Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Security Update 2005-007 v1.1 Released

shamino said:
64-bit application code is not something most people need.

It is needed for apps that require more than 4G of memory (64-bit memory pointers), and it is very useful for apps that do a lot of 64-bit integer arithmetic (scientific applications, simulations, maybe some games.)
Click to expand...
To be clear.... to do 64 bit integer arithmetic (internally) an application does NOT need to request 64 bit virtual addressing.

Mac OS X as of 10.2.8 (and matching xcode) has supported the use of 64 bit integer / general arithmetic when running on a G5. As of 10.4 Apple added support for applications to request a 64 bit virtual address space, this later aspect was broken by mistakenly not including the 64 bit version of libSystem in the first security update.

(possibly because of the update system not being configured to correctly deliver the package containing 64 bit libSystem to folks that have G5s)
 
Apple should just release separate security fixes instead of an all in one roll-up.

Over three dozen updates in this release! Could save IT people all sorts of headaches in the future.
 
Problem with the Security Update update

Hi,

Seems like I'm the only poster to have this problem - since I updated to the updated update, I have had big problems completely loading certain pages.

Like Apple UK home page, also Apple Support pages... (also Internet Banking - HSBC)..

The loading bar will get a third/half way (in address bar) there, but no further. Bizarrely, no beachball. When I tried the HSBC (log-out page), I checked Activity window and it seemed like it was mostly GIFs and JPEGs that werent loading (not sure if this explains the problem at hand)..

I have repaired permissions and restarted, but no joy. Also, this is happening with both Safari and Firefox..

Any help most appreciated, especially as I can't load Apple's Support page!

Thanks,

Darren
 
jicon said:
Apple should just release separate security fixes instead of an all in one roll-up.
Click to expand...
That could in theory let people install only the updates that might affect them (especially useful over slow dial-up connections), depending on which system services they use, but I think most people would prefer not to have to read the details and decide which they need. In fact, I don't think most Mac users could do that. For the most part, people know which applications they use but don't know which internal system services/frameworks/features they depend on.
 
NEED SERIOUS HELP

Now I restarted my computer and I logged back in to the blue default screen and all of my music and pictures are missing, yet my movies are still there. My HD says that the space that my mp3's (40+ GB) exsisted in, is still taken up, so I don't think they are lost. I can deal with all my programs being defualt, but not losing data. Can anyone help me out with this? I'm really stressing because I didn't back up anything. All my apps, documents, and movies are still there but mp3's and pictures are not. I don't know even the first thing to do.
 
re-rip

nashstradamus said:
....but mp3's and pictures are not. I don't know even the first thing to do.
Click to expand...
The MP3's can be re-ripped from your CD's - which gives you an opportunity to even re-rip in a more modern, higher quality, format.

It's a pain, but at least you have the "backups" in the CD's that you originally bought.
 
It should not have affected anyone.

coolfactor said:
Hey SiliconAddict, please explain to us how this oversight affected you. Millions upon millions of Mac users were not affected by this problem, only a relatively small percentage. Apple engineers are working day and night to bring us the best OS, but I guess they're not as perfect as you are.
Click to expand...

I should start by saying I use both MacOSX and WinXP day to day and like/hate them both equally for the most part.

The truth of the matter is, Apple is getting sloppy.

By saying millions upon millions of mac users were not affected by this you are avoiding the issue. I mean when Microsoft screws up - often times 90%+ of people are never affected - and guess what 90% of all MS users is a LOT more people than even all of the Apple users.

The point is that Apple is dealing with a much more closed - controlled system than Microsoft. There is no reason that Apple should be releasing updates that break something as obvious as 64bit processing - or - keyboards or ...whatever. Apple has always had a HUGE advantage over Microsoft - all the systems are defined - there are only a handful video cards, audio cards, etc.... Compare that to Microsoft that has to deal with literally hundreds of video cards etc....

It's not like this Apple update only affected one piece of third party software - sure they were the ones to report it - but it affected ALL 64 bit programs.

I love Apple products - but come on people - they are not incapable of mistakes - AND - they have been getting slightly worse. It is a disturbing trend that can easily be reversed but one that we should not stick our heads in the sand over.
 
AidenShaw said:
The MP3's can be re-ripped from your CD's - which gives you an opportunity to even re-rip in a more modern, higher quality, format.

It's a pain, but at least you have the "backups" in the CD's that you originally bought.
Click to expand...

But I assume his photos aren't on CDs, and if he downloaded any music on iTunes that would be completely lost.

If the amount of space indicates that your photos and mp3s are still somewhere on the hard drive, there hopefully is a way to recover them. I'm not an expert so I can't offer any advice. I wouldn't start reimporting the CDs right away though, I'd be too afraid that it might write over the data you're missing.

Maybe you can get some help in this forum.

https://forums.macrumors.com/forums/78/

Good luck!
 
Hmm, not sure if it was the first or second update, but it seemed to have messed with some of my web javascript. I use Firefox pretty much exclusively, and as of today when I go to http://www.yesasia.com and click on the javascript button to endlarge an image, to see a full sized DVD package it doesn't work, though it always has before, and even made sure to add it to my allowed sites. Then i tried it in Safari, and the same thing happens, though it does work in IE (I can't believe I even still have IE around). And strangely other sites with javascript popups still work, like http://www.fivestarlaser.com. Not sure what is going on here, but it is really F***ing annoying. I have been adding my DVD library to Delicious Library, and since their is no Amazon Hong Kong or Korea, I need to manually add all the information, and now I can't even see the full sized labels from the sites without using the completly wonky IE.
 
jonahlee said:
Hmm, not sure if it was the first or second update, but it seemed to have messed with some of my web javascript.
Click to expand...

I went to Yesasia and I can confirm the same thing happens to me with Safari there. :( However, in Deer Park, The pop-up opens in two background tabs... one that says the image is loading and another with the image. :rolleyes

I have not done the second update, FWIW.

And *sobs* Tampopo is available on DVD for $50... I wish I could convince myself it was worth it. That's such a classic movie. *sobs*
 
how could anyone rate this as positive?

Just to counter the "how could anyone rate this as negative?" posts - how could anyone rate a story about:
- Apple breaking a highly-visible feature in a service pack
- Apple showing that their QA system has huge flaws
- Apple rushing a re-release of the service pack

as positive?
 
AidenShaw said:
Just to counter the "how could anyone rate this as negative posts" - how could anyone rate a story about:
- Apple breaking a highly-visible feature in a service pack
- Apple showing that their QA system has huge flaws
- Apple rushing a re-release of the service pack

as positive?
Click to expand...

Easy - when Apple's updates start getting sloppy it means that the bulk of Apple's engineering staff have moved on to the next version of the operating system showing that Xi work is going full speed ahead.

OK I'm stretching here but Apple's updates do get buggier the closer you get to a major new release
 
ROTFLOL !!

trrosen said:
Easy - when Apple's updates start getting sloppy it means that the bulk of Apple's engineering staff have moved on to the next version of the operating system showing that Xi work is going full speed ahead.

OK I'm stretching here but Apple's updates do get buggier the closer you get to a major new release
Click to expand...
Thanks for the genuine laugh - definitely "looking for the silver lining" in this cloud :)

Has a 10.5 schedule been released? Didn't Apple say that it would slow down major releases - that one a year was too many?

Maybe all the engineers are porting 10.4 to Intel.... (And adding Blu-Ray support so that all those fat binaries will fit on one disc.)
 
AidenShaw said:
Thanks for the genuine laugh - definitely "looking for the silver lining" in this cloud :)

Has a 10.5 schedule been released? Didn't Apple say that it would slow down major releases - that one a year was too many?

Maybe all the engineers are porting 10.4 to Intel.... (And adding Blu-Ray support so that all those fat binaries will fit on one disc.)
Click to expand...

Ugh. As you mentioned earlier - this isn't about the development group and what they are doing. It's about the build group, and QA and what they are not doing. As a developer on an application that gets FDA'd, HIPPA'd and QA'd to death, I can tell you that SNAFU's like this do occasionally happen on service releases because the full QA usually be somewhat circumnavigated when you are delivering a patch instead of a full release. It is likely that the library in question was delivered several times to QA in 64 bit form, but a problem with the "official" build machine caused the final delivered build to be built 32 bit only.

In other words, they could know the source is 64 bit clean, but their procedures did not guarantee that the official build was generated correctly. It also shows that they may be a little too trusting in their build process in that they never thoroughly verified the final release build before putting it out in the wild. This still does not cast anything in glowing terms, but it isn't as bad as "we've never actually tested 64 bits in our regression suites" as some people have mentioned, and which is patently ridiculous.
 
panther problems...?

What about Panther users who have experienced a multitude of sudden crashes since this update? WTF?!
 
shawnce said:
To be clear.... to do 64 bit integer arithmetic (internally) an application does NOT need to request 64 bit virtual addressing.
Click to expand...
Addressing and arithmetic are two completely separate issues.

Where did I say one requires the other?

And yes, I know you can do 64-bit artithmetic without 64-bit registers. It means that every math operation must be a (hopefully inline) function call, instead of a single machine instruction. This will impose a severe performance penalty if your application is doing a lot of 64-bit math.
 
dernhelm said:
In other words, they could know the source is 64 bit clean, but their procedures did not guarantee that the official build was generated correctly. It also shows that they may be a little too trusting in their build process in that they never thoroughly verified the final release build before putting it out in the wild. This still does not cast anything in glowing terms, but it isn't as bad as "we've never actually tested 64 bits in our regression suites" as some people have mentioned, and which is patently ridiculous.
Click to expand...
Please re-read what you claim to have read.

Nobody said that they never tested any 64-bit code.

What I (and others) wrote was that either the final release version of the patch was never run through a regression testing suite or that suite has no 64-bit tests. I don't know which, but in either case, it shows somebody (maybe several somebodies) getting careless and sloppy.

If you can come up with a situation where neither of these statements are true, and still managed to produce that broken patch, please explain how.
 
AidenShaw said:
Apple want you to spend $129 to fix Panther.... This is their subtle reminder.
Click to expand...

My wife hasn't experienced any problems on her 15" PB running Tiger. The update twice went smoothly. So I'm wondering are there actual those with Panther having problems, I'm using Panther?
 
mkrishnan said:
I went to Yesasia and I can confirm the same thing happens to me with Safari there. :( However, in Deer Park, The pop-up opens in two background tabs... one that says the image is loading and another with the image. :rolleyes

I have not done the second update, FWIW.

And *sobs* Tampopo is available on DVD for $50... I wish I could convince myself it was worth it. That's such a classic movie. *sobs*
Click to expand...

Thanks for confirming that problem, so it was the first security fix that did it, and stayed with the second, this is completely annoying and unnaceptable.
 
shamino said:
Addressing and arithmetic are two completely separate issues.

Where did I say one requires the other?
Click to expand...
It seemed implied in your first message and possibly again in the below.

shamino said:
And yes, I know you can do 64-bit artithmetic without 64-bit registers. It means that every math operation must be a (hopefully inline) function call, instead of a single machine instruction. This will impose a severe performance penalty if your application is doing a lot of 64-bit math.
Click to expand...
My point was that you can do 64 bit math with 64 bit registers using 64 math instructions PPC without requiring 64 bit virtual addressing. In other words an application compiled to request 32 bit virtual addressing can use 64 bit math operations supported by the G5 using 64 bit registers, etc.
 
same for the P4....

shawnce said:
My point was that you can do 64 bit math with 64 bit registers using 64 math instructions PPC without requiring 64 bit virtual addressing. In other words an application compiled to request 32 bit virtual addressing can use 64 bit math operations supported by the G5 using 64 bit registers, etc.
Click to expand...
Note that the 32-bit Pentium 4 can do 64-bit and 128-bit integer arithmetic in hardware using SSE2/SSE3 instructions. The G4 and G3 cannot do 64-bit integer in hardware, but compilers will silently generate the additional code to do it for you.

Also note that virtually every Mac or PC sold in the last 10-15 years can do 64-bit double precision floating point in the hardware (or FPU). This includes the 68K, 80x86, Pentium, 60x, G3, G4,....
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back