Other Weird, accidental, but nifty things you’ve discovered about Early Intel Macs

[B S Magnet]

Seems there wasn’t a thread like this before on the Early Intel side (but a couple along those lines on the PowerPC forum), so why not make one?

I’ll start:

Recently, whilst cleaning out and setting up a mid-2007, turn-up-the-base iMac I got for the low, low price of free, I learnt something unexpected — sort of the way I remember learning the inverse with a DLSD PowerBook G4 (those were the very last of the PowerBook G4s, which had a completely redesigned logic board, in anticipation of a PowerPC future which never came to be).

Over on the PowerBook G4, once upon a time, I verified how, yes, one can run a GUID-partitioned boot volume on those last models, and the system will boot into Leopard. I remember there being a couple of, for want of a better word, hinky quirks about this combination which, in the end, had me re-doing it all with an Apple Partition Map.

Over on this A1224 (iMac7,1), base-how-low-can-you-go 2.0GHz Santa Rosa C2D, I stuck in an SSD previously used in my mid-2004 Power Mac G5 (running Leopard Server). I hadn’t even thought about the SSD’s partition tables when I threw it in there, but I mostly wanted to make sure the system was detecting the SSD and also to boot an already-extant OS X build.

Well, turns out it booted fine, but there were other quirks which kept me from using ethernet to connect to the network to install SL on it — namely, ethernet was also acting hinky. It was only able to use a connection from another Mac sharing internet via wifi. When trying to connect to the router, the ethernet port showed no activity, and in Network prefs, ethernet was red-lighted (wifi, on the channel I use at home, wasn’t seen at all).

It was only once I began to install a fresh copy of SL on it (imaged onto a second SSD partition on the just-installed SSD), when I got the “cannot install on this volume” warning, with an added note that the partitioning scheme was incompatible with an installation of Snow Leopard. Reflexively, I thought, “Well why wouldn’t this be a GUID SSD?” — forgetting it came from an earlier G5). It was only after I opened Disk Utility when I remembered how the SSD was (of course) set up with APM!

So yah, there was a time, a weird slushy window, when very late PowerPC Macs could boot from a GUID-parition scheme installation of OS X, and the earliest Intel Macs could boot into OS X on an Apple Partition Map disk. But within both oddball environments, there were weird quirks which managed to find their way to the user’s attention — sometimes fairly quickly.

So yah. I had no idea!

What unusual quirks have you all discovered on Early Intel Macs?

 

[rampancy]

These are all old-hat to most folks by now, but jump starting Macs using the motherboard power pads is a nifty trick that I'd previously never encountered on my iBook or any other PPC Mac I'd worked with. Ditto for early Intel Macs downclocking to 1.0 Ghz when using a dead battery.

Other odd things I've gleaned over the years:

• Most A1181 MacBooks simply won't boot past a black screen if there's no RAM installed. I always imagined they'd sound a hardware error chime or flash a sad Mac screen. I had a perfectly good working 5,2 MacBook languishing in my parts pile for years because I assumed it was dead. Once I popped known good RAM in both of the slots, it worked fine. Go figure.
• The Core Duo (and Core 2 Duo GMA 950?) MacBooks also seem require to a working battery to prevent damage from short circuits. I think it's because the battery is involved in helping to ground the Mac when its plugged in? I've lost a couple of MacBooks due to startup issues when run plugged in with dead batteries that seem to be related to short circuits.
• El Capitan will fail to install if doing a clean reinstall from Apple's recovery servers, unless you set the time/date to a value before the expiry of the installer's bundled security certificate. It took me days until I figured this out.
• If you're doing a thermal compound replacement on an A1181, remove the whole fan assembly as well as the heatsink (contrary to iFixIt's instructions). You'll save yourself a whole lot of trouble that way. Also, be gentle, and don't overtighten the heatsink screws when reinstalling. There's a risk of cracking connections or components due to motherboard flex.
• The GMA 950 A1181s provided full bus power over both USB ports; every revision that came after only provides full bus power to the port closest to the front.
• For some reason I don't understand, Linux live DVDs simply flat out refuse to work (in my experience) on an A1181 (regardless of whether it's 32-bit) when run off of an external USB DVD drive.

 

[GMShadow]

Used to be common knowledge but it's pretty lost/buried at this point and not the easiest to find in searches - Core Duo MacBook Pros often had coil whine problems when the CPU would sleep the cores.

However, unlike a lot of modern coil whine issues, this can be alleviated by running an app like QuietMBP (download link near bottom) that simply runs just enough junk background processes to keep the CPU from putting cores to sleep, eliminating the whine.

That said, if you're running on battery, remember to turn it off or you'll wonder why it's draining so fast. 😄

 

[sdfox7]

These are all old-hat to most folks by now, but jump starting Macs using the motherboard power pads is a nifty trick that I'd previously never encountered on my iBook or any other PPC Mac I'd worked with. Ditto for early Intel Macs downclocking to 1.0 Ghz when using a dead battery.

Other odd things I've gleaned over the years:

• Most A1181 MacBooks simply won't boot past a black screen if there's no RAM installed. I always imagined they'd sound a hardware error chime or flash a sad Mac screen. I had a perfectly good working 5,2 MacBook languishing in my parts pile for years because I assumed it was dead. Once I popped known good RAM in both of the slots, it worked fine. Go figure.
• The Core Duo (and Core 2 Duo GMA 950?) MacBooks also seem require to a working battery to prevent damage from short circuits. I think it's because the battery is involved in helping to ground the Mac when its plugged in? I've lost a couple of MacBooks due to startup issues when run plugged in with dead batteries that seem to be related to short circuits.
• El Capitan will fail to install if doing a clean reinstall from Apple's recovery servers, unless you set the time/date to a value before the expiry of the installer's bundled security certificate. It took me days until I figured this out.
• If you're doing a thermal compound replacement on an A1181, remove the whole fan assembly as well as the heatsink (contrary to iFixIt's instructions). You'll save yourself a whole lot of trouble that way. Also, be gentle, and don't overtighten the heatsink screws when reinstalling. There's a risk of cracking connections or components due to motherboard flex.
• The GMA 950 A1181s provided full bus power over both USB ports; every revision that came after only provides full bus power to the port closest to the front.
• For some reason I don't understand, Linux live DVDs simply flat out refuse to work (in my experience) on an A1181 (regardless of whether it's 32-bit) when run off of an external USB DVD drive.

Good points. I found this pjobson Github page a while ago and it works perfectly for the old OS.

OSX Install Dates

10.16 - 0.5 Leopard   - date 0101010121 <- Currently not needed
10.15 - Catalina      - date 0101010120 <- Currently not needed
10.14 - Mojave        - date 0101010119 <- Currently not needed
10.13 - High Sierra   - date 0101010118
10.12 - Sierra        - date 0101010117
10.11 - El Capitan    - date 0101010116
10.10 - Yosemite      - date 0101010115
10.9  - Mavericks     - date 0101010114
10.8  - Mountain Lion - date 0101010113
10.7  - Lion          - date 0101010111
10.6  - Snow Leopard  - date 0101010110
10.5  - Leopard       - date 0101010108

 

[B S Magnet]

I’m adding a new, nifty thing which, in most if not all instances, applies to both Early Intel Macs and also PowerPC Macs with FireWire:

One can daisy-chain multiple, FireWire-connected, target mode Macs together. All volumes across those daisy-chained will, on the destination Mac (the one mounting the target mode drives) appear on the desktop.

On the face of it, knowing how IEEE1394 protocol works, this isn’t a surprise. But it isn’t often one thinks to connect a FireWire target mode Mac to another FireWire target mode Mac, then to another Mac running OS X/macOS. It works exactly as any other daisy-chained FireWire external hard drive enclosures would.

Once upon a time, for those of you who’ve worked only sporadically with FireWire devices, this was a fairly common scenario, especially when relatively fast external storage and an external optical drive burner were key components in a professional setup (bonus if the setup also sported an MO — or magneto-optical — drive).

 

[Alexis Trinquet]

One can daisy-chain multiple, FireWire-connected, target mode Macs together. All volumes across those daisy-chained will, on the destination Mac (the one mounting the target mode drives) appear on the desktop.

TIL. That's something I really wanna try now 🤣

---

I figured, while I'm waiting on parts for my MDD, might as well throw that one here that I don't know if it has been brought up in the past or not:

On some (?) Intel Macs, you can boot from a MBR-partitioned hard drive, in my case, this was off a SSD I got from a (now long gone) Core2Duo PC that ran Linux that I had put into my A1225.

It just booted from it without any fuss, unlike my now-dead Santa Rosa MBP that would be extremely picky about anything that wasn't EFI partitioned.


Oh and one last tiny tidbit, regarding FireWire: on Macs that have both FW400 and FW800 ports, you can use them as a passive FW400 to 800 bridge, even if said Mac isn't running