Slashdot Mirror
A Closer Look At Apple Leopard Security
Last week we discussed some of the security features coming in Leopard. This article goes into more depth on OS X 10.5 security — probably as much technical detail as we're going to get until the folks who know come out from under their NDAs on Friday. The writer argues that Apple's new Time Machine automatic backup should be considered a security feature. "Overall, Mac OS X 10.5 Leopard is perhaps the most significant update in the history of Mac OS X — perhaps in the history of Apple — from a security standpoint. It marks a shift from basing Macintosh security on hard outside walls to building more resiliency and survivability into the core operating system."
Why doesn't everyone (Apple, Microsoft, Linux/Unix people) work together on security? Its the one thing that everyone benefits from.
"Overall, Mac OS X 10.5 Leopard is perhaps the most significant update in the history of Mac OS X -- perhaps in the history of Apple
Maybe in the history of Mac OS X, but definitely not the history of Apple itself. I'd say that would be, oh, the shift to Unix.
Mikey-San
Karma: +Eleventy billion (mostly affected by watching Celebrity Jeopardy)
Reading this made me wonder. What would happen if you had an important file you temprarly drop it in a public location then move it out. once the person downloaded it. Then someone goes and runs time machine on the public directory and picks up the file that you deleted.... Also will time machiene pick up different permissions set on a file at different time. You made it and tested it as 777 then after you assure it physically works you bring it down to 755 will it allow you to go back in time and get the permission 777 of the file...
While I do agree having good backups is important part of security... Perhaps just perhaps because it is so easy there is a security problem with it.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
The much-needed focus on availability is a real breath of fresh air. If one can recover a previous state (i.e. if it is available), it's a great deal easier to restore integrity. Confidentiality improvements are always welcome, of course, but they'll never be complete, and availability allows us to recover after the fact.
Also, Time Machine is a great forensic tool.
Overall, of course, I'm lauding the article more than 10.5, since I'm unaware of any of these features being truly new to the IT world.
I am the one true god. However, as an atheist, I don't believe in myself. I guess I have a self-esteem problem.
Wait... don't tell me they implemented RFC 3514 .
A Human Right
How did the parent make it past the lameness filter? Seriously
And it sounds from many of these changes, that Apple can see a future where they would be suffering like Windows because of being a larger target.
App signing and stack randomization has already come to windows.
System restore and shadow copy exists in Windows, though it looks like Apple will be providing a better backup system out of the box.
And the sanboxing sounds a lot like UAC with the exception that you wont get a prompt. The sandboxed app will just be denied.
It sounds like Apple OS is not inherently more secure than Windows. It is now a larger target, so it needs these new protections.
Apple just made it easier to recover deleted files, if you're using backups. If you're not using backups, there is no problem. OS X has also long had a "secure delete" option that not only deletes the file, but writes over it with random data multiple times, ala DoD requirements. I'd be willing to bet that also does the same on your time machine backups.
That means then Apple would have to support unknown hardware..... won't happen. Thats the benefit to owning apple hardware and OS... I can point my finger at one company and expect to get it fixed right the first time.
Another poster has addressed the core issues (secure delete, etc), but one other thing needs to be pointed out: At least anecdotally, I suffer data loss far more often than I have hackers breaking into my system (at least that I know of) or having to deal with the compromise of sensitive information from my hard drive.
There is a greater risk for many people in lack of backups vs. outside threats who have sufficient access to the machine to see data we've deleted without bothering to secure delete it or delete the backups.
There are jobs that demand that level of security, but there you are dealing with taking every hard drive that touches the system out and locking it in a safe at the end of the day. Backups, in and of themselves, are not the issue.
Integrate Keynote and LaTeX
"With Time Machine making it easier to back up for all users, especially individuals not already protected by some corporate backup system, Apple is doing more to improve security than any upgrades to firewalls or Safari ever could."
Although I am a fan of backups, this is really silly. Even if we assume that users have Time Machine turned on, that they have external media on which to back up, that they manage to actually have everything turned on and hooked up to do the automated backup, there's still one hole in this argument.
In order for a backup to offer protection you need to know that there is something that needs to be restored from the backup. If most security attacks are by nature silent then you won't realize that you have been compromised and will not preserve a recent backup much less restore it.
Unless there is unlimited storage space for backups there will come a point when Good Data Set A will be replaced on your backup by Corrupted Data Set B. Time Machine likely has no way of knowing that the data it has just backed up is not your good current file, but one that has been damaged. All that it knows is that the file changed.
Three Squirrels
Deleting from Time Machine is as easy as deleting from any other folder in finder.
Here are some step-by-step directions if you really need it: Leopard Time Machine: Delete Files or Folders from Backup
AC
Yes, it's really too bad that it's not going to try to support all the cheap-shit, generic hardware that China can pump out. You do know that's why Windows has sucked so hard in terms of stability, right?
Mac OS X has the "it just works" reputation because of the limited number of hardware configurations on which it runs. They can take full advantage of what's there, because they know exactly what's there. Windows has to take the 'lowest common denominator' approach, to its detriment. Microsoft has tried and failed (though they have gotten better) for twenty years to get Windows to work with generic hardware as well as OS X works on Apple hardware. But when they decided to try to take over the game market, what did they do? They rolled their own hardware instead of just leveraging the existing Windows-on-generic-hardware market. That should tell you something.
I do agree with you on the second point though, Apple does indeed have a gaping hole in its product line where a midrange tower should be. If for no other reason than to make all the people who have been whining for one shut the hell up and buy the goddamn thing. Although, who am I kidding? If one was made, they'd just bitch about the price or specs or something and still not buy a Mac.
Just exclude your homemade porn folders from the Time Machine backup set. Easy. If you forget to do this, just delete the files on your Time Machine drive; it uses the standard .snapshot-style folder layout. No binary databases or big backup blobs that you can't parse and delete yourself.
If you want public key encryption of the backups, set an encrypted DMG to be your Time Machine target. You can even use AES-256 in Leopard.
If you are looking for a rundown of all the new features, you can check out Apple's official listing of the 300 new features. Tech-Recipes has already started releasing screenshots and tutorials detailing many of these.
Apple has to do very little with security, honestly. Compared to a serial-killer, even the car thief looks good. Apple keeps their solid history of security and adds a nice backup platform. If anybody asks, all they have to say is that we are better than Microsoft.
"Code randomization" is a terrible idea. Virus writers will write something that searches around for the right place to patch. Developers will think buffer overflows are now OK, and write worse code. Worst of all, bugs become nonrepeatable and harder to debug. (Great for tech support. Much harder to pin blame on the vendor now.)
They will complain about anything.
They want OS X to be realed for common hardware not realize that apple tried that (with their older OS) and it nearly killed them. And right now they are doing stellar, they way they are going now. Basicly they are just jelious that Linux isn't as good as OS X is.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
I am wondering if some even more basic holes have been filled here.
I have been given to understand that one of the problems with OSX is that in order to make some legacy software work such as applescript, apple had to make a few file settings more open than they should be.
The big example is the one which allows a USB drive with a correc tly set up copy of OSX on it to automatically become the boot drive with full root access to all drives on a restart. IIRC there's even a company that sells these things pre-configured for unnecessarily large sums money.
This is the stuff I most want to see fixed
If you look at Apple's description of the time machine functionality, it's not possible for it to work the way they claim. Suppose my backup drive has a capacity of 80 Gb, and so does my primary drive. I record 79 Gb of data onto my primary disk. I run out of space, delete all of that video, and then record 79 more Gb of video, filling the disk again. Then let's say I go through the cycle for a third time. They're claiming that I can then go back in time and get back my first or second video. No way. I don't have enough total disk space to store all three videos. So realistically, there are implementation limits, which they conveniently don't mention. Their description makes it sound as if everything Just Works, and will never fail to let you recover old files. In reality, it will Just Do Its Defaults, which may or may not be what you would have liked. Does it default to deleting the oldest files first? If so, then that's probably not what you would have liked in many cases, because you probably care more about preserving the 500 kb manuscript of your novel than about preserving the 70 Gb video of your kids' soccer games. Maybe it has some heuristics, so it tends to delete bigger files first, or files of a certain type first. Well, maybe that's what you wanted, but maybe it's not. Or maybe it asks you to make the decision whenever the backup drive fills up. Well, maybe that's what you want and maybe it's not, but it wouldn't be the same thing as the zero-work solution that Apple claims in their description.
In reality, I think you can have some, but not all, of the following:
Personally, what works for me is the unison file synchronizer (I use it on Linux, but it's cross-platform), plus monthly backups on CD or DVD. Using the network file synchronization takes care of two things: (1) I have an off-site backup that's always fairly up to date; (2) it makes it easy to undo mistakes like "oh no, I didn't want to delete that file." The CD backups let me (3) go back in time and get very old versions of files. I'm not saying that my solution is right for everyone. No solution is right for everyone. However, my OSS solution works much better for me than Apple's expensive, proprietary system would work for me.
Find free books.
"Mac OS X has the "it just works" reputation because of the limited number of hardware configurations on which it runs."
I've heard this for years but I still haven't seen ANY hardware sample where Windows "just works". I'd put more value on the fact that Apple based the core of their OS on a unix-like system not the registry/spaghetti mess that has been windows for the past decade plus. I'm sure that eliminating poorly written drivers from the mix does help prevent some of the problems that plague windows but it's not the whole story by a long shot.
Besides, with that argument, Linux should be even more unstable because very few of it's hardware drivers are written by the device manufacturers - many are reverse engineered.
Alex, I'll take keybindings not used by Emacs for $400....
You can exclude things from TimeMachine backups. Exclude your ~/Homemade_porn directory.
I'm hoping that this is meant to be sarcastic, though I'm certainly stretching to find it.
Security hole from hell? Okay, if a person has that kind of access to your machine, your files are really already compromised; cause unless you frequently leave your Mac out in the open with the root password pasted to it, people will rarely get to the point where they can recover incriminating files. On top of that, you can control what time machine does and does not back up.
I'd be willing to bet Time Machine doesn't delete old copies. Otherwise, what's the point of having Time Machine? It's *supposed* to recover deleted files.
There are two types of people in the world: Those who crave closure
and how does wear leveling affect writes over with random data?
In the time people have been complaining about the lack of a mid-range Mac, those same people would easily have saved enough for a Mac Pro...
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
How the hell is that a flaimbait? Parent is right, the implementatios on the new MacOS and Vista are nearly identical. Of course what do you wait from a rabid macboi moderator.
That's true for the normal delete, but I don't know about "secure delete." Secure delete could very well go back through your entire backup set and delete the file utterly.
We won't know for sure until it comes out and someone tests it.
Integrate Keynote and LaTeX
Watch the Apple leopard video. I believe in there, they talk briefly about how TM has the option to permanently remove all versions of a file. It should also be mentioned on the TM feature page Apple has on the web site... in any case it's possible.
It's such an obvious feature it's no surprise it's included. This is versioning 101 stuff.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Mod the parent up to 11 :-).
Besides, with that argument, Linux should be even more unstable because very few of it's hardware drivers are written by the device manufacturers - many are reverse engineered.
I couldn't say it better myself!!
Think Deeply.
OS 9 was more responsive, yes. But, due to cooperative multitasking, if any program crashed, your entire computer did as well. I did some fairly memory intensive Photoshop work for a newspaper on an OS 9 Mac that was packed to the gills with RAM, and I'd have an average of two reboots a day. This can be maddening to the point where you'll want to throw the Mac out the window if you just lost an hour's painstaking work to the fucking bomb.
The OS X came about. Systemwide crashes are a rarity, and in my experience mostly due to hardware failure. If some beta version of Firefox crashes, it dies a lonely death while the other programs keep on chugging.
It's srm rather than shred, but yeah, same idea.
Its OK to truncate data lots of the time, I would go so far as to say most of the time. Do you gnutards really think its a good idea to try to allocate many MBs of RAM to a string storing a filename, when the operating system limits path lengths and filename lengths anyways? Or when you are grabbing data to put into a varchar(255) column in the database, you want to using snprintf to truncate the data and then use the return value to see that it was truncated, and give the user an error message.
What's to keep the virus from just using the underlying trap instruction for its system calls? This is a UNIX system, friends, you don't need to call printf(), you can call write().
I've heard this for years but I still haven't seen ANY hardware sample where Windows "just works".
It really depends on what you mean by "just works". The truth is that Windows does suffer from supporting a larger variety of hardware. Specifically, if you have a Windows XP computer that crashes on a regular basis, there's a very good chance that you either have some sort of malware installed or else have some really crappy drivers. Ignoring malware and crappy drivers, Windows XP is actually a pretty stable OS.
So when you talk about how things "just work", are you only speaking about stability? Because OSX makes it much easier to image machines and put that image on other (different) hardware, for example. OSX also doesn't go into "reduced functionality mode" when you install new hardware. OSX also keeps up to date with the most common hardware so that you don't have to hunt down drivers if you install new hardware. After using a Mac for a few years, I'd say that OSX does a lot of things more sensibly than Windows, and cuts out a lot of annoyances from day-to-day computer usage.
Also, as you mentioned, Linux does a number of things better than Windows these days. When I install Windows XP (or even Vista) on a machine, I usually have to spend an awful lot of time afterwards hunting down drivers. Then I have to install a bunch of different pieces of software, one at a time. I have to find the disks for that software and the serial numbers, and click "next" 50 times for each install. On the other hand, when I install Ubuntu on the same machine, it discovers all my hardware automatically and also automatically installs a bunch of common apps. I can then install additional apps very easily, and update all the software on my machine at the same time, using Synaptic. No user intervention is required during the install.
So are these things included in "just working"?
Application signing, warning dialogs for downloaded files, and the like... these have been Microsoft's first line of defense against cross-zone exploits for a decade now and they have systematically failed. Now Microsoft is using Sandboxing, and that will also fail.
I wish that Apple would decide to photocopy good ideas from Microsoft rather than bad ones. The single set of application bindings for helper applications and URL handlers? That comes from Windows. The idea of giving users the opportunity to open potentially hostile files directly from mail and browser software? That comes from Windows. Open Safe Files? That comes from Windows. Popping up dialogs before automatically doing stupid things, instead of not automatically doing stupid things? That comes from Windows.
The last straw for me was when Safari on OSX warned me that I was downloading an EXE file because it's executable. Not that I was running it. Just that I was downloading it. Holy Mother of Turing!
*sigh*
At least they don't have anything like ActiveX yet.
That's easy. It tracks the changes to the files. If you create a new file by using "save as" that won't be deleted and neither will it's history, but that is obvious because the original file still exists. If you move a file, it is still the same file. If you copy a file, you've made a new file, based upon the old one.
Okay try this one on for size. Make a hard link of a file. Now edit one of the hardlinks and save it (not save-as, just save). Now which one is the copy? From the file systems POV the edited one will be a copy. But from the users point of view it might be the original, especially if they had no way of knowing the hard link had been made.For example, since I don't have Time Machine yet I currently snapshot my home directory by making a image of it populated by hardlinks. this happens in the background so I don't even know it is happening. Nor do the other users on the computer. You can't really say which is the link since it's a hard link not a softlink or alias. A hard link is an identical file system entry as the original and should be indistinguishable. The save will sever the link.
Some drink at the fountain of knowledge. Others just gargle.
Not only does it work together with the hardware, but the software works a little better with the software too. It's a little bit less frustrating than using software under windows. A little bit more stable, a little bit more intuitive, simple, and less maintenance. Hardware doesn't have to play a role here.
Every experiment which ends in a big bang is a good experiment.
No, Apple did NOT try that. The hardware that was released by PowerComputing, UMax, Motorola, Radius, etc. was not generic hardware. It was Apple designed motherboards. I think in some cases they were even manufacturered by Apple and placed in the other company's cases. Apple had deals with those guys that didn't make any damn sense (for Apple). Very different from trying to support "generic" hardware.
Avoid Missing Ball for High Score
The consequences of a privacy breach are incomparably more grave than that of data loss. You could be put in jail, face a divorce, get fired or have your reputation permanently tarnished by content leaked on Internet. Companies will face lawsuits based on intermediate versions of a memos that were never actually distributed.
Suppose you were writing a letter to an old friend and, in a moment of weakness, add a paragraph on how you still have a crush on her and would like to meet. Later you think better of it and send a version without untoward sentiments. What would your wife think if she stumbles upon an earlier draft while looking for your daughter's accidentally deleted school essay? Is the inconvenience of doing manual backups so great to risk suffering for "thought crimes" that were never carried out?
When people burn a letter, cut up a CD or flush something down the toilet, they trust that the stuff stays gone. Computers should follow the same metaphor accurately by default and only retain information on opt in basis. At the very least, they can ensure that archives are only accessed with your permission by asking for a password before showing old files.
If it is an important file, why would you drop it in a public location in the first place ...
GUIs are prone to errors, just like consoles. All that has changed is how the error manifests. When your finger slips at the console you get a typo. When your finger slips during a drag you may inadvertantly issue a mouse up and drop the file being moved prematurely, in the wrong folder. It can be a PITA when you were dragging over a bunch of subfolders in a list view.
It has little to do with stability. Although it is less of a problem than before, one of the biggest hurdles to Linux adoption is drivers for hardware. Without support from the manufacturers, OSX would have the same problem.
From article submission: http://yro.slashdot.org/article.pl?sid=06/07/31/0044201
Re: Vista Previous Versions (Also in 2003 Server)
Some users will find the feature objectionable because it could give the bossman a new way to check up on employees, or perhaps it could be exploited in some nefarious way by some nefarious person. Previous versions of Windows were still susceptible to undelete utilities, of course, but this new functionality makes browsing quite, quite simple.
From today's article:
The writer argues that Apple's new Time Machine automatic backup should be considered a security feature.
- So the same feature that first appeared on Windows Server in 2003 and then on Vista is considered a security risk, especially because it is too 'easy' to use.
- And now the same freaking feature in OS X is considered a 'security feature', and they claim it is even 'easier' to use than Vista's version?
How can logical people even accept information like this? Can we officially rename SlashDot - Apple's new bitch?
Doesn't anyone else find things disingenuous when you can get modded down attacking OS X faster than if you attack FreeBSD or Linux on a OSS site? We now see the same 'coveted' features in Vista are bad, but good in OS X.
SlashDot, I miss the real tech news, OSS information, and honest debate...
"It's Schadenfreude. Making me feel glad that I'm not you!" -- "Gary Coleman", Avenue Q
GLaDOS for President 2016! "Well here we are again. It's always such a pleasure." -- GLaDOS, 2011
Hell, Apple, when importing BSD drivers (read: they don't have to write their own drivers since they benefit from other people's hard work) has even excluded perfectly working BSD drivers. And this doesn't just exclude people who want to run on non-Apple machines. Let's say I buy a stock off-the-shelf component and want to put it in a perfectly legit Mac. Let's use a PCMCIA bridge as an example, as I ran into this once. Even though the project that they took their PCMCIA code from supports the chipset, I can't use it because Apple won't let me. Meanwhile, MS, Linux, and BSD are supporting these without considerable wasted effort. In fact, since the interface to this PCMCIA bridge is nearly indentical to (if not 100% the same as) another one they support, they could make the change by just adjusting an array of integers representing PCI IDs. But they don't. Why is this a good thing?
Get the apple, boot off of the OSX cd, use the password reset utility, now that you ARE root you can do what ever you want to that machine.
Having that kind of access with almost any OS means your data is no longer secure.
So what? Microsoft, Linux, and BSD derived OSs already do this, and it works pretty well for them.
No it doesn't work pretty well for them. I can't claim an authoritative figure, but let's call it most Windows crashes are down to third party drivers. When this happens, you have the dodge-the-blame dance going on between the computer manufacturer (say, Dell), the OS manufacturer (MS) and the peripheral manufacturer (say, Logitech). With Apple, there's no such opportunity for blame avoidance, so they have to keep their shit together.
Linux (and to an even further extent BSD) sidesteps the problem by supporting far less products, and keeps the "no right to complain about free software" ace up its sleeve.
Install linux without the network cord plugged in and with the wireless turned off. Then see what is missing. Windows doesn't connect online during the install process to check for drivers. Hell, even after the install process windows check for drivers. Most linux installs are network card drivers and maybe video card the rest is search online against a very large database of drivers. Microsoft are complete fools for not having a driver database that can be checked during the install process.
Second, this bit about "and to an even further extent BSD" isn't necessarily fair. For example, OpenBSD has Linux beat as far as wireless support. (Although that might have changed, now that Linux developers have integrated some of their work into Linux)
The software was still written for generic hardware, and very device-independent. That's what made it so damned easy to run MacOS 7.5 on my Amiga. The Mac emulators for the Amiga were essentially just device drivers.
Some people complaining would really like a machine that that isn't as large as the Mac Pro. even if they could have saved up enough money to make up the price difference between the cost of the fictional "midrange Tower (or Desktop)" and the MacPro, is that any reason for them to spend the money on the computer when they don't need it? They could use the extra money on software to make their Mac more useful.
Or you can turn it off, or use another backup solution...?
So what? Microsoft, Linux, and BSD derived OSs already do this, and it works pretty well for them.
Wrong.
It may appear that it works well (rI'm referring to only the big name corporations that create OS's), however to keep control and certification of all of the third party device companies, let alone their code changes is a very large task, which leads to mis-communication and eventually more unknown support issues.
Yes, I choose to pay more for Apple products, but when it comes to service I know (once again) I can go to one company to get my issues resolved - period.
Stick with what you know, support what you build, and if people don't like your products - go build something else.
Sure, I'll bite.
This has been rehashed over and over again, but... Time Machine is not Volume Shadow Copy. See also here and here. See also this comment in this article.
One of the big problems I have with System Restore is that only certain key files are "backed up," and they're backed up as versioned, hidden files on the same volume. Although VSC attempts to be more comprehensive, it has the similar flaw of storing everything on the same volume. (The VSC solution also has the ability to store deltas, as block level changes, to a normally hidden part of the file system -- the shadow copy storage area.) My understanding is that the Microsoft-branded technologies rely on snapshots taken at periodic intervals (roughly once a day), and if you need a particular version of a particular file that falls in between a couple different snapshot intervals, you could be screwed. Time Machine is way more granular, providing comprehensive versioning (i.e., every revision that gets written to the file system is tracked) for each file, and on another volume, typically another drive. While there's been much talk about using external hard drives for Time Machine, Mac Pro users will no doubt use one of their many extra drive bays internal to their machines -- perfect since installation and removal is a snap.
Tracking every single revision makes it easier to track down where in time a particular file may have gotten corrupted or maliciously modified. It also becomes easier to then find a "last known good" version of a specific file, without having to pore over sets of snapshots.
Note that I'm only touching on a few small details here. But a Google search would easily enlighten you... or you could start with the links I've provided above.
Incidentally, Microsoft has a good resource explaining How Volume Shadow Copy Service Works.
Yes, I use OS X. I'm planning to migrate to Leopard. I also use a Windows machine somewhat less often.
.... but extends to routers, firewall, software, encryption)? I suppose you could remove a computer from a networked environment, but then, that only provides adds a layer of security. Most of us either don't have more than one computer, don't have time to read up on *the* latest computer security, and rely on reviews of OSes, software, routers, from "experts" to say what is best. I've been reading manuals on 'securing OS X' from reliable sources, but I've no idea how much extra security this provides.
.... but what do I recommend family and friends to do when fixing their computers? I don't think 'Use OS X Leopard thats secure" cuts it as this issue to begin with is really complicated. Identity theft is a huge problem. Home computer security I'm sure plays a role in this. As I at least know quite a bit about computers, etc. as at least an average computer geek, I should at least be able to take extra precautions.
That being said, I've taken some basic precautions with security. More than anything, I've not put financial information on my computer instead of relying on 'encryption' to take care of it.
Computer security though is a really specialized field. You need to know alot about hardware, networking, software, specific OSes, patches, encryption, math, etc etc. At that, and at best, because this falls under so many domains, no one can ever really know about everything about computer security.
As I'm sure a lot of people here are as well aren't security experts, the question becomes how much do we rely on the "security provider" (for this thread say Apple with OS X or MS with Windows
I'm not protecting Fort Knox here
Don't get me wrong, I still think OS X is better overall, because of its underlying architecture and a functional CLI, but the Classic Mac GUI had been honed incrementally over almost two decades before Steve just decided to bin the whole thing and reinvent the wheel. It was that interface which made the crappiness of OS 9 worth dealing with, despite the fact that you could hang the whole system by holding down the mouse button, and had to manually allocate memory, and everything else. It was the Mac's saving grace -- perhaps its only saving grace -- throughout the 'lean years' of the platform. And that's why a lot of users just never got over its elimination; it was, for many people, the only reason why they'd stuck around for so long.
There was no real reason to change it when the old codebase was dropped for NeXT's: even if none of the code needed to be kept, the interface guidelines that had evolved as best practices, arrived at by painstaking trial-and-error by generations of Mac programmers, could have been retained. What I think happened is that Steve Jobs wanted more eye candy, and wanted to make the entire desktop reflect the OS's "newness." It was a sales tactic, and although I don't think there's any debate that it worked, it was a pretty huge cost.
OS 9 was an operating system with a great GUI and a terrible backend; OS X had a great backend, but a GUI that was almost unusable at first, and which has only very recently come back on par with the Classic OS circa System 7.5 or so. (They just recently snuck the option-click-to-close-all-Finder-windows trick back in, which I believe originated on the IIgs, and was definitely missing for a while in early OS X versions...)
(Incidentally, the interface scizophrenia isn't limited just to the Mac OS; you also see this behavior in some of the major Apple apps [e.g. iTunes] -- every time there's a whole-number version increase, some part of the interface gets changed, apparently for the sake of changing it. It's as if they realize that some people won't believe that anything is different unless the widgets change, so they scramble everything around periodically, just to keep everyone on their toes.)
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Sorry to tell you this but you are mistaken. In most application (but not all) that modify files the program unlinks the file before writing the new file. Try it. However there are also many cases where this is not the case, an example is "cat > filename" or pico or bbedit. Those will write to single file and both hardlinks are changed.
Now you might say, well yes that is true but that's just the application doing that not the hardlink. True, but re-read my original question. How does the user know (or the OS know what the user meant) when he asks it to erase all version of a file? In the example I gave the result would change depending if a hardlink existed or not. The alternate answer is that the system would not be able to backtrack versions for any application that did that (which would be basically all major apps).
Some drink at the fountain of knowledge. Others just gargle.
Hmm... what movies did you have in mind? Primer, perhaps?
I don't care if it's 90,000 hectares. That lake was not my doing.
Data: I was tempted by the offer.
Jean-Luc Picard: How long a time?
Data: Zero point six eight seconds, sir.
Well, you're in luck... Time Machine is something you have to manually activate/configure before it will do anything.
I don't care if it's 90,000 hectares. That lake was not my doing.
I think you're confusing snprintf() behavior with that of strncpy(), which does as you say.
You're bluffing. Citation please.
But, see, most of the drivers in Mac OS X were not written by people at Apple, but were already written by the free software community. This includes, for example, the PCMCIA bus driver the poster was commenting on. Why don't they just include the full set of drivers from this project? They have nothing to lose, since they can say that only their own configurations are "supported" officially... But you'd be able to use more third party stuff without adding any drivers. And... In reality, these drivers are good quality. If they weren't, Apple wouldn't use them for their own purposes.
You seem to be living in some sort of dreamworld where Apple wrote these drivers. They didn't. They ripped most of them off of the efforts of free software volunteers. Yes, the idea with corporate backing is that if a driver starts malfunctioning, people can go to Apple to get support... But... Honestly, I doubt you can find me a case study of someone who did.
I've actually run into a fair share of kernel panics on Macs and I doubt Apple would care if I called them up and reported it. Does MS investigate every BSOD?
Sorry but you are utterly and completely mistaken. If you modify a hardlinked file in Tiger, it copies the file and modifies the copy.
You have to be a total idiot to post that after seeing from the other comment I replied to that is not the case!
Sitting right here in Tiger, if I edit a file, then create a hard link to that file, then edit the link - both files change, because there is ONE FILE!!!!
What would be the point of hard links if it were NOT a single file? You have zero clue what you are talking about. Try reading up on what a hard link even is before you spout nonsense. Or at least try it before embarrassing yourself so.
Sorry to be harsh, but you can't go around spreading complete fabrications about something so easy to test.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Emacs exhibits the behavior you note - because by default, it renames the file you are editing (to filename~) and then saves out a whole new version. Of course only the old version is still linked...
Thus as I said, the behaviour you are talking about has Zero to do with what hardlinks do, and everything to do with how applications treat files they are about to modify. If they simply modify the file in place of course the link behaves as expected, if they move the file without telling the user then blam - the APPLICATION has destroyed the link.
So TIGER doesn't behave at all as you say, it's only some applications. Again you fail to understand the very clue the name "hardLINK" is giving you, that you get a link to a file. Once you read the linked data and start modifyng the name and/or existance of that link, all bets are off as far as the user knowing what iis going on.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Care to send a screenshot with a clear, prominent warning of privacy implications inherent in enabling Time Machine?
Pay for software? Are you mad?
So the point is, that on tiger, when you hard link an document then edit it, the hardlink gets split to two files.
Since it doesn't happen with all applications, that's not the point at all. That's simply wrong, or at least application dependent. And again it has nothing to do with hard links and how they work.
The greater issue is how does Time Machine know what to delete when the users asks it to delete all version of an file?
Since I know how hard links work, there's no question at all. TM removes the file with that name as far back as it finds it. In later versions if it's really duplicated contents of a link destroyed by the application, then the new file is removed. If it is a hard link throughout, you simply remove that link and whatever other link remain persist (because, once more, there is only one file with a hard link).
There's no mystery to how TM works at all.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
On this one, what I expect to happen is TM to copy the individual links as new files, which would break the link but also preserve backup status. I agree it would be a bit of a bummer not to have it handle hard links by re-creating them, it's just kind of what I expect from a first version. I may be surprised though.
But, even if TM does understand and honor hard links, removing R/x9 and all previous revisions (basically just R/x8 as well) has no impact on any of the other file links you have in P or Q, old or new. Remember that removing one link of a multipy hard-linked file does not remove the file.
There is some question over TM re-creating hard links or not, but I think you're overthinking the trickiness of what happens with hard links and different revisions. As links move they stay links. When links get removed files remain until the link count goes to zero.
"There is more worth loving than we have strength to love." - Brian Jay Stanley