Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I liked the Sparc systems until I had to fix the dead RTOS batteries. Did it on an Ultra 5, an Ultra 1 and some old thing with a 50Hz processor, an LX something or other. An absolute pain to do. That and the keying in all the OF strings to reset things, find hard disks and find a workable resolution on non-Sun monitors. No fun.

But NT.. I bet it runs like agony. I want to hear File Mangler scream.
 
I liked the Sparc systems until I had to fix the dead RTOS batteries. Did it on an Ultra 5, an Ultra 1 and some old thing with a 50Hz processor, an LX something or other. An absolute pain to do. That and the keying in all the OF strings to reset things, find hard disks and find a workable resolution on non-Sun monitors. No fun.

But NT.. I bet it runs like agony. I want to hear File Mangler scream.
What type of batteries do the SPARCs use?
 
What type of batteries do the SPARCs use?
Those dreadful Dallas batteries buried within iCs. You have to break into the IC to kill the connection to the dead battery and solder on a new one without breaking the crystal. Luckily, these are socketed in Sun machines but I have done a couple where the Dallas battery was soldered onto the mainboard - one on a no-name German laptop and one IBM i think. Scary stuff as there is no room for error.


NT Sparc looks like a no-go. It's not even been confirmed that one made it out of a "proof of concept" build and if the Betaarchive doesn't have a copy, then nobody has leaked one yet.
 
Last edited:
  • Like
Reactions: Amethyst1
Those dreadful Dallas batteries buried within iCs. You have to break into the IC to kill the connection to the dead battery and solder on a new one without breaking the crystal. Luckily, these are socketed in Sun machines but I have done a couple where the Dallas battery was soldered onto the mainboard - one on a no-name German laptop and one IBM i think. Scary stuff as there is no room for error.


NT Sparc looks like a no-go. It's not even been confirmed that one made it out of a "proof of concept" build and if the Betaarchive doesn't have a copy, then nobody has leaked one yet.
Ok, thanks for the info! That's too bad about NT SPARC, that would be a neat piece of history.
 
  • Like
Reactions: Amethyst1
Now that the source code for Windows XP has been leaked, would it be possible to compile XP for PowerPC Macs?
Try to use AI to port the code and all the drivers and components.

AI become powerful enough to port any code to unsupported architectures and reverse if it's needed. I'm saying this from my own experience.
 
Try to use AI to port the code and all the drivers and components.

AI become powerful enough to port any code to unsupported architectures and reverse if it's needed. I'm saying this from my own experience.
I haven't looked at the XP codebase but I bet it is massive enough that AI will choke on several Intel specific parts. If you are saying in a roundabout way that you have already done it successfully, tell us more.
 
I haven't looked at the XP codebase but I bet it is massive enough that AI will choke on several Intel specific parts. If you are saying in a roundabout way that you have already done it successfully, tell us more.
Well, I meant that I have already done similar porting, but not on such a large scale, since the free models have a very limited context and usage limit, XP would probably require a paid and the most powerful model available, and it would also require porting many programs without which the OS itself would be useless.
Also, as a reference, one could use the source code of Windows 4.0 leaked into the network, which was already on the NT kernel, but at the same time had support for the PowerPC architecture, and just as I heard that there were projects to launch it on the PowerMac G5, which also means that the required drivers are available for this hardware, which means these drivers can be ported to Windows XP, and part of the PowerPC ASM code can also be taken from Windows 4.0 src.

I did porting only small projects - decompiling C# programs like MCCToolChestPE and games (Angry Birds, Fruit Ninja, PvZ, Chess) with Windows Phone 7 (Silverlight Runtime / XNA C#) through ILSpy, then processing through AI to refine the pseudocode into a working source code, while the AI coped even with heavily obfuscated code, which is already quite a lot. After that, I also used AI to port the existing code to Windows 10 ARM32 15035 (a networked ARM 32-bit build) and Windows 10/11 ARM64 (and x86, x64), which had never had native ports before.
Thus, the AI helped me transfer one Win x86 program and several WP7 games to previously never-supported platforms. Obviously, these projects are not comparable in complexity to Windows XP, but this is at least the equivalent of a very successful porting to another architecture using AI.
 
Those dreadful Dallas batteries buried within iCs. You have to break into the IC to kill the connection to the dead battery and solder on a new one without breaking the crystal. Luckily, these are socketed in Sun machines but I have done a couple where the Dallas battery was soldered onto the mainboard - one on a no-name German laptop and one IBM i think. Scary stuff as there is no room for error.


NT Sparc looks like a no-go. It's not even been confirmed that one made it out of a "proof of concept" build and if the Betaarchive doesn't have a copy, then nobody has leaked one yet.
Sounds just like the suicide batteries in Capcom CPS-II arcade PCBs; at least from the "margin for error" standpoint. The batteries backed the encryption IC, and once they died (the power drops below 2v), that was it for the game. I think the batteries for these were outboard, but I don't remember without checking (just checked- they are outboard). Nowadays, thanks to MAME developers and others, the games themselves have been decrypted and they can now fix those PCBs just by burning new ROMs with the decrypted program code.
 
Last edited:
  • Like
Reactions: TheShortTimer
Also, as a reference, one could use the source code of Windows 4.0 leaked into the network, which was already on the NT kernel, but at the same time had support for the PowerPC architecture, and just as I heard that there were projects to launch it on the PowerMac G5, which also means that the required drivers are available for this hardware, which means these drivers can be ported to Windows XP, and part of the PowerPC ASM code can also be taken from Windows 4.0 src.
I see. The bit about the PMG5 is interesting. Lately, we have had success in getting the MS PPC ports of NT4 and NT3.51 to boot on G3 hardware thanks to an existing bootloader that got successfully tweaked to fit Apple's hardware. If such a thing exists in any form for the G5, then it makes the job of porting that much easier. MS never initially intended Windows PPC to boot on anything but a limited selection of IBM PPC hardware.
 
  • Like
Reactions: MaxRM
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.