You can't get spyware/adware/viruses in Windows unless you a genuinely stupid person who clicks every link and downloads every odd program. Mac OS X is hardly safe, either, from the stupidity of certain users.

1. Giving up Mac OS X and the entire platform because of no firewire on a consumer machine is ridiculous, no matter how you justify it with dimensions.
2. Price? I get that. No arguments here.
3. Windows HAD innovation. They lost their way when they got to where they are. Hence the mass-exodus to Mac.
4. Having 90% market share does not mean it isn't crap. Slavery probably had 90% market share and that was crap. (Dramatic reference? Yes. But it makes a point.) Another reference: there are more Honda Accords on the road around the world, but does that make them better or more innovative then a Mercedes Maybach, BMW 7 series or Aston Martin? Not so much.
5. Hardly safer? Dude, seriously? Do you really wanna argue this point? I don't even know where to start with that? There's an entire industry making millions, if not Billions, off of poor Windows security. Can the same be said for OS X? Not so much.

Just my 2 cents.

 

[Krevnik]

Those that want to find out that information are unlikely to be put off by the hurdles an OS puts in their way. Just look at those who broke the iTunes DRM scheme even though Apple put up quite a few defences to stop it. As far as I am aware they did it in the native OS and not in a VM although I do see your point.

Comparing the designs of Fairplay and AACS is like comparing a toy safe with a bank vault. Apple's incentive with Fairplay is to keep the providers happy and no more than that. Anything more is potential profit lost in the eyes of Apple. AACS has requirements that make it very expensive and difficult to comply with. Data does remain encrypted end-to-end, and the vulnerable spots are where unencrypted keys are passed around. PowerDVD had to rush and release an update to not lose its AACS license for accidentally storing unencrypted keys in its own RAM space.

If you actually want to know why SJ called Blu-Ray a bag of hurt, he was likely talking about the AACS requirements. Requirements of trusted drivers (can't let a 3rd party driver slip in), encrypted channels in RAM, through the system bus, and to the display via HDCP is time intensive. Apple doesn't want to do that work until there is clear benefit/cost ratio (i.e. lost sales due to lack of Blu-Ray support actually starts costing them more than it would take to implement Blu-Ray and maintain such a tightly controlled security model).

 

[drsmithy]

If you had an ounce of knowledge about operating systems you would realise that Windows is not secure by design.

Christ, you are allowed to modify the memory of another running process under Windows. Hardly secure is it?

If you had "an ounce of knowledge about operating systems" you would realise that you were wrong.

 

[drsmithy]

You really can't be serious? That OS X API has strict permissions to set whether something is writable in memory.

WriteProcessMemory from the MSDN documentation has no such restrictions, requiring you to just make a call to OpenProcess and then to use the handle returned in WriteProcessMemory to write an entire buffer to arbitary memory points. It is not even limited to debugging either (the Windows functions).

How you can compare the two is beyond me.

You cannot write to a process's memory space unless you have appropriate privilege levels or permissions to do so (and the privileges/permissions system in Windows is *far* tighter and more granular than the classic UNIX-style one in OS X, where UID==0 means you can do whatever you want).

You might find this helpful.

 

[Cromulent]

You cannot write to a process's memory space unless you have appropriate privilege levels or permissions to do so (and the privileges/permissions system in Windows is *far* tighter and more granular than the classic UNIX-style one in OS X, where UID==0 means you can do whatever you want).

Windows also implements privileged access controls in the same way. The only difference is that the superuser is not called root in Windows (source: Modern Operating Systems).

As far as Windows providing far tighter and more granular access controls, I'm yet to be convinced. Given that Windows allows the same superuser security overrides as Unix I don't think that argument holds much water. At the very least they are comparable.

You might find this helpful.

Interesting article. The function quoted earlier though is completely different and does not provide the same level as access even given the appropriate permissions. Where as WriteProcessMemory allows you to write a buffer of unchecked code into another process the Unix function is mearly there for setting breakpoints with a debugger.

If you know of a comparable Unix function to WriteProcessMemory I would be very interested to hear about it. All I've been able to find on the subject (and I've looked before for a couple of projects of my own) are to use IPC for process communication rather than being able to directly modify another processes memory space even with the required privileges.

 

[drsmithy]

Windows also implements privileged access controls in the same way. The only difference is that the superuser is not called root in Windows (source: Modern Operating Systems).

Then your source is wrong (or you are misunderstanding it). Windows NT has no superuser concept. *ALL* user accounts are subject to ACLs (whereas root is not subject to security restrictions).

As far as Windows providing far tighter and more granular access controls, I'm yet to be convinced.

Then keep reading your operating systems books.

Given that Windows allows the same superuser security overrides as Unix I don't think that argument holds much water. At the very least they are comparable.

They're comparable, sure. But the security capabilities of Windows are a superset of those found in classic UNIX.

[/quote]If you know of a comparable Unix function to WriteProcessMemory I would be very interested to hear about it. All I've been able to find on the subject (and I've looked before for a couple of projects of my own) are to use IPC for process communication rather than being able to directly modify another processes memory space even with the required privileges.[/QUOTE]

Likely this is due to the different programming mindsets of each environment.

 

[wrldwzrd89]

Windows also implements privileged access controls in the same way. The only difference is that the superuser is not called root in Windows (source: Modern Operating Systems).

As far as Windows providing far tighter and more granular access controls, I'm yet to be convinced. Given that Windows allows the same superuser security overrides as Unix I don't think that argument holds much water. At the very least they are comparable.

Windows DOES provide more granular access controls than traditional UNIX, but how much tighter they are is debatable, and this is only true compared to a pre-Tiger Mac OS X system, because Tiger and later support Windows-style ACLs on top of traditional UNIX permissions.

Interesting article. The function quoted earlier though is completely different and does not provide the same level as access even given the appropriate permissions. Where as WriteProcessMemory allows you to write a buffer of unchecked code into another process the Unix function is mearly there for setting breakpoints with a debugger.

If you know of a comparable Unix function to WriteProcessMemory I would be very interested to hear about it. All I've been able to find on the subject (and I've looked before for a couple of projects of my own) are to use IPC for process communication rather than being able to directly modify another processes memory space even with the required privileges.

I remember reading somewhere about a UNIX function that is far more powerful and far-reaching than Windows's WriteProcessMemory. It is similar to what fork and execve do when used together, except that it allows you to essentially inject code into some other running process's code space and execute it. True, you can't just inject code into anything - its use is restricted to processes owned by the currently logged-in user. I wish I could recall where I read this though.

 

[scottwaugh]

Nice to hear

I'm definitely looking forward to a rewrite of Finder as it has a habit of occasionally disappearing (silent crashing?) in OS X, which is most annoying - as I'm normally moving files when it does this.

Hopefully they'll extend its functionality as well. Bring it on Apple.

 

[AppleMojo]

I notice a lot of users bashing Apple for "poor quality control" and saying that osx has not been really stable since panther. I came from M$ a few months a go and osx is heaven compared to it. I think you've been treated too good by Apple . There are some flaws but nothing in the realm of what windoze users suffer.

Finder rewrite in cocoa sounds good though. Hopefully it has tabs.

Well said.

Software isn't perfect and I agree that there have been issues, but lets not forget that not only does OS X have superior quality and features, it was also converted to x86 within the last few years.

Just keep it all in perspective.

 

[AppleMojo]

i think all of us should at least remember that it was microsoft that have brought us here today as far as computer technology is concern? it was not apple and their os, but microsoft that opens the door of possibilities..

Note: i am not a microsoft fanboy, i am getting a mac soon.. but i think we should all be aware how much microsoft have contributed to the computer scene.. i still appreciate my PC very much, instead of condemning their os like some other people in this forum does..

/puke

 

[AidenShaw]

WriteProcessMemory from the MSDN documentation has no such restrictions, requiring you to just make a call to OpenProcess and then to use the handle returned in WriteProcessMemory to write an entire buffer to arbitary memory points.

You apparently aren't familiar with reading the Windows documentation. Note:

The OpenProcess documentation says that the call will not give you a handle with the ability to write remote memory unless your process is specifically authorized to modify the remote process, for example by holding the SeDebugPrivilege right.

To open a handle to another local process and obtain full access rights, you must enable the SeDebugPrivilege privilege. For more information, see Changing Privileges in a Token.

It is the usual case that privilege issues are described in the call that creates the handle, not so much in the APIs that use the handle.

You've been pwned on this memory protection claim, give it up

It is not even limited to debugging either (the Windows functions).

And OSX doesn't permit arbitrary writing, even though the casual mention in the vmmap page is that the goal is breakpoint insertion? LOL

Can the OSX debugger change the values of variables in the target process? That's certainly writing to arbitrary locations in the other process!

How you can compare the two is beyond me.

It's pretty obvious that the two operating systems provide similar functions for debuggers and other privileged software to modify other processes - so it's pretty easy to compare.

It would be hard to *contrast* them, however, since the capabilities are similar.

 

[raccoontail]

Since they are rewriting the Finder anyways... I hope they'll re-implement the "Put Away" command I miss from the 9" macintosh days... It was so intuitive - you just pulled a folder out from deep in your hard drive, set it on the desktop while you were working on it, and when you were done, just clicked on it and pressed Apple-Y and it refiled itself.

 

[wrldwzrd89]

Since they are rewriting the Finder anyways... I hope they'll re-implement the "Put Away" command I miss from the 9" macintosh days... It was so intuitive - you just pulled a folder out from deep in your hard drive, set it on the desktop while you were working on it, and when you were done, just clicked on it and pressed Apple-Y and it refiled itself.

Totally agreed there - that command was so useful. I made use of it as a sort of undo if I accidentally put a bunch of things in the Trash I shouldn't have - Apple-Y sent them all back where they came from.

 

[jlewis2k1]

I am beginning to really like this upcoming release of OS X. I am very curious if 10.6 would be the last update before moving on to XI? If so, I think that would be very interesting update.