Any one want to help build a community made OS 9.3?

[MacOS Plus]

WOW !  You just added another model to the list !

I know you are very busy, but when time permits, can you make a "Step by step" or "How to" guide for that model even if it relies on OF commands ?  we should re-post this info at https://www.thinkclassic.org/ and let iMic know

No problem. I have a funny feeling that any G4's that are not working will work with this fix with the exception of the mini and the xserve.  I could be completely wrong though.

  Well, at least there's no integrated sound device in the Xserve to contend with!    (That said, I've successfully used USB audio devices with the Xserve in 10.5.)

Booting from the Unsupported G4 disk will probably hang.  The ATI extensions cause a hard lock up where you can't even drop into a debugger which is triggered by a hardware interrupt.
You will have to set the iBook into target disk mode and install everything to the disk and make sure no ATI extensions (this includes the OpenGL and Quicktime ATI extensions).  Other than that so far it looks like almost all the common extensions that would come with a standard install are not causing issue.

  When I first restored from the "Unsupported G4" disc to a FW800 MDD it booted fine the first time but then froze-up the next boot.  It turned out that at least one of the ATI extensions that applied only to much older hardware caused this freeze.  Removing it made everything work correctly.  (I believe this was when using my Radeon 9800 card.)

  Not sure if this helps your particular endeavor but it might be a contributing factor.  Also try to make sure you're testing with the final ATI drivers that are later than what is on the CD.  I'm really hoping in future that there will be a better way to have a system restore or installation detect what video hardware is actually present and only copy the needed extensions.  I've had numerous related issues over the years with particular combinations of ATI driver version versus OS version versus hardware version.  (Even worse a history of ATI bugginess in the Windows environment too.)

  Even if the 9200 ROM version in that machine isn't recognized by the ATI drivers it still might crap out on one of the older extensions.  What you mention sounds correct though - that the acceleration extensions fail because a suitable driver never loads first.  There's gotta be some sort of workaround given that there were OS 9 supported 9200-based cards, both stock and flashed.  Someone wrote on The Mac Elite site that he'd created a custom driver for some variants of them.  One way or another it should be possible to get that chip to work.  (The ATI ROM is probably flash-updatable in that machine, though I'm sure that's a risky proposition to mess with.)

[nanopico]

  When I first restored from the "Unsupported G4" disc to a FW800 MDD it booted fine the first time but then froze-up the next boot.  It turned out that at least one of the ATI extensions that applied only to much older hardware caused this freeze.  Removing it made everything work correctly.  (I believe this was when using my Radeon 9800 card.)

  Not sure if this helps your particular endeavor but it might be a contributing factor.  Also try to make sure you're testing with the final ATI drivers that are later than what is on the CD.  I'm really hoping in future that there will be a better way to have a system restore or installation detect what video hardware is actually present and only copy the needed extensions.  I've had numerous related issues over the years with particular combinations of ATI driver version versus OS version versus hardware version.  (Even worse a history of ATI bugginess in the Windows environment too.)

  Even if the 9200 ROM version in that machine isn't recognized by the ATI drivers it still might crap out on one of the older extensions.  What you mention sounds correct though - that the acceleration extensions fail because a suitable driver never loads first.  There's gotta be some sort of workaround given that there were OS 9 supported 9200-based cards, both stock and flashed.  Someone wrote on The Mac Elite site that he'd created a custom driver for some variants of them.  One way or another it should be possible to get that chip to work.  (The ATI ROM is probably flash-updatable in that machine, though I'm sure that's a risky proposition to mess with.)

After testing more last night, the ATI drivers aren't the only thing that can hang it up.
For one the Multiprocess Apple CPU stuff causes a lock up (no real surprise there).
The HID extension disables the trackpad (kind of funny actually)
Some of the open transport stuff causes a freeze.
And there is few more that I haven't yet identified that cause a freeze that are on the cd.
So I'm still working through the extensions to see what is going on.

I'm going to continue details on the iBook on this thread
http://macos9lives.com/smforum/index.php/topic,2744.0.html

[MacOS Plus]

  I think it would be very helpful if the Multiprocessing file wasn't installed in the system folder by default.  It would be better to put it somewhere else first so the user can manually move it later if they know they need it.  It seems in most cases of DP machines the stock MP file doesn't work at all anyway so it probably could be dumped entirely.  The hacked one I have that works with my Sonnet dual 1.8GHz MDX module should also be included on the disc but also not in the system folder by default.

  For the most part this practice would likely be best for the ATI drivers too since you'll get basic display without them present.  The extensions should be installed to a separate folder with instructions on which ones to manually copy to the system folder depending on the users particular ATI hardware (if scripting this to occur automatically were to be too difficult).  Like I said, there's no point in copying any of the extensions for very old built-in video chips, nor for Rage if using a Radeon card like in my case where the Rage extension's presence froze the machine on second reboot after system restore.

  I guess this sort of stuff will come with time as the project gets refined, along with the modified ROM to eliminate the OF commands on many machines being included in the SR image.  Hopefully we can get to that stage soon so a lot more users can get involved in the testing more easily.

[MacTron]

  I think it would be very helpful if the Multiprocessing file wasn't installed in the system folder by default.  It would be better to put it somewhere else first so the user can manually move it later if they know they need it.  It seems in most cases of DP machines the stock MP file doesn't work at all anyway so it probably could be dumped entirely.  The hacked one I have that works with my Sonnet dual 1.8GHz MDX module should also be included on the disc but also not in the system folder by default.

  For the most part this practice would likely be best for the ATI drivers too since you'll get basic display without them present.  The extensions should be installed to a separate folder with instructions on which ones to manually copy to the system folder depending on the users particular ATI hardware (if scripting this to occur automatically were to be too difficult).  Like I said, there's no point in copying any of the extensions for very old built-in video chips, nor for Rage if using a Radeon card like in my case where the Rage extension's presence froze the machine on second reboot after system restore.

  I guess this sort of stuff will come with time as the project gets refined, along with the modified ROM to eliminate the OF commands on many machines being included in the SR image.  Hopefully we can get to that stage soon so a lot more users can get involved in the testing more easily.

This was already done. All of the ASR images for unsupported hardware and some of the Rescue CDs that included the famous "Mactron's System Folder"  have the Multiprocessor folder out of the System folder.
Eventhroug I don't recomend disabling this extension unless it causes serious issues,  because it handles the CPU energy savings.
Furder more, my last edition of the Mactron's Rescue & Install CD included the iMic/nanopico's ROM as optionall install. I have to update this info on the proper thread.

About the ATI and nVIdia issues, there is no easy way to determine what file include the drivers for each card. IIRC the Radeon 8500 file enable 3D acceleration in a wide number of R2x0 cards.

Any way, for testing purposes, it is allways advisable to try to boot with a ROM, Finder and System files only.As nanopico suggested.

[MacOS Plus]

Cool.  I didn't know you had done some of that already.

  My concern about the ATI drivers being present as an 'all-or'nothing' proposition is that having all of them present by default can actually be worse (freezing) than leaving them out so the user can move them manually.  Like I said I was more than a little surprised to find my MDD freezing on the second boot because of the presence of the Rage extension which was redundant with my card.  Obviously the CD should still have them but the ASR copying all of the ATI extensions blind was an issue once I was booting from hard drive.  To be clear, I'm really only implying the implications of what the ASR does when creating a new hard drive.

  Perhaps an 'expert' in extension optimization could chime in here with respect to the extensions required for certain cards.   I'm wondering if we could start another thread and get feedback from the members on how they've optimized their own ATI configurations so we might put some kind of chart together.  I've never found one good central place of reference for that information.

  I look forward to reading the update about your CD image!

[nanopico]

Learned something new in my journey today.
I think it was asked how to delete a node from the device tree in open firmware.
I had provided some instruction somewhere on how to sort of disable and it would work for most things, but depending on how the interrupts are set up in hardware I found it can not work well.

If you can delete the node then it clears out those dependent interrupts so things are happy.

so with out further delay's here's how you can delete a node (and it deletes all the children below it as well)

I am not responsible for what you choose to do with this information.

So let's say we want to delete the mac-io node (fyi this will pretty much make your machine do pretty much nothing after this until you restart) located at /pci@f2000000/mac-io

Code: [Select]

" /pci@f2000000/mac-io" find-package drop delete-node

then just change the path to the node you want in the first part.  The space between the first " and the / is important.

[MacOS Plus]

Does this finally mean less typing?

[MacTron]

Obviously the CD should still have them but the ASR copying all of the ATI extensions blind was an issue once I was booting from hard drive.  To be clear, I'm really only implying the implications of what the ASR does when creating a new hard drive.

Although The ASR method have some benefits, it isn't all time suitable. I don't even use it.

  Perhaps an 'expert' in extension optimization could chime in here with respect to the extensions required for certain cards.   I'm wondering if we could start another thread and get feedback from the members on how they've optimized their own ATI configurations so we might put some kind of chart together.  I've never found one good central place of reference for that information.

I'll make a new thread to try to collect  info about this, I'll start with what I know about .

  I look forward to reading the update about your CD image!

Here you have :
http://macos9lives.com/smforum/index.php?topic=3177.msg20361#msg20361

[nanopico]

Does this finally mean less typing?

Yes now that i figured that out.

[nanopico]

Not that we don't still have some issues with machines once they are booted, but I do believe the only one we haven't gotten to a desktop yet is the Mac Mini. (correct me if I'm wrong here).
I have one on order will be here sometime this week.
So hopefully I can crack that one too.

[nanopico]

So after a fairly productive weekend getting somethings working, I've got an update list of items I'm targeting.
The goals for OS 9.3 (at least that's what I'm calling it) are the following.

• Remove the 1.5 GB RAM limit
• Create drivers/get hardware working on machines we have now got booting.

Might not be a long list of features, but we don't have a huge amount of resources and I think those these would get us pretty far.

After that I might tackle updated graphics drivers. 

[Texas_RangerAT]

All I can say is you've all been doing some incredible work! If there's any way I can help you I gladly will!

The non-supported machines I have access to are:
- iBook G4 1 GHz (Powerbook 6,5)
- two MDD FW800 Dual 1,25 GHz (although one seems to have a broken SuperDrive)
- a Gigabit Ethernet with a Sonnet 1.8 GHz upgrade (it isn't mine so I'd have to look up the exact details), I think it's currently running 10.4 only (installed on a Quicksilver, transplanted to the Gigabit Ethernet by simply swapping the hard drives, the QS is now happily running the 9.2.2 with any Sonnet mods)

All my other machines are either Intel or officially supported (QS, MDD 1,25 GHz 2003 FW 400, B&W G3, beige G3 etc.). BTW, one of the MDD FW800s has 2 GB of RAM if that helps in any way (i.e. testing any solutions for the 2 GB issue).


Regarding an older post on the availability of 68k hardware:

a quick search on ebay germany gives you 220 results for "mac pro", 15,000 results for "apple g4", and 0 for "apple 68k".

Well that's obviously because sellers don't call them 68k. A quick search for specific models unearthed a Classic, an LC, an LCIII and a bunch of other stuff, all next to free except the - supposedly - professionally serviced Classic. The eBay classified ads offer even more, although sometimes for ridiculous prices (250 Euros for a Performa 630? Are you kidding?).

[LillyOmegaWolf]

I think this is awesome!
I'm happy to test anything. It would be great to see OS9 become more than what it is as I think its a great old OS. I got a few G4s and one G3 iBook clamshell all of which I'm happy to play around with as my work system is OSX Sierra (for FCPX).

[devils_advisor]

quote author=nanopico link=topic=2727.msg20905#msg20905 date=1466362731]
So after a fairly productive weekend getting somethings working, i've got an update list of items I'm targeting.
The goals for OS 9.3 (at least that's what I'm calling it) are the following.

• Remove the 1.5 GB RAM limit
• Create drivers/get hardware working on machines we have now got booting.

Might not be a long list of features, but we don't have a huge amount of resources and I think those these would get us pretty far.


After that I might tackle updated graphics drivers.
[/quote]




I saw somewhere that the os7 sources where either leaked or released.  Would it be a good source to study that or is it to old to be useful?
Just thought the core is almost the same except addons over time.

[nanopico]

quote author=nanopico link=topic=2727.msg20905#msg20905 date=1466362731]
So after a fairly productive weekend getting somethings working, i've got an update list of items I'm targeting.
The goals for OS 9.3 (at least that's what I'm calling it) are the following.

• Remove the 1.5 GB RAM limit
• Create drivers/get hardware working on machines we have now got booting.

Might not be a long list of features, but we don't have a huge amount of resources and I think those these would get us pretty far.


After that I might tackle updated graphics drivers.

I saw somewhere that the os7 sources where either leaked or released.  Would it be a good source to study that or is it to old to be useful?
Just thought the core is almost the same except addons over time.
[/quote]

In theory it would be good as a lot has not changed.  It also might have change history in it so a person could see the changes made over time and the logic so that a person could also infer how it progressed to today.  But that's assuming "one has looked at it". 
I would say that would be not right as the source is proprietary, and what not, but then again? 
Reality I have seen it.  It is very interesting indeed.  It has confirmed some of my suspicions and raised more questions. But it is a very interesting.

[ck]

@Nanopico and OS 9.3 team:

THANK YOU for inspiring the rest of the Mac Classic holdouts to keep re-inventing the old into new, or should I say newer 

I may have next to zero-programming skills, but I enjoy everything mac-related from 7 to 9. I've set up an off-shore call center using Mac laptops running 7.x to X just eleven years ago and even now I'm very happily emulating 8.1 on both my chromebook and smartphone.. heck, I'm even skinning all my classic OSes in Sierra-themed icons and windows, lol!

All of your efforts are COMPLETELY appreciated by Mac lovers the world over so please include me in the long list of fans!!!

Finally, just a mini-interjection that may hopefully be of some use in your AWESOME endeavors:

While not endorsing to do so exclusively, reverse-engineering a few critical select pieces of the unsupported hardware puzzle, which looks like is the heart of this project, may very well be used constructively to learn more about what can be made to work. So, say some rogue hacking gets the job done in a quick-and-dirty way, having these working elements can shed some new light on what (and where) to tinker with. Then the real work of re-achieving successful exploits really begins, at the very least with the revealed knowledge that it is definitely doable because it's already been done in a rogue fashion.

Ps. and if anyone wants to go Sierra on their old systems, purely for kicks of course, let me know and I can send over a cool little 'unfinished' gui package!

Pss. if it helps, I do have a 17" Powerbook G4 that I believe doesn't boot natively into OS 9 to test future compatibility-enhanced OS 9 builds and rom/boot tools.

[Daniel]

Hi Everyone,

You can probably already tell this because of my custom ROM file hacking, but I'm in!
I officially want to help develop 9.3! I am currently only good at forth and open firmware, but I can learn quickly and I have too much time on my hands. I will mostly only be able to ask interesting questions about non Open Firmware stuff, but I am willing to try out any crazy experimental software you make 

[nanopico]

Hi Everyone,

You can probably already tell this because of my custom ROM file hacking, but I'm in!
I officially want to help develop 9.3! I am currently only good at forth and open firmware, but I can learn quickly and I have too much time on my hands. I will mostly only be able to ask interesting questions about non Open Firmware stuff, but I am willing to try out any crazy experimental software you make 

You will be useful with that knowledge.  It's something we need some help with. I have worked a lot to learn it and have a pretty good knowledge of it now, but it seems you have some additional useful knowledge to add.

[purakee]

I bought an iBook G4 A1133 1.33 GHz (PowerBook6,7) and an iBook G3 800 Mhz A1005 last gen. (Early 2003, PowerBook4,3) a while ago. Now, when I looked at the compatible property for /pci@f2000000/mac-io/i2s/i2s-a on the iBook G3, the value was "i2sbus," which is the same as the iBook G4. This property seems to hang many of the iBook G4s and Aluminum PowerBook G4s without the "fix" that Nanopico posted; however, those G4s don't have sound on Mac OS 9. The iBook G3 Early 2003 has a custom CD for installation, and therefore, a custom Mac OS 9 system, and the Sound Manager recognizes the "i2sbus" without a problem on the G3 system.

So my main point is that if we restore the OS9General.dmg file from the iBook G3 Early 2003 disc to the hard drive on the semi-functioning iBooks and PowerBooks, along with the custom ROM, we could make sound work on those systems! Unless there is another impediment in the properties of the sound device on the G4s, this should work. I have attempted this on the iBook G4 with the custom ROM, but the hard drive became corrupt, so I cannot even boot into Mac OS 9. I do not have access to my PowerBook G4 right now, but when I do I will try this on it.

Could anybody with a G4 that requires Nanopico's sound card hack try this hack as well?

Thank you so much, and I am so grateful for the progress we've made since 2015!

[nanopico]

I bought an iBook G4 A1133 1.33 GHz (PowerBook6,7) and an iBook G3 800 Mhz A1005 last gen. (Early 2003, PowerBook4,3) a while ago. Now, when I looked at the compatible property for /pci@f2000000/mac-io/i2s/i2s-a on the iBook G3, the value was "i2sbus," which is the same as the iBook G4. This property seems to hang many of the iBook G4s and Aluminum PowerBook G4s without the "fix" that Nanopico posted; however, those G4s don't have sound on Mac OS 9. The iBook G3 Early 2003 has a custom CD for installation, and therefore, a custom Mac OS 9 system, and the Sound Manager recognizes the "i2sbus" without a problem on the G3 system.

So my main point is that if we restore the OS9General.dmg file from the iBook G3 Early 2003 disc to the hard drive on the semi-functioning iBooks and PowerBooks, along with the custom ROM, we could make sound work on those systems! Unless there is another impediment in the properties of the sound device on the G4s, this should work. I have attempted this on the iBook G4 with the custom ROM, but the hard drive became corrupt, so I cannot even boot into Mac OS 9. I do not have access to my PowerBook G4 right now, but when I do I will try this on it.

Could anybody with a G4 that requires Nanopico's sound card hack try this hack as well?

Thank you so much, and I am so grateful for the progress we've made since 2015!

I have both those iBooks as well.  The i2sbus compatibility doesn't seem to make much of a difference.  That is a generic identifier for things with early boot. As the machine boots, it does some other checks to determine the exact device.  i2s is just the communication protocol.  Not the actual communication instructions on how and what to do. The G4 iMac have i2sbus as well.  When the sound manager starts up it is seeing it as a generic sound device when it isn't so the initialization sits in an infinite loop.  Your theory may work and it's worth a try, but the base install of my G4 when I tested it was done from the G3's restore disc (which would contain the needed info) and the custom ROM.
It is really worth a try. I can try it, but I'll be honest and say that it won't really happen for several months. I get too busy with other things at the moment.