Awesome breakdown and really interesting results

Still. I have to admit, the single 1.25 2003 running OS 9 holds a special place in my heart.

Quieter that the Dual 867, 2MB cache, and most apps really didn't use the 2nd CPU in OS9, so the thermals were so much better. I guess we all have our fond memories and favorites.

Not one to ever use GeekBench, before now (because as you know… OS9 Lives!) — but after assisting someone upgrading an 867 DP MDD with a 1.25 GHZ DP MDD CPU (via removal of the R676 resisitor from the 867’s mobo)… in order to increase the Bus Speed from 133 MHz to 167 MHz… thought I might compare EveryMac.com’s MDD stats (which use GeekBench 2.0).

Could only source GeekBench 2.2 and that only works with Tiger 10.4.11 (not 10.4.6). So some machines here were quickly updated to OS 10.4.11.

Now, if you compare the GeekBench score of the original 867 DP MDD (727) with that of the 2002 1.25 GHz DP MDD (1078) that seems a good OA performance boost. Yet, if you compare the original 867 DP against the Single Processor 2003 1.25 DP, somehow the GeekBench results are the exact same 727? Whaaat?

Apologies for the large, attached MDD graphic - but uploaded or embedded any smaller, it becomes nearly unreadable. So you might temporarily switch to the Forum’s “Blu” theme - OR drag the graphic out of your browser and onto your desktop and then view it there @ full size.



Just for fun, compare the 2002 1.0 GHz DP MDD to the 2003 1.0 GHz Single Processor MDD. Geekbench 2.0. 727 vs. 570. But the 2003 FW 800 model was less expensive (by a mere $200). OR, the 2002 1.0 DP against the 2003 1.0 SP. (888 vs. 570) And $2499 vs. $1499. Wonder why I prefer to transplant 1.42 GHz DP processors onto FW 400 mobos?

EveryMac.com doesn’t list the 2003 1.25 DP MDDs’ specs (that I have yet to find anyway) but I did run GeekBench 2.2 on a 1.42 DP that I transplanted onto a 2003’s 1.25 DP’s mobo and came up with the following GeekBench 2.2 rating of 1167. Now I should put the original 1.25 DP CPU back in and GeekBench it again. AND maybe change the RAM. (Might also miss that 2.0 MB of L3 cache?)



AND this is all (of course) part of yet another little project on the ever-always-running list here.

So, anyone have a copy of GeekBench 2.0? •  Awaiting an answer from EveryMac.com on this.

---

PowerBook G4 1.5 17 - Geekbench score here 808. Every Mac, 745.

G4 Quicksilver with 1.67 GHz Sonnet CPU, 855. Every Mac, 1.0 GHz QS, 888.

I still do not like GeekBench.

I've updated the Mac mini 7.6.1 CD to Test Version 0.2.

You can download the iso from the Macintosh Garden if you want to try it out:
https://macintoshgarden.org/apps/mac-mini-g4-system-7

CD 0.1 had some incorrect icons; the System itself is unchanged from 0.1.
Also now just a 43 MB download.

it comes with 20 custom "effects" which are nothing more than transitions (wipe, dissolve...) and it can convert to different formats and sizes in realtime. that is all i know about it.

All the downloads are on Hotline and public, like always. The website is moving to a new host and when that is done the links on the website will be updated.

Sorry for the inconvenience.

Does anyone here have any experience with the CineWave RT cards for Final Cut Pro? I'm starting the process to save some miniDV tapes before they die and it occoured to me that it would be fun to do some editing in Final Cut Pro in OS9 on the MDD, but the lack of real time effects is killing my workflow.

Back in the day the CineWave cards added some RT (Real Time) effect functionality to Final Cut, but I've never experienced them first hand. Does anyone here have any experience with them?

The cards are avaliable on eBay and the software seems to be avaliable on the garden, but I have no idea if these things require some kind of hardware dongle to work, a serial code that doesn't exist anymore, or if I need one of the breakout boxes to even get the video inside (or is my camcorder connected over miniDV enough if the correct codec is used?).

And if all of that is in place, what do I get in terms of added functionality? Is there a general hardware acceleration, or do I need to use CineWave made plugins to Final Cut?

Many questions! Help me out here

you guys are my heroes. just let the rest of us know once there is an easy method to bless our mac mini g4s with system 7/8/9 multiboot with an easy iso install

Quote from: Jubadub on December 04, 2025, 10:14:23 PM

Quote from: jjuran on December 04, 2025, 07:00:23 PM

I use Metrowerks C++ 2.4.1 from CodeWarrior Pro 6.3, by way of MPW's ToolServer, driven by Apple event from MacRelix.  Some of my code also builds with CW Pro 4 (the last version hosted on 68K).  I've written my own post-linkers (which run in MacRelix) to remove the dependency on StripAddress() in Metrowerks' runtime code.

Git 2.2.2 indeed ships with MacRelix.  I wouldn't recommend using it with massive repositories like metamage_1.git (except natively in Mac OS X), but it should be fine with small repos.

Hey, great to have you here as well, @jjuran! Thanks for porting Git to Mac OS 9 one way or another.

Just an idea, but do you think you could record a short video capture of this whole process? I would love to see it in action! I think it'd also encourage people to use Git and your other ideas on Mac OS more, that's an excellent workflow.

Two questions:  (a) How much of the process do you have in mind?  It's all very well to install MacRelix, run it, and even create a Git repo, but it's another thing entirely to set up the CodeWarrior/MPW environment that the build system depends on.  It really just needs to be automated...

Well, you did say "a short video", so I guess that's just watching the build happen.

(b) How do you recommend recording a video?  Back in 2012 when I made the FORGE demo[1], I ran SheepShaver in Debian and used ffmpeg with the x11grab input mode.  For the MacPaint screencast[2], I used my own recording facility built into xv68k and a tool to convert the result to MP4.

If I had a build of SheepShaver that could connect to an AMS front end, I'd use that.  Alternatively, I might write something that records from within Mac OS in a VBL task, and writes to a file on the extfs volume.  That won't happen right away, though.

FORGE:  Hello World
[1] https://www.youtube.com/watch?v=JgivHy-oC2g

MacPaint in Advanced Mac Substitute
[2] https://www.v68k.org/advanced-mac-substitute/video/MacPaint-in-AMS.mp4

(a) As much as you feel like showing, really. I'd say there's no need to overdo it. Sometimes less is more.

(b) Hmmm, I was thinking of a video file (quicktime?), but honestly, again, anything you are comfortable or pleased with would be best.

As an example to both (a) and (b), I once "recorded" a GIF of me compiling Mini vMac from beginning to end to show everyone how easy it is and that it is not anything "scary", and that anyone can do it:

http://macintoshgarden.org/sites/macintoshgarden.org/files/screenshots/Compile_Run_Mini_vMac.gif

(From this Macintosh Garden page)

FYI. I use one specific software for programming HAM radios that relies heavily on Python. The consensus there is that one has to use nothing newer than Python 3.10, because later versions have bugs/compatibility problems.
I'm using MacPorts version 3.10.5 on Yosemite.

As to the mouse and sound issues, did you try to zap PRAM, remove battery and also press the CUDA button? Is everything OK after these procedures too?

EDIT. Another question. Does your ROM file have ATI Via driver integrated into it?

It seems they removed some HQX-related tooling in such later Python versions, yeah. Perhaps others things, too. Can't say I'm a fan of Python and nor of their decision-making, both recent and old.

I zapped NVRAM with the "PRAM" keyboard shortcut, only. Although from other people's testings, it seems zapping isn't really a requirement to reproduce the system sounds and volume control afterall. Something else triggered it (although still NVRAM-related). Some System 7.5.x installer I ran previously, I think from another System 7 system, seems to be what has set it off and enabled it, from what I could tell.

I intend to figure out how to dump my current NVRAM state so that we can get to see what exactly is allowing these to work. Then, perhaps, we might get something closer to being worthy of a "v10" Mac OS 9.2.2 disc for the minis.

About the ATI Via driver, AFAIK, it doesn't come integrated into the ROM. As a ROM that has the patches seen in the "v9" disc, though, it does contain some ATI-related patching (the NDRV?). The exact ATI patching present is in the end of the "macmini.py" Python script: https://github.com/elliotnunn/tbxi-patches/blob/master/macmini.py

Interesting! So far I haven't been able to reproduce this.
To confirm, do you have sound coming from the Mac mini's internal speaker?

Oh, no, sadly, I did not get that to work. To be sure, I just retested that with the mini's USB audio adapter, that I always have, unplugged before booting it fresh, and the internal speaker is still completely mute.

What I meant was 2 sound-related things started working:

- System Sounds: For example, when you get an alert window, if you click outside its window, a regular Mac OS system will beep at you if it's not on mute, with whatever you have configured as your system sound ("Indigo", "Sosumi", "Beep", "ChuToy" etc.). Likewise, if you get an oldschool file select prompt, and you decide to click outside, some programs are coded to beep at you if you do, and that's exactly what we also get now on the mini. There is one exception to this, though: There's also supposed to be a beep when an alert window itself pops up, but that still won't play, curiously enough.

- Overall volume control (via the Control Strip and Sound Control Panel): Before, moving the volume slider did not save its position in the slider, and the top-bar menus would flash blue for a moment, which is what Mac OS does when a system sound is supposed to play, but the system is on mute, as if we had just selected mute each time (despite the slider bar remaining in the middle in the Control Strip IIRC). This is no longer happening: Now I pick a volume via the slider, the system then beeps accordingly with a loudness matching the volume setting (or it doesn't beep at all, if I chose the lowest setting, which mutes the OS), then, in the case of the Control Strip, when I reopen the slider, we see the current volume slider position is still in the position we previously chose. ALL sound output is now louder/quieter accordingly, as it should be.

(Then there's the mouse freeze bug during boot, which I still have to encounter since.)

To get the internal speakers working, we will have to tweak the ROM scripting further, so that we either tweak Apple's sound drivers to somehow cooperate with the existing speakers in the mini, OR write brand new drivers for it at worst, I'd suspect. I think we cannot unfortunately dodge meeting one of these requirements with Rairii's patching alone.

I'm afraid that I've done that. Subjectively the mouse freezing issue seems to occur less often. I rebooted 15 times after I reset the NVRAM before it happened. But it did happen.

Also the sound is still muted. I even rebooted into 7.6, changed the volume via the Monitors and Sound control panel, then rebooted into 9.2.2. Still the same. But if you have another tests to do then let me know and I will. Thanks!

Thanks for checking all that again. I guess there's still some step of the puzzle I missed out on documenting... When I think back to the things I did up until the point I noticed I had system sounds suddenly, I remember I was installing various System 7.5 ~ 7.5.3 versions, then booting into them, and perhaps even installing them at one point from within another System 7 instance? Or... did I always install from Mac OS 9?

So, I really ought to capture the state of my NVRAM somehow while it lasts... This way whatever it is doing right, it can be transferred. (E.g. specific values for sound-related varibles stored in it.)

The hard way would be to get System 7.5.x installs / booting to trigger these sound features again, and pin down which install and install procedure exactly did it. That will be the last resort.

Same for the mouse freeze... There is something preventing it from happening in my machine entirely now, it would seem... (Could the mouse freeze bug, all along, have a correlation with the other sound bugs I described? Meaning that if those sound features are in order, so is the mouse movement...?)

BTW I did notice that in the Sound control panel in the INPUT tab if I toggle on SIGNAL METER it locks solid. Not sure if that is new.

I can guarantee this was always the case: anything involving a microphone / sound input has always caused a hard crash or failed to work on the mini under Mac OS 9.2.2, for all 4 mini models, documented in the vastness of the original Mac mini thread somewhere.

I just rechecked it on my mini in its current state, and I can see that behavior remains unchan--- WTH?! It's actually working now on my end! What the?!

On the "Input" tab, I see I have an "External Mic" option whose device is marked as "USB Audio", which I can click on to be my input source, and if I click on "Check signal level", it doesn't freeze, but actually does as it advertises. (Meanwhile, "Play sound through output device" is greyed out, which I guess makes sense?)

So I went ahead and started digging through my old stuff to see if I had an old style microphone I could plug into the USB audio adapter and... it... works? As in, the OS didn't crash, the "Sound" Control Panel didn't crash, and when I picked the "Alerts" tab and then clicked on "New Sound" and then recorded new audio, it actually captured the sound input! All of this would previously freeze or crash the OS, as far as I'm aware!

Thing is, however, the volume of the sound was extremely low, even if I was extremely loud (being loud helped a bit, though). So I decided to search for another microphone candidate I'd have immediate access to...

... which I did. At first, it seemed to take no sound, so I rebooted. After that, it'd register the audio, but... the audio was just some random kind of static! Still "worked" in that nothing crashed and a sound file generated, but that wasn't sound from the outside per se like that...

So now I ordered in an ultra cheap microphone that should be a bit more appropriate for the task. Once it arrives, I will try again later, to see if any of them will record audio properly.

I really would love to figure out how exactly this "sound features medicine" was set up in my mini (almost guaranteed to be NVRAM, as I touched nothing in my Mac OS 9 install). Or, at the very least, how I can "capture its state" so that we can manually set it on any Mac we want (e.g. NVRAM script).

Oh, for the record, I also went ahead and retested the "External Mic" option of the "Built-in" type, and that doesn't get to record any audio at all, still. So just how the internal speakers still don't work, the internal mic also doesn't. At least there are no crashes normally if I try using them, though at one point I got the Control Panel to crash the Finder in a way I had to reboot, and an actual bomb which also required a reboot. (But I never got any of these with the "USB Audio" source for the mic, now that is stable OS-wise.)

Hey, welcome, Stephen!

Thanks... but I do have to make a small correction there. You're not the first to overlook it, but my name is an alternative spelling of "Stefan", not of "Steven". ("a" and "e" just look too similar at small sizes.)

That was a flawless introduction!

In hindsight, it felt a little too much like a self-centered ramble to me. All repetitive "I..." "I..." "I...".

And yes, OS X, to me as well, is just a UNIX box, and not Mac OS per se.

I'd also say the GUI layer is trying too hard to be an appliance to be a good UNIX box. While they had different philosophies for what to prioritize, classic Mac OS and proper UNIX are both designed around giving power users the freedom to customize their systems.

(As Eric S. Raymond's The Art of UNIX Programming pointed out, the UNIX ecosystem was so focused on good architecture that their UI design was neglected, while the classic Mac ecosystem was so focused on good UI design that elements relating to stability and robustness were neglected.)

And indeed, Mac OS' modularity which encourages more user autonomy, ownership and tinkering is definitely among the many of the highest distinctive qualities the OS has to offer.

It's not just the modularity. It's little details that other platforms never caught up to, like how much more polished "Hold shift to start with extensions disabled" is than Windows Safe Mode or UNIX's single-user mode, how easy it is for even a novice user to disable or uninstall problematic system extensions as a side-effect of not shying away from having users see and interact with files, etc.

Glad to have you participate here on "Mac OS 9 Lives!" and thus also in the Mac OS communities at large! Looking forward to seeing what you might end up working on for Mac OS!

We'll see. Right now, I'm still trying to climb out of the backlog that my hobby stuff fell into when the psychological and social effects of the pandemic had everything fall apart, so all I've really accomplished that's suitable for sharing is the mostly-finished set of custom icons for the Samba/Netatalk share I use to make it easy to load stuff onto a new/wiped hobby machine.

Here's a screenshot from before I added a Mac_68k folder since I don't have a newer composite. (It leverages that Windows 9x/XP and classic Mac OS use smaller icon sizes than Windows Vista/7/... and OSX to present different styles of icons, and leverages how Samba will use a different serialization of the Icon\r filename than Netatalk without vfs_fruit to serve up different icons to Classic Mac OS and PPC OSX vs. Intel OSX. The PPC OSX one is just a proof of concept at this point.)