Mac OS X 10.3.3 Update Released
joelhayhurst writes "Mac OS X 10.3.3 is now available via Software Update. Full details are available on the Apple site - key enhancements include: 'Network volumes are now available in the Finder sidebar and Desktop for easier access... Improved file sharing and directory services for Mac (AFP), UNIX (NFS) and PC (SMB/CIFS) networks... Improved PostScript and USB printing... Updated Disk Utility, DVD Player, Image Capture, Mail and Safari applications... Additional support for FireWire and USB devices... Improved compatibility for third party applications... Previous standalone security updates and Bluetooth Update 1.5."
Funny. I was downloading the update only a few minutes before this article was posted.
:(
I sure do hope the improvements to the SMB protocol are substantial. I've had no end of problems using it to access SMB shares on Windows and Unix.
You're right, I wouldn't steal a car. But if it were possible, I sure as hell would download one!
When ever I used a network share and I changed something from the server or from another computer I'd have to play with the finder for a long time to get it to refresh.
I happen to like the icon preview and I had it on globally by default. The first time I hooked up my digital camera the finder jumped to 100% cpu usage and stayed that way even after generating all the thumbnails.
I'm glad they're fixing these bugs.
-Francis
I'm just this guy, you know?
They STILL haven't fixed the Mail.app bug that i probably submitted in October last year. If you use a mail server with imap-ssl that doesn't have a verified certificate, you get a pop up complaining whenever you open mail. If you click show certificate, there's a certificate icon you can supposedly drag to the desktop and then use in keychain according to apple's support webpages. Well, dragging that image still crashes Mail.app (and hard).
I always figured "complete crash" bugs were high priority. sigh.
- tristan
Anyone know if the bug with Energy Saver not working on mondays is fixed? I don't know when I'll get to update my machine and am curious if this was tended to. They didn't mention it in the support document, but it's such a minor bug anyway...
fs
>1) Sometimes (10% of the time?) I just get a black screen on wake and have to reboot (CAPSLOCK key LED turns on/off though...so it ain't "dead").
same thing happens with 12" PB. i've noticed that the problem is usually when i put the machine to sleep when it's "busy." for example, you can reproduce the problem almost every time if you plug or unplug an USB mouse then shut the lid immediately.
the screen goes black, keyboard LEDs work but the machine will not wake up. my solution has been to give the machine some time before you put it to sleep after you plug/unplug stuff or launch some apps.
The 10.3 Finder's Network Browser now seems to be finished:
1 - Connected servers now appear on the Desktop and in the Sidebar when the preferences have been set for this option.
2 - Connected servers can be ejected via normal means.
3 - Password options are now present including Keychain functions.
4 - User folders no longer appear twice.
I'm a happy camper.
Also, a blurring problem when in edit mode in iPhoto 4 appears to be fixed as well.
My suspicion that the "snappier" feeling comes from the massive prelinking that the OS goes though when installing stuff.
While OS X has nothing approaching the DLL hell of Windows, prelinking associations can at least theoretically break, and I am fairly sure that the entire OS is only prelinked when installing a major update. Otherwise installing apps would take even longer than usual
I get a similar problem consistently on my old FW PowerBook:
- put machine to sleep with an external monitor attached
- remove monitor
- wake machine before the case is opened (by attaching a USB mouse while the machine is plugged in)
the machine will awaken without recognizing that a screen is attached. The drive will be spinning, the caps lock will work and you can connect to it if you're attached to a network. You can hit the power key and return to make it shut down in a controlled manner.
Because it doesn't recognize the LCD, closing and opening the case won't trigger sleep (on my old 'book, at least), and therefore won't force it to detect monitors after a proper case-opening awakening.
My lame solution is to leave the machine until it falls asleep; a better solution is to have a script that detects monitors attached to a function key. Even better would be if this update fixes it outright, but I'll be waiting a few days for bolder people to test this patch out before I try.
Ok, I give up. I have NFS volumes exported/mounted across a few Linux systems (no issue). OS.X mounting the same Linux NFS mount can only handle 2.something gig files (it's bigger than 2 and smaller than 3 is all I know right now).
/. ideas out there?
:)
A 4,707M file shows as 611M on OS.X. No, I can not create files larger than the "limit" -- they eat themselves. The same applies to both the command line and the Finder. Any
(yes, I want to move 50G files around sometimes
I realize that Apple doesn't officially acknowledge problems with the OX911 chipset (only the 922), but does anyone know if they have fixed the bug yet?
All other bugs are low priority for me, as I have a TiBook and external storage that needs to be accessed - RELIABLY!
Is it only me, or are all Java-based apps now unusable? I tried a few and they all start (show up on the dock) then disappear. I haven't done anything to Java external to Apple's updates to make this happen. Not critical as I don't really use Java that much, but I do have a couple of handy ldap tools written in Java that would be nice to get working again.
Here's hoping the mysterious freezing bug a lot of people have experienced, including me, is fixed. It was evidently introduced in 10.3.2.
this is less of a but and more of the finder being stupid.
i NEVER, EVER trust what the finder says because it only updates that figure once every 10 minutes or so. a much more reliable option is to open up the terminal and run df -h. this tells you the current usage information for all mounted disks. (rather than the free space of the disk that is up to 10 minutes out of date) just like in every other unix environment.
the good news is that this update does have the finder updating other stuff correctly, so this may be covered. i still won't ever trust it again.
Don't call me back. Give me a call back. Bye. So yeah. But bye our, well, but alright we are on a shirt this chill.
I updated my Quicksilver at work and had it save the update. Update went fine, so I copied it to my 15" AL PB. When I start the installer on the PB, it runs the 'little program' then goes to the list of updates. Then the buttons on the install screen die, and I have to force quit out of the installer.
Leery to install, but perhaps it's because I didn't do the software update version.
Performance doesn't seem to be lacking on the QS, but I didn't give it a good run-through either.
I just updated and now Reason won't recognize my M-Audio usb keyboard (Core MIDI error). Aaaarrrgggh!! Beware . . .
harmonious design
I deleted some extraneous Keystation drivers in MIDI Setup Utility and all is well . . .
harmonious design
Sure it would be nice to sort by date in column view - but there has to be something to "invent" in 10.4, right?
And I can no longer play DVD's with my Lacie DVD R/RW-drive.
"There was an initialization error. A valid DVD drive could not be found (-70012)
Does anyone else have that problem after updating?
Of particular interest to audio professionals, this update includes native mLAN support.
mLAN is a standard developed by Yamaha which allows lots of audio and MIDI info to travel over FireWire.
Yamaha's hottest new box, the 01X, has been without OS X drivers since it began shipping a month ago, and as of today it will finally work (in theory anyway)...
Of course time will tell how effective this new version of mLAN will be in terms of latency and stability. Now if you'll excuse me, I've got some downloading to do!
I've been burned by this one too many times on a variety of UNIX systems. I've had a policy for some time now that all edits to precious config files (regardless of where they reside) are done against an RCS archive.
A simple rcsdiff now tells you exactly what has changed so you can accept or edit, lock, and checkin your munged changes to the archive. A bit more straight-forward than trying to figure out which installer-created backup copy is the last "good" one.
Guards against accidental deletion, as well. Also especially helpful when testing with a variety of settings for a new daemon you are configuring. You can always go back to a known good condition, and/or save those clever tweaks for posterity!
Remember kids: ci -u is your friend.
-- clvrmnky
Yank the added memory and reboot. I had a number of kernel panic probs with the same model due to the memory stick I'd added.
(hehe)
Anti-aliased lines have long been borq'n under Panther, but no longer! AA lines now work properly with both my GF4MX card and Powerbooks Radeon 9600 (Maya & C4D).
I just installed the update about a half-hour ago, and upon trying to launch Mathematica 5.0.0, I was greeted with the registration screen. I went through it and the program generated a different machine number than it had before (I checked it against the old one in the email). Did anyone else have this problem?
First, I want to make it clear I'm a hobbyist developer... not a professional.
t ml ) and found that OpenGL's the culprit. There's nothing I can do about it! At least not that I know of. My OpenGL code's not the best in the world, but it's simple and fast and caches all geometry in display lists and so on and so forth. I spent a fair amount of time profiling and optimizing already.
So, after reading largely good comments about this update I installed it on my laptop, my development machine. I'm developing a simulation framework for development/testing of behavioral AI agents. The simulator uses the ODE physics engine for dynamics and OpenGL for visualization. Ho hum. Only 10'000 CS students have done the same thing. But this one is mine.
I'll be the first to admit I'm not a professional OpenGL programmer. I'd never written a line before June 2003, but I did the Right Thing and bought the _Red Book_ and went through it from nape to nuts. I also made the requisite vists to NeHE and gamasutra for performance tips.
Anyway, until I installed 10.3.3 my opengl code amounted for a whopping 5 to 10 % of my program's running processor time. In other words, no problem. The physics engine took up another 10 to 30% and that left *plenty* for the separately threaded AI logic. It worked beautifully on my rev A PB 12" -- most of my simulations ran at 100% "simulation/real" time. Which is to say, I set a fixed simulation timestep to 0.05 seconds, allowing for opengl @ 20fps and 20 synchronized physics steps per second. Enough to be "smooth" but not so much as to take away too much processing time from the AI. This on average left 2 or so millis per frame for AI logic, if you want to think of it in a single threaded way.
Immediately ater installing 10.3.3 my average performance dropped to 35% to 50% "real" time. Meaning my average simulation timestep was taking not 0.05 seconds but at least 0.1 or more. No good.
I sharked it ( http://developer.apple.com/tools/shark_optimize.h
I'm so frustrated... I used to be able to run my simulations in real time but now I can only site and watch it crawl ( and listen as my PB's fan kicks in ). Considering it often takes several minutes observation to determine if a new behavior has improved or ruined an agent's performance, it's quite a pain to have to wait at least twice as long.
Let's just hope that this is a dorked driver, and apple releases an update. & perhaps, I can somehow dig up the 10.3.2 opengl drivers somehow.
Oh well. Since I love cocoa programming, and since I love my PB, and since I drank Steve's Kool-Aid long ago, I still love apple.
What would love be if the object of your affection didn't disappoint you now and then?
lorem ipsum, dolor sit amet