Panther Eats FireWire 800 Drives

the_webmaestro writes: "Apple has announced that Panther (Mac OS X 10.3) may cause corruption with external FireWire 800 drives (anything with an Oxford 922 chip). Fortunately for me (unlike the poor souls who've already had problems), I guess I'm glad I ordered a lowly a 250GB Firewire 400/USB2.0 Combo Drive..." maccw reports that Firewire 400 customers are also reporting problems, as detailed from this Wired story.

Comments

Re:2 for 2

[Anonymous Coward]

Just FYI -- the Panther issue just ruins data, not the hardware. It's not clear where the problem is but it seems to affect a lot more chipsets than Apple is acknowledging right now.

Updates

[dynayellow]

According to this Apple page the problem is with the chipset.

OWC has posted a firmware update for their drives, as has Wiebetech. It looks like the Wiebetech requires you to update the firmware in Jaguar, and then they don't recommend using it IN Jaguar afterwards... Sheesh.

Re:Updates

[pjcreath]

Note that while OWC's firmware update has only been tested with their drives, other people in the forums have successfully used it to update other vendors' drives using the Oxford chipset.

Their updater will also tell you which chipset (911 or 922) your drive has and which firmware it's running. Note that some FW400 drives use the affected Oxford 922 chipset rather than the 911.

As of this writing, nobody's quite sure yet whether Oxford 911 (FW400) chipsets are also affected. Some evidence indicates that they are, Official Statements to the contrary.

Re:Updates

[gunnk]

THANK YOU!

The parent post is much more informative that the story itself, which links to the Apple Special Message page while appearing to have no knowledge of the content on that page (i.e.: problem is a chipset bug, not a Panther bug).

After the previous story in which Apple was said to be refusing to release security patches for Jaguar anymore (with no links to any actual Apple statement), it looks like any story that seems to paint Apple in a bad light gets no review at all. Come on -- we should be able to expect a LITTLE better editorial review of story postings than this!

In this case the title of the story is misleading and sensational, while the content puts the blame on the wrong party. Apple is doing a good job of getting the word out to people that may have the faulty chipset so that they don't get burned by the problem. Making this sound like a failure by Apple is just plain wrong.

What is up with software destroying drives?

[mhesseltine]

First Mandrake 9.2 and LG CD-ROMS, now this. Is this something the hardware manufacturers are doing that violates the common standards (IDE, ATAPI, IEEE1394)? Or, is this software companies trying to squeeze out a little more performance by not adhering to a standard?

Apple motherboards kill firewire bridges

[littlerubberfeet]

I work at a small sound studio. We use macs for their stability and simplicity. I had been trying to figure out what to do with the 4 dead WD Caviar 30gig drives, 5 dead Maxtor 80 and 160 gig drives, and 4 dead Que! CD-RWs. They are all external firewire devices. After getting no where with Maxtor or Apple, and not being able to pinpoint the failure to a single G4 or circumstance, I took apart the damn things.

ALL the drives were still perfectly good. The firewire bridges were bad. ALL the drives were advertised as hot-swappable. Almost all of the brideges died during, or after a hot-swap procedure. Indigita, a bridge company, has been gracious enough to test some stuff for us.

I suspect that the problem isn't just with panther or bridge firmware. I think there is a problem with the mac firewire interface generally, expecially when hot swapping.

Never Never buy anything the first day.. Never

[G4from128k]

This just goes to show the hideous complexity of all the myriad permutations of software, platforms, and peripherals that defines modern computing environments. All software for a mainstream platform is beta when it is first released because it is impossible for the maker to test every possible combination (or even most of the common ones). Even with lead user groups and the developer community testing the so-called "betas", there are always missed combinations of equipment. Only when millions of people try something do you see the real problems appear.

So don't buy new software the day it appears. Wait and see what problems it causes and then buy the x.x.1 update after a suitable wait to see if that's safe.

BTW, this rule applies to new pharmaceuticals. I'd recommend waiting at least 10 to 100 million doses and 5-20 years before taking any "new" medication unless it is a total blockbusting lifesaver.