Welcome to the MacNN Forums.

kiskynet · May 2, 2003, 01:35 PM

I'd never been able to get this hidden feature before (maybe it was left over from os 9)... but try this... update to 10.2.5 then plug or unplug a usb mouse or anything else a couple o times... works for me sometimes.... i'd saved all the work so didn't get the whole retro experience but was still kinda fun....

Adam Betts · May 2, 2003, 01:57 PM

You think it's an easter egg?

SMacTech · May 2, 2003, 02:44 PM

I tried this with my M$ intellimouse, Handspring cradle, Epson 870 printer, Epson scanner and the USB hub in my monitor. No Easter Eggs left over from last month. YMMV. do you have a Belkin hub?

dampeoples · May 2, 2003, 03:42 PM

I'm confused...where is the easter egg?

brainchild2b · May 2, 2003, 04:49 PM

It's just you dude. That and mostly likely some crappy usb device from a crappy company who can't program their crappy drivers worth crap.

I can unplug my mouse a billion time, the printer at the same time, the handspring can be syncing....

somebody crapped on you, but it ain't apple.

juanvaldes · May 2, 2003, 06:51 PM

Originally posted by dampeoples:
I'm confused...where is the easter egg?

something that is not supposed to be there. Jokes and whatnot that programmers put in that no one catches.

But, a KP is NOT a easter egg.

CharlesS · May 2, 2003, 07:07 PM

Only MS would consider a crash to be a hidden feature.

SupahCoolX · May 2, 2003, 07:23 PM

Originally posted by CharlesS:
Only MS would consider a crash to be a hidden feature.

Except that MS doesn't hide them very well

Kenneth · May 2, 2003, 07:25 PM

Funny.. Easter was passed weeks ago...
You need some better USB devices.

OpenStep · May 2, 2003, 08:17 PM

dampeoples · May 2, 2003, 09:10 PM

Originally posted by juanvaldes:
something that is not supposed to be there. Jokes and whatnot that programmers put in that no one catches.

But, a KP is NOT a easter egg.

OK, that clears things up, heh

Colonel_Panic · May 2, 2003, 11:50 PM

kiskynet,

Your either a OS 9 switcher or a windows user huh? Well here ya go:

A kernel panic is a faliure of the system at a low level. IE Windows Blue screen of death, OS 9 full system freeze.

Those are other examples of things LIKE Kernel panics. They are not the same though.

A kernel is the base of your system, Your talk to your apps your apps talk to your kernel thats basicly how it goes, it your app screws up a word or slurs its speech you get a kernel panic. (those are about the dullest terms I can put it in, Please hardcore *nix guys that was not ment to be a true def of a kernel panic)

on a side note if Apple can tell the kernel to bring up that nice screen and load the fade screen animation there has to be some way for them down the road to make OS X recover from a KP.

Cipher13 · May 3, 2003, 12:29 AM

Originally posted by Colonel_Panic:
on a side note if Apple can tell the kernel to bring up that nice screen and load the fade screen animation there has to be some way for them down the road to make OS X recover from a KP.

That doesn't always happen, though. Sometimes you don't get the polite, pretty KP screens, and just get the dump...

Colonel_Panic · May 3, 2003, 12:42 AM

That doesn't always happen, though. Sometimes you don't get the polite, pretty KP screens, and just get the dump...

Yeah but when it does that means apple is calling up other commands after the Kernal has already crashed - if they could write some kind of program or something to recover from it, even if it means all my applications quit and what not..

Thats still better than rebooting

CharlesS · May 3, 2003, 02:13 AM

Originally posted by Colonel_Panic:
Yeah but when it does that means apple is calling up other commands after the Kernal has already crashed - if they could write some kind of program or something to recover from it, even if it means all my applications quit and what not..

Thats still better than rebooting

UNIX has been around forever. If there was a way to recover from a KP, I think someone would have done it by now.

Besides, doing so would likely put your machine in an unstable state anyway (kind of like Norton CrashGuard did on OS 8 or like the Windows Blue Screen does when it recovers). Better just to make the KP screen appear as rarely as possible.

geekwagon · May 3, 2003, 02:27 AM

Originally posted by Colonel_Panic:
Yeah but when it does that means apple is calling up other commands after the Kernal has already crashed - if they could write some kind of program or something to recover from it, even if it means all my applications quit and what not..

Thats still better than rebooting

There is a very good reason why they always make you reboot after a kernel panic. Most panics are caused by the kernel hitting an error so bad that the programmer felt that recovering from that error reliably was impossible. So, since the environment the kernel runs in is now unknown, it just bails to minimize the amount of corruption to your data.

Same thing with parity errors with bad RAM in the old days. Basically, if the computer had a parity error it would just stop, as that was considered less likely to corrupt your data then trying to actually use RAM that is corrupting your data.

Colonel_Panic · May 3, 2003, 03:28 AM

ah your right, I forgot about corruption and things like that.

boma · May 3, 2003, 04:25 AM

I don�t know if you can call it an easter egg but When my iBook crached in verbose mode it displayed a message that said something like: "Oh oh. Hold on. We're hanging here."

MilkmanDan · May 3, 2003, 01:00 PM

Congradulations, you've found the hidden Kernal Panic! Your Prize* is to lose all that useless data!

*Prizes donated by Microsoft. User responsible for all taxes, shipping, production of said prize, keeping of the prize, lifting, rolling. In event that contestant does not want the prize, a large man with a chainsaw will come over and take one of your arms. If you accept the prize a large man with a chainsaw will come over and cut off your leg. Any objections to said prize will result in the loss of your first born child. By reading this statement you agree to above terms, and the loss of all individual right. We own your soul. MS, everywhere you want to be.

theory · May 4, 2003, 08:16 PM

The reason know bothers writing code to recover from KPs is that
it is easier to write code that will prevent KP (mostly better drivers) and KP are supposed to be very infrequent (supposed to

).

kingskel · May 4, 2003, 09:12 PM

Sure, it's understandable that it's best to reboot after a KP. However, here's a cool suggestion that just occurred to me upon reading this thread: How about these?:

1. A "feedback" feature that would send an email to Apple after your KP. It would send only relevant console / low level info on what likely caused the crash.

Or...

2. A console log with the crash info sitting on your desktop after reboot. It would be "edited" by the system to have only the relevant (to the KP) info listed, and would clearly identify the likely cause of the KP.

Either of these would help either Apple or the user identify the source of the kernal panic, and work towards a solution or workaround.

What do you think?

Gul Banana · May 4, 2003, 09:22 PM

Both of those are actually possible, although the first one would be difficult. Maybe we'll see more improvements in this area later...

moki · May 5, 2003, 02:19 AM

Originally posted by geekwagon:
There is a very good reason why they always make you reboot after a kernel panic. Most panics are caused by the kernel hitting an error so bad that the programmer felt that recovering from that error reliably was impossible. So, since the environment the kernel runs in is now unknown, it just bails to minimize the amount of corruption to your data.

Well, yes and no. Technically you're right, but it depends on *what* is implemented in the kernel. With Mac OS X, USB devices are implemented in the kernel. This means that any mistakes in the USB kext will result in kernel panics. It makes sense that things such as this are implemented in the kernel.

It works like this. Consider Mac OS X to be two mostly separate operating systems (this is not really true, but work with the analogy). On the one side you have the kernel, on the other side you have applications and other "userland" code.

Let's say I have a chunk of code that runs in "user-land" (eg, an application), and it crashes. What happens is my application/process is killed by the kernel, and you get a little notice saying it unexpectedly quit.

If I have the same chunk of code running in kernel space, and it crashes, the machine kernel panics and dies.

Same code -- same severity of error -- but where the code is running depends on the severity of the result.

There are certain realtime tasks such as handling input devices that are best served by kernel space code (and indeed, at least on some level, they *must* be implemented there, at least as stub code).

I'm merely pointing out that the error that causes a kernel panic isn't necessarily that bad or dire, it just happens to occur in kernel space code.

As for USB, writing USB device drivers is notoriously tough, and Apple has broken, fixed, and re-broken the USB kernely code many times in Mac OS X. I'm sure if there is an error, they will fix it in future versions of the OS.

eightoeight · May 5, 2003, 03:02 AM

I think he was being sarcastic...

moki · May 5, 2003, 08:45 PM

Originally posted by eightoeight:
I think he was being sarcastic...

hrm... I don't think so, but perhaps...