"Apple to Users: 'It's Not Our Problem'":
April 6, 2001 7:28 PM Subscribe
"Apple to Users: 'It's Not Our Problem'": Apple is doing nothing about their new firmware update which doesn't recognize some third party RAM chips.
tsk. apple should let other people do their hardware and try to make a good os for once. Actually, the only thing they do well is make nifty cases, come to think of it, maybe they should just become third-party?
posted by Hackworth at 8:28 PM on April 6, 2001
posted by Hackworth at 8:28 PM on April 6, 2001
Piffle. There's a long standing tradition amongst Mac users to find workarounds for the curveballs Apple throws at us from time to time. In this case, at Macintouch, you can find links to DIMMCheck, a lovely wee program which has fixed the problem for many people, including myself.
posted by gsh at 8:54 PM on April 6, 2001
posted by gsh at 8:54 PM on April 6, 2001
It's coming out that there's more to it than "not meeting specs". In some cases the RAM is faster than the RAM that Apple sells and is still rejected. (Fast RAM can always be run slower; faster RAM should always work.)
This information from Macintouch:
Date: Tue, 3 Apr 2001 17:57:44 +1200
From: Glenn Anderson
Subject: Mac Firmware upgrade and memory problems
At the weekend I tracked down the problem with a PC100-222 DIMM in my PowerBook G3 and firmware 4.1.8, and fixed it. The information in the Serial Presence Detect (SPD) EEPROM on the DIMM did not conform with the PC100 standard, in particular, there were no settings for CL3 mode, only CL2 mode. I was able to fix this by getting the specs for the chips on the DIMM (from the web site of the chip maker), and reprogramming the SPD EEPROM with settings for both CL3 and CL2 mode. The DIMM is working just fine now.
The SPD EEPROM can be reprogrammed from software, but I will need the details from some more problem DIMMs and more details on other hardware before I can put together a program that anyone can use to do this. I have a program called DIMMCheck available from my web server. It dumps the contents of the SPD EEPROM and it should identify problem DIMMs. If people with problem DIMMs can run this, copy the results, paste it to an email, and send it to dimms@mactcp.org.nz, I will see if I can put together a fix this weekend.
Glenn.
(Note: DIMMCheck in its current form doesn't solve the problem. It does tell you ahead of time whether the upgrade will deactivate your RAM, so that you can decide not to upgrade.)
posted by Steven Den Beste at 10:21 PM on April 6, 2001
This information from Macintouch:
Date: Tue, 3 Apr 2001 17:57:44 +1200
From: Glenn Anderson
Subject: Mac Firmware upgrade and memory problems
At the weekend I tracked down the problem with a PC100-222 DIMM in my PowerBook G3 and firmware 4.1.8, and fixed it. The information in the Serial Presence Detect (SPD) EEPROM on the DIMM did not conform with the PC100 standard, in particular, there were no settings for CL3 mode, only CL2 mode. I was able to fix this by getting the specs for the chips on the DIMM (from the web site of the chip maker), and reprogramming the SPD EEPROM with settings for both CL3 and CL2 mode. The DIMM is working just fine now.
The SPD EEPROM can be reprogrammed from software, but I will need the details from some more problem DIMMs and more details on other hardware before I can put together a program that anyone can use to do this. I have a program called DIMMCheck available from my web server. It dumps the contents of the SPD EEPROM and it should identify problem DIMMs. If people with problem DIMMs can run this, copy the results, paste it to an email, and send it to dimms@mactcp.org.nz, I will see if I can put together a fix this weekend.
Glenn.
(Note: DIMMCheck in its current form doesn't solve the problem. It does tell you ahead of time whether the upgrade will deactivate your RAM, so that you can decide not to upgrade.)
posted by Steven Den Beste at 10:21 PM on April 6, 2001
This is one of the biggest pieces of crap I've seen in years.
Some Macs aren't running right because they have out-of-spec DIMMs -- memory that shouldn't have been installed in the first place.
Some memory makers sell DIMMS that are within hardware specifications, but the firmware on them is not.
Apple updates the firmware so that DIMMs that identify themselves as out-of-spec are disabled, because they shouldn't have been installed anyway. In many cases, the DIMMs can be fixed by simple reprogramming so the DIMM firwmare accurately reports what the hardware is capable of doing. But as far as the Mac firmware is concerned, a DIMM that doesn't say it meets specifications is dangerous and shouldn't be used.
Now that it's hit the fan, the memory vendors are all saying they'll reprogram the DIMMs -- when they couldn't be bothered to have done it in the first place and avoided the problem before it started.
Yes, Apple could have done a better job of warning people that out-of-spec memory would be disabled after applying the firmware upgrade, and maybe the installer could even have checked first. That's a valid complaint. But to bitch that Apple should enable DIMMs that identify themselves as not meeting specs when doing so has caused stability problems? That's ridiculous. Don't blame Apple, blame the memory vendor who didn't do the right thing the first time.
posted by mdeatherage at 10:23 PM on April 6, 2001
Some Macs aren't running right because they have out-of-spec DIMMs -- memory that shouldn't have been installed in the first place.
Some memory makers sell DIMMS that are within hardware specifications, but the firmware on them is not.
Apple updates the firmware so that DIMMs that identify themselves as out-of-spec are disabled, because they shouldn't have been installed anyway. In many cases, the DIMMs can be fixed by simple reprogramming so the DIMM firwmare accurately reports what the hardware is capable of doing. But as far as the Mac firmware is concerned, a DIMM that doesn't say it meets specifications is dangerous and shouldn't be used.
Now that it's hit the fan, the memory vendors are all saying they'll reprogram the DIMMs -- when they couldn't be bothered to have done it in the first place and avoided the problem before it started.
Yes, Apple could have done a better job of warning people that out-of-spec memory would be disabled after applying the firmware upgrade, and maybe the installer could even have checked first. That's a valid complaint. But to bitch that Apple should enable DIMMs that identify themselves as not meeting specs when doing so has caused stability problems? That's ridiculous. Don't blame Apple, blame the memory vendor who didn't do the right thing the first time.
posted by mdeatherage at 10:23 PM on April 6, 2001
I'll second that, third it. The original linked story and the summary by paladin are slanted and generally ill-intentioned. Apple's firmware upgrades increase the auto-spec checks of the ROM to weed out ram that is sub-par. That is all. This is not, has not been, and should never be a problem blamed on Apple. Certain RAM vendors are offering exchanges or offering to re-program the EEPROM: they realize, as does anyone who's been following this since the beginning, that they are at fault.
posted by Mo Nickels at 2:30 AM on April 7, 2001
posted by Mo Nickels at 2:30 AM on April 7, 2001
I have to admit that I updated my firmware before I saw any of the warnings about RAM. I was praying with I took a look at About this Computer...
posted by Brilliantcrank at 8:36 AM on April 7, 2001
posted by Brilliantcrank at 8:36 AM on April 7, 2001
Anti-Apple bias in the media? Heavens, that couldn't be true.
posted by Zeldman at 2:44 PM on April 7, 2001
posted by Zeldman at 2:44 PM on April 7, 2001
Glenn Anderson (mentioned above, author of DIMMCheck) has gone ahead and written a hack to restore disabled DIMMs.
posted by darukaru at 4:34 PM on April 7, 2001
posted by darukaru at 4:34 PM on April 7, 2001
So when a video card vendor releases a card that is incompatible with certain versions of Windows, is it Microsoft's fault? Lame...
posted by camworld at 5:13 PM on April 7, 2001
posted by camworld at 5:13 PM on April 7, 2001
Again we see the holy war flare up again. I'll break it down for all:
Apple did something to hose it's users by not informing them of the possible ramifications of this firmware upgrade.
and
MS is still evil
I think the people on the Windoze side aren't really pointing fingers saying "neener neener, look what stupid Apple did", so we can all continue to play nice
posted by tj at 10:13 PM on April 7, 2001
Apple did something to hose it's users by not informing them of the possible ramifications of this firmware upgrade.
and
MS is still evil
I think the people on the Windoze side aren't really pointing fingers saying "neener neener, look what stupid Apple did", so we can all continue to play nice
posted by tj at 10:13 PM on April 7, 2001
The original linked story and the summary by paladin are slanted and generally ill-intentioned.
I own an Apple. I am writing this on a PowerMac G4. That does not mean that when I see something like this, where, as tj put it,
Apple did something to hose it's users by not informing them of the possible ramifications of this firmware upgrade...
that I'm going to overlook it. I like my Mac, and I'm not going to switch over to the Wintel world because of this, but I would not have liked to have been someone who suddenly couldn't use RAM they paid good money for.
posted by paladin at 1:41 AM on April 8, 2001
I own an Apple. I am writing this on a PowerMac G4. That does not mean that when I see something like this, where, as tj put it,
Apple did something to hose it's users by not informing them of the possible ramifications of this firmware upgrade...
that I'm going to overlook it. I like my Mac, and I'm not going to switch over to the Wintel world because of this, but I would not have liked to have been someone who suddenly couldn't use RAM they paid good money for.
posted by paladin at 1:41 AM on April 8, 2001
But why are you - and, by extension, other Mac users - blaming this on Apple? If there's a "villain" in this situation, it's the RAM vendors. Apple hasn't changed the specs of the RAM that is required by any Macintosh that is shipping today. The only thing they did was to step up their checking of that RAM to make sure the chips were, in fact, who they "said" they were at boot time. The vendors of some chips - I refuse to speculate why - had either intentionally or unintentionally made their chips "say" they were something other than what they are. This would be bad for any computer that made assumptions based on that, and bad for you as the user of that computer. You want to bitch? Call your RAM vendor. And while you're near the phone, why not give Apple a call and thank them for forcing the RAM vendors to stop dicking around with the chips in your Mac...
posted by m.polo at 9:06 AM on April 8, 2001
posted by m.polo at 9:06 AM on April 8, 2001
People are mad at Apple because, without warning, their computers became markedly less functional and in some cases completely non-functional, forcing them to do without their machines until the memory could be replaced or else forcing them to visit CompUSA and pay retail for more memory. The least they could have done is tested the memory before installing the upgrade and warned you if some of your RAM was out-of-spec, or provide a way to re-install the old firmware. Disabling functionality without warning is not very friendly and is completely antithetical to what Mac users expect.
All indications are, however, that Apple simply did not realize just how many out-of-spec memory chips are out there. No manufacturer, not even premium name brands like Crucial, seem immune to the problem. The vendors themselves rely on the manufacturers to test their products (although some may spot-test them) and to meet spec; if they work in the computer they're intended for, then they are assumed to meet spec. Even a dedicated testing rig might pass the "defective" memory modules since it might only check to make sure the module meets all the specs, not whether it has correct data in its EEPROM (which is the issue here).
In short, the things Apple is testing for in its new firmware have never ever mattered to any other computer ever made, and for this reason are probably not routinely tested for by manufacturers or vendors. Perhaps they should be, but when no functionality depended on it, it's understandable why they wouldn't. Apple should have tested the firmware update on a much wider range of third-party memory modules before releasing it to the public. If they had, they would have realized the necessity of testing the installed memory before proceeding with the update.
posted by kindall at 1:09 PM on April 8, 2001
All indications are, however, that Apple simply did not realize just how many out-of-spec memory chips are out there. No manufacturer, not even premium name brands like Crucial, seem immune to the problem. The vendors themselves rely on the manufacturers to test their products (although some may spot-test them) and to meet spec; if they work in the computer they're intended for, then they are assumed to meet spec. Even a dedicated testing rig might pass the "defective" memory modules since it might only check to make sure the module meets all the specs, not whether it has correct data in its EEPROM (which is the issue here).
In short, the things Apple is testing for in its new firmware have never ever mattered to any other computer ever made, and for this reason are probably not routinely tested for by manufacturers or vendors. Perhaps they should be, but when no functionality depended on it, it's understandable why they wouldn't. Apple should have tested the firmware update on a much wider range of third-party memory modules before releasing it to the public. If they had, they would have realized the necessity of testing the installed memory before proceeding with the update.
posted by kindall at 1:09 PM on April 8, 2001
« Older George Jetson, your car is here! | A dot com (sort of) that's making money Newer »
This thread has been archived and is closed to new comments
posted by jragon at 7:44 PM on April 6, 2001