Mac OS X 10.4.2 update to patch dozens of Tiger glitches

AppleInsider

The second update to Apple's Tiger operating system will continue to focus on stability and reliability, delivering dozens of bug fixes to the month-old operating system.



The first pre-released builds of the software -- Mac OS X 10.4.2 -- began to surface on the internet over the Memorial Day weekend, according to several published reports. The most prominent build appears to be Mac OS X 10.4.2 build 8C21, which reportedly weighs in at just over 43MB in Combo Update form.



Along with the software, Apple is rumored to have asked developers to focus their testing efforts on nearly four-dozen components of the Tiger system. Some of these areas include: Automator, AddressBook, AppleScript, Dashboard Widgets, Directory Services, Display Preferences, File Manager, Graphics, iCal, iWork, JavaScript, Navigation Services dialog boxes, Printing, Safari, Spotlight, Sync Services, and iChat video conferencing.



Mac OS X 10.4.2 also appears to place an emphasis on improving Tiger's Mail application and Networking capabilities. Apple is said to have fixed glitches between Mail and Automator and further asked developers to test Mail scripting, attachments and iMAP-based e-mail functions.



On the subject of networking, Mac OS X 10.4.2 is expected to deliver enhancements to general system networking, NFS, NSL DS, Net SNMP, AirPort and ethernet networking. The release should also pack a fix to iChat 3.0, which has presented some users with insufficient bandwidth issues.



Already sources say the current build of Mac OS X 10.4.2 includes well over two-dozen new bug fixes to Tiger, in addition to those delivered in the mid-May release of Mac OS X 10.4.1. Some of these fixes pertain to new user accounts not showing up properly in the system, issues with smartcards not functioning in certain situations, and blank media failing to mount.



Apple is also reportedly working on a fix for users who saw their modems crippled by the release of Tiger. Meanwhile, several additional fixes target .Mac notification, iDisk functionality, Weather widget crashes, and video conferencing frame rates.



It's believed that Tiger users will be able to apply Mac OS X 10.4.2 to either Mac OS X 10.4 or Mac OS X 10.4.1.



Sources have previously stated that maintenance release is slated for a released within the next 45 days.



Apple last updated Mac OS X 10.4 Tiger on May 16th with the release of Mac OS X 10.4.1 -- just two weeks after Tiger went on sale.

[ View this article at AppleInsider.com ]

melgross

Here we go again.

buonrotto

Are you going to do this for every point release? One of us is going to be very tired when it's all said and done.

zenatek

YAY, iChat is getting updated!

melgross

Quote:

Originally posted by BuonRotto

Are you going to do this for every point release? One of us is going to be very tired when it's all said and done.

Assuming you're talking to me, sure, if they come out every few weeks.

nagromme

Welcome to reality, for ANY major OS version from ANY company.



No matter how big the internal test program, the public is bigger still. And no OS is ever perfect or ever will be, yet it must ship at some point.



So the updates will ALWAYS be more frequent when a new version is young. Knowing that, you may choose to wait for a couple of them before installing. You do have the choice

louzer

Quote:

Originally posted by nagromme

Welcome to reality, for ANY major OS version from ANY company.



No matter how big the internal test program, the public is bigger still. And no OS is ever perfect or ever will be, yet it must ship at some point.



So the updates will ALWAYS be more frequent when a new version is young. Knowing that, you may choose to wait for a couple of them before installing. You do have the choice

Excuses, excuses. The point is that there's some MAJOR bugs in the OS at this moment, bugs that should never have been let out in the wild. I'm not talking mild iChat issues like icons not appearing in some cases. But iChat not working at all for people? That's supposed to be acceptable? Errors with sparse disk images that corrupts them? How does this get out (esp since it was really easy to find: Create one, copy data to it, eject, hey, its corrupt!). And this on top of the MAJOR problems in 10.3 with file vault and sparse images. Spotlight seems to have more problems then anything it actually helps. Large groups of people can't get spotlight to find files with some criteria, but can with other. And if you turn it off for a couple of directories because they're chock full of files you don't need/want indexing and don't want it to suck your processor usage, you can't even find files in those places at all. And apparently Apple couldn't even figure out how to do spotlight searching on large databases, because they've changed their database designs (Mail, address book, etc) to a "file per element" model, which is truly inefficient for any use (except backups).



I guess if you're one of the people not affected, its fine. You can spout your "its never perfect" spiel (which, by the way, you people never say about windows, its always "Its so full of bugs and security lapses, you're risking your life just turning on your computer!")



But if you hit these bugs, then its basically more of a "screw you dude, that's what you get for using our crappy OS. But thanks for the bug report, we might fix it when we get to it". And I'm sure the people who lost whole sets of data because 10.3's firewire issues corrupted their disks, its their fault for not backing up regularly too. How long do you people make excuses for Apple before realizing there's a long and downward trend by Apple here to make sure things ship, not whether they're actually tested or reliable (Bulging capicitators on the iMac G5, anyone? iBooks which still have video problems? Fan issues with the PowerMac G5s AND the iMac G5?)

nebula

Any speculation as to whether Apple will enable Quartz 2D Extreme?

joealamaiz

Im still not able to use managed accounts over 10.3.9... when will they ever fix this? Am I missing an update?

melgross

Quote:

Originally posted by Louzer

Excuses, excuses. The point is that there's some MAJOR bugs in the OS at this moment, bugs that should never have been let out in the wild. I'm not talking mild iChat issues like icons not appearing in some cases. But iChat not working at all for people? That's supposed to be acceptable? Errors with sparse disk images that corrupts them? How does this get out (esp since it was really easy to find: Create one, copy data to it, eject, hey, its corrupt!). And this on top of the MAJOR problems in 10.3 with file vault and sparse images. Spotlight seems to have more problems then anything it actually helps. Large groups of people can't get spotlight to find files with some criteria, but can with other. And if you turn it off for a couple of directories because they're chock full of files you don't need/want indexing and don't want it to suck your processor usage, you can't even find files in those places at all. And apparently Apple couldn't even figure out how to do spotlight searching on large databases, because they've changed their database designs (Mail, address book, etc) to a "file per element" model, which is truly inefficient for any use (except backups).



I guess if you're one of the people not affected, its fine. You can spout your "its never perfect" spiel (which, by the way, you people never say about windows, its always "Its so full of bugs and security lapses, you're risking your life just turning on your computer!")



But if you hit these bugs, then its basically more of a "screw you dude, that's what you get for using our crappy OS. But thanks for the bug report, we might fix it when we get to it". And I'm sure the people who lost whole sets of data because 10.3's firewire issues corrupted their disks, its their fault for not backing up regularly too. How long do you people make excuses for Apple before realizing there's a long and downward trend by Apple here to make sure things ship, not whether they're actually tested or reliable (Bulging capicitators on the iMac G5, anyone? iBooks which still have video problems? Fan issues with the PowerMac G5s AND the iMac G5?)

This has been my point.



Apple fixes two dozen bugs two weeks after release. Several weeks later they fix, what, FOUR dozen more?



Most of these bugs were known before release. To tell us that we should wait until it becomes stable before we upgrade is not an answer. I've upgraded two machines here at home. Neither machine has much third party software, and neither has third party pref panes. But both machines are exhibiting problems from the release. From Apple's own software.



I don't believe that most people who have upgraded have not had any problems. In my usergroup, most have said that they have some trouble, some have had major problems. It's inexcusable.



We read the heading here that people are switching for reasons of security and STABILITY. Well, this upgrade isn't stable.



To say that those of us ( a majority here in these forums) are whining is wrong. Those who are whining are those who would have installed an Alpha release if they could have gotten it. And they would say how stable it was as well.

booga

This looks like the first fully-functional version of MacOS X 10.4. Releasing an operating system with SEVERAL common data corruption scenarios, videoconference that doesn't work for large portions of the population (and disabling the ability to downgrade iChat,) and tons of security issues is just unacceptable.



Of course, for this unacceptable behavior, Apple was rewarded in the short term with huge sales. And the long term, where people, especially recent converts, are driven away from the platform due to low quality and bad implementation doesn't seem to concern them (yet.)

jms698

Quote:

Originally posted by nagromme

Welcome to reality, for ANY major OS version from ANY company.



No matter how big the internal test program, the public is bigger still. And no OS is ever perfect or ever will be, yet it must ship at some point.



So the updates will ALWAYS be more frequent when a new version is young. Knowing that, you may choose to wait for a couple of them before installing. You do have the choice

Well said (could we make this sticky?)

melgross

Quote:

Originally posted by jms698

Well said (could we make this sticky?)

I wasn't well said, because it isn't true.



Apple DECIDES how their testing programs work, and how widely spread they are. Because of that, it's Apple's fault if that program fails to catch serious bugs, or if it does, Apple choses to ignore them.



We think of MS as being arrogant, but Apple is known for that as well. The more successful they get the worse it gets.



As the platform moves away from it's core base of stalwart Mac users (such as myself) who live with these problems, and have even minimized them in the past, newcomers to the platform might not be as forgiving. This is not the time for Apple to be carefree about these problems.



I remember, from a long time ago-in an industry publication, the difference between Sony's concept of quality, and Zenith's.



Zenith said that they had a strong warranty. If it broke, they would fix it.



Sony said that they made sure that it didn't break.



I prefer the Sony approach.

somynona

Is there any news regarding whether or not Quartz 2D Extreme will be enabled by default?

melgross

Quote:

Originally posted by Somynona

Is there any news regarding whether or not Quartz 2D Extreme will be enabled by default?

I was told that it would be included in the first update, if they had enough time to work on it, but they didn't. With all of these bugs to squash, they might not get to it now either. I haven't had a chance to speak to anyone about it yet.

blindingforce

I've had a fraction of the problems most users have had but due to the fact I'm about to finish up a comic going to press in 2 weeks or so I went back to Panther. My photoshop 7 and Illustrator 10 wouldn't work at the same time and Illustrator left nothing but temp files in the folders of the files I worked on.



I'm still having my iPod issues and at this point it's a useless piece of hardware because it locks up even when I do a complete restore. The new iTunes and it's sync options are garbage. Tekserve couldn't even figure it out.



I do love little things like the new preview options and burn folders so until I upgrade later in the year, again, I'll miss those. But it's just not as stable as I need it. And due to the fact Adobe likes to screw people over in not doing their own updates...just large ones you have to buy...well I don't have the money to deal with the BS.



The other thing is Automator is great for native programs but that's about it. That does me no good at all. So until Adobe and or other developers like Corel, and etc. make actual functioning actions it'll do nothing but collect dust on my machine. Why can't I use it to batch open a program, open a file, export a file with said options, and save. Not all of us are programmers and Adobe couldn't be bothered making Illustrator actions...so Automator is worthless to me personally.



Tiger is great but if dragging a file from mail onto the desktop makes Finder do nothing but an infinite loop of restarts I'll just wait a bit longer.

louzer

Quote:

Originally posted by melgross

I was told that it would be included in the first update, if they had enough time to work on it, but they didn't. With all of these bugs to squash, they might not get to it now either. I haven't had a chance to speak to anyone about it yet.

Well, my assumption is that its disabled because ITS (or is it IT'S, I can never remember, but, then again, I have so much trouble getting the whole "which is left and which is right" thing memorized too, so its not too surprising) buggy as well, so we'll probably have to wait for (a) they fix the bugs in the stuff that's enabled, then (b) they fix the bugs in the stuff that's turned off, then (c) wait for them to turn it on.



And then we get to wait for all the 10.4 users (or beta-testers, depending on your point of view) to complain about all the problems with it and get those items fixed.

melgross

Quote:

Originally posted by Louzer

Well, my assumption is that its disabled because ITS (or is it IT'S, I can never remember, but, then again, I have so much trouble getting the whole "which is left and which is right" thing memorized too, so its not too surprising) buggy as well, so we'll probably have to wait for (a) they fix the bugs in the stuff that's enabled, then (b) they fix the bugs in the stuff that's turned off, then (c) wait for them to turn it on.



And then we get to wait for all the 10.4 users (or beta-testers, depending on your point of view) to complain about all the problems with it and get those items fixed.

Yes, that was my implication. It is buggy. Those who have turned it on (only from Terminal) have found problems, depending on what setup they have. By the way, if you rush out to turn it on in Terminal, it turns off again upon a restart (or start).



I argue with some editors about the its it's its' thing as well. The " ' " implies possesive. Most times there's no problem with using the " ' ", but some editors think that it isn't "modern".

mcdawson

The reality is that there's a balance in deciding when to release a major OS--you have to balance the veriication time vs how much $$ you're burning every day you're not selling it vs how bad the "known" bugs appear to be vs the risk of introduction NEW bugs. To fully verify a major OS release probably takes about 3 weeks--one week of internal testing, plus 2 weeks of release to outsiders (who aren't forced to test right away and who answer questions less quickly). You also have to realize that the operation cost of any division does NOT have the same value to the group as to the corporation--i.e., say it costs Apple's sw divison $100K per DAY for its engineerings, managers, etc. While that may be a small # to Apple as a whole, that's a BIG # to that division. Waiting 21 days for a full verification for that the last build in the above senerio would be $2.1 million. There could also be marketing costs and production costs (holding a spot at the DVD duplicator) for missing a deadline. Then you add in the risk of introducing a new (bad) bug. THAT is a scary one for executives--something that hits everyone. So, as project winds down, you make the bar higher and higher to allow bugs to be fixed--the bugs that you know are better than a really nasty one that you don't. These bugs are more likely to appear at the end of project as your engineers are tired (they've been working long hours the last few months). So you focus only on the "showstoppers"--bugs that would cause data loss, create buggy reputations, etc. Every week the definition of "showstopper" becomes harder to make. Additionally, while every bug IS probably known, the risk analysis over how many/what % of people are likely to see it doesn't always hold true (testers and developers don't always do things exactly like mom & pop do).

You also know that it takes 3-4 weeks for the sw to get through duplication and distribution, so you've got a window. What you do is slowly stop teams working on this release, and start working on the update. The idea is that starting NOW gives you 5-6 weeks of development fixing the bugs, whlie the paying public only has used the OS (if they bought it the 1st day) a few weeks. Because you don't have distribution times (internet download vs disk duplication and packaging), you gain even more time.

So, while it would be great if more time could be given (most engineers argue for it), its rarely "if only they had worked on it another week". The risk of new, nasty, bugs is just too great, so any major company has to "wind down" the project only working on those bugs that it preceives are the worst. The cost of waiting around to see if bugs are fully fixed is very costly, too.

Because of the winding down issue and production issues, while 10.4.2 may reach consumers mid-June, that same level of bug fixing at the start would have meant that 10.4 couldn't have gone on sale until Aug. And it still wouldn't likely be as good as 10.4.2 will be because what the company thought the biggest priority bugs wouldn't necessarily have been what the consumers think, while 10.4.2 WILL benefit from feedback.

At some point, some executive decides that any particular bug isn't worth $1 million to fix. So they ship it and fix it later. I believe that Apple has been and is good in at least the "fix it later" part--many other sw companies rarely ship fixes, or bundle them with new features and charge $$ for those updates.

melgross

Quote:

Originally posted by mcdawson

The reality is that there's a balance in deciding when to release a major OS--you have to balance the veriication time vs how much $$ you're burning every day you're not selling it vs how bad the "known" bugs appear to be vs the risk of introduction NEW bugs. To fully verify a major OS release probably takes about 3 weeks--one week of internal testing, plus 2 weeks of release to outsiders (who aren't forced to test right away and who answer questions less quickly). You also have to realize that the operation cost of any division does NOT have the same value to the group as to the corporation--i.e., say it costs Apple's sw divison $100K per DAY for its engineerings, managers, etc. While that may be a small # to Apple as a whole, that's a BIG # to that division. Waiting 21 days for a full verification for that the last build in the above senerio would be $2.1 million. There could also be marketing costs and production costs (holding a spot at the DVD duplicator) for missing a deadline. Then you add in the risk of introducing a new (bad) bug. THAT is a scary one for executives--something that hits everyone. So, as project winds down, you make the bar higher and higher to allow bugs to be fixed--the bugs that you know are better than a really nasty one that you don't. These bugs are more likely to appear at the end of project as your engineers are tired (they've been working long hours the last few months). So you focus only on the "showstoppers"--bugs that would cause data loss, create buggy reputations, etc. Every week the definition of "showstopper" becomes harder to make. Additionally, while every bug IS probably known, the risk analysis over how many/what % of people are likely to see it doesn't always hold true (testers and developers don't always do things exactly like mom & pop do).

You also know that it takes 3-4 weeks for the sw to get through duplication and distribution, so you've got a window. What you do is slowly stop teams working on this release, and start working on the update. The idea is that starting NOW gives you 5-6 weeks of development fixing the bugs, whlie the paying public only has used the OS (if they bought it the 1st day) a few weeks. Because you don't have distribution times (internet download vs disk duplication and packaging), you gain even more time.

So, while it would be great if more time could be given (most engineers argue for it), its rarely "if only they had worked on it another week". The risk of new, nasty, bugs is just too great, so any major company has to "wind down" the project only working on those bugs that it preceives are the worst. The cost of waiting around to see if bugs are fully fixed is very costly, too.

Because of the winding down issue and production issues, while 10.4.2 may reach consumers mid-June, that same level of bug fixing at the start would have meant that 10.4 couldn't have gone on sale until Aug. And it still wouldn't likely be as good as 10.4.2 will be because what the company thought the biggest priority bugs wouldn't necessarily have been what the consumers think, while 10.4.2 WILL benefit from feedback.

At some point, some executive decides that any particular bug isn't worth $1 million to fix. So they ship it and fix it later. I believe that Apple has been and is good in at least the "fix it later" part--many other sw companies rarely ship fixes, or bundle them with new features and charge $$ for those updates.

Yes, yes, yes. We've heard all of this before. And while some of what you say is correct, most isn't, and it simply isn't relevant.



You don't understand that to add updates is a much more time consuming, and costly way of doing things than fixing them in the initial release.



Essentially, it's a political decision to release or not. In Apple's case, from everything that I've been reading, they wanted to book the sales this quarter, because they estimated a somewhat flat quarter, and the market pounded them for it (I know, I've got a fair amount of stock!). They apparently figure that the July-September quarter is going to be good enough that they won't need the lift then, so they took it now.



If Apple was a software house, I could possibly understand that, though it still wouldn't be proper. But OS sales are a very small component of Apple's sales, though it does contribute disproportionately to profits.



And that's the problem. In order to make things look better in the short term they were prepared to allow massive bugs through.

mcdawson

Quote:

Originally posted by melgross

You don't understand that to add updates is a much more time consuming, and costly way of doing things than fixing them in the initial release.

What believe is that it is NOT more time consuming and costly way of doing things. If you subscribe to the "don't allow all fixes in at the last momement due to the risk" theory, then as you get closer and closer to a release, more and more of your engineering team is idle. However, if you have a concurrent update running, those people simply stop working on the "current" release and start working on the update. Thus you are NOT losing engineering time. And given that the updates are done via the internet, your first release + update is MUCH, MUCH quicker than a release, because you have to wait for production/disc duplication (3-5 weeks at least!). During that "dead" time until the product is in the user's hand, you can be working. In theory, you could have an update ready on the day the user gets the box, thus delivering MUCH more quality in the same time frame than just a release.

Quote:

Essentially, it's a political decision to release or not

It ALWAYS is this. NO major software can be released bug free, so its always some kind of cost/# bugs left tradeoff. Those decisions are made by executives based on their reality. That's their job.

Quote:

In Apple's case, from everything that I've been reading, they wanted to book the sales this quarter, because they estimated a somewhat flat quarter, and the market pounded them for it (I know, I've got a fair amount of stock!). They apparently figure that the July-September quarter is going to be good enough that they won't need the lift then, so they took it now.



If Apple was a software house, I could possibly understand that, though it still wouldn't be proper. But OS sales are a very small component of Apple's sales, though it does contribute disproportionately to profits.



And that's the problem. In order to make things look better in the short term they were prepared to allow massive bugs through.

Short term gains over long term gains? That's an issue with all public companies. The quesiton: do shareholders (those who aren't also Tiger owners) CARE about the bug list? Are the amount of Tiger bugs so bad that analysts are reducing their ratings; are the Tiger bugs so bad that sales are slipping due to negative feedback? Is it so bad that people will remember and NOT flock to buy 10.5.0? Are the service costs (telephone, mainly) costs eating up all the profits? Another 3 months would have done wonders to the initial quality, but my guess that 10.4.2 would probably be of better quality than waiting 3 more months (due to my arguments above).

A car manufacturer, due to liability, needs to make the care safe. They need to make their cares as "bug free" at sale because those numbers are tracked (via JD Powers, etc) and used as buying tools. That isn't the case in software (unless the sw is so bad that a reviewer says "don't buy until update #1"). There really is no business incentive to wait 3 months, especially if what you are saying is true. "Good enough" for most people, and fixing the bugs later works until they either severly misjudge the risks or there is some "bug" competition like there is in cars?