Jump to content

AnV Chameleon boot loader


Andy Vandijck
 Share

136 posts in this topic

Recommended Posts

As I remember the primary idea of EFI on PC comes from Netkas but not chameleon group.

 

You remeber terribly wrong. The fake EFI implementation is David Elliott's work, netkas just grabbed his sources, added device-properties support, GPT partitioned disk support, then he called it as PC_EFI :( Chameleon is also based on dfe's work, but such improvements on boot1h to able to use /boot in your root folder is totally written from scratch. And there's more coming...

Link to comment
Share on other sites

@ this point and most likely from here on out...everyone wishes you luck along with all your fanboys.

 

No more negative emotions of sentiments will be found here by anyone in the development camp and the devs should return underground.

 

Good Luck, anv.

Link to comment
Share on other sites

OK guys, give me a bit of time I'll post updates soon...

Link to comment
Share on other sites

Thanks Andy, its working here. I just want to know how to check whether my DSDT.aml file, which is on my / is working or not. Because I want to update to 10.5.6 and I don't want to screwup my hack.

 

I would also like to say, that don't bother about the negative inputs people are giving, just continue your good work as there is a saying that you should have foes beside you, they always inspire you more.

 

Please do reply to my question, Thanks again.

Link to comment
Share on other sites

OK guys, give me a bit of time I'll post updates soon...

Thanks !

But please ANDY tell me ( i asked some posts above) if i can copy only the boot file (of your files) into / to overwrite my chameleon with dsdt boot ?

I have installed chameleon with dsdt already (using only the boot file) over chameleon v11 (which alreadx installed that bootxy files)

Or do i must dd and startupfile commands to to overwrite that bootxy files also ?

Link to comment
Share on other sites

Thanks !

But please ANDY tell me ( i asked some posts above) if i can copy only the boot file (of your files) into / to overwrite my chameleon with dsdt boot ?

I have installed chameleon with dsdt already (using only the boot file) over chameleon v11 (which alreadx installed that bootxy files)

Or do i must dd and startupfile commands to to overwrite that bootxy files also ?

 

Forget startupfile, and put your boot into your root folder where boot1h got dd'ed to. The startupfile loading method won't be supported with next versions of boot1h, but can load 64k+ boot files in advance.

Link to comment
Share on other sites

Forget startupfile, and put your boot into your root folder where boot1h got dd'ed to. The startupfile loading method won't be supported with next versions of boot1h, but can load 64k+ boot files in advance.

Thanks - i did think same, but wanted to know it from an professional to avoid problems :P

So i can now test it.

Link to comment
Share on other sites

I'll be adding all the stuff soon.

I'm pretty busy, but I'll get there.

Again... patience... ;)

Link to comment
Share on other sites

I'll be adding all the stuff soon.

I'm pretty busy, but I'll get there.

Again... patience... ;)

 

I did not tested your bootloader yet !!! But i'am actualy using PCEFI v9 and booting an Apple Raid Volume.

I've noticed that efi v9 isnt supporting Apple Raid natively as Chameleon 1.0.11 was !

 

So first question : does your bootlaoder support Apple Raid ?

 

PCEFI needs a kernel + /S/L/Extensions.mkext + /L/P/SystemConfiguration/com.apple.Boot.plist + /Extra/Extensions.mkext + /dsdt.aml on the "Boot OSX"

partition to boot properly. Chameleon 1.0.11 can natively use the Apple Raid file system migrated in the "/Boot OSX/com.apple.boot.R" folder by Apple Raid.

I'm actualy using an home made Startupitem script to update files to the right place for PCEFI to use them. That's working fine hopefully !!!

 

So second question : Do you think you can add that to your bootloader ?

 

Thanks.

Link to comment
Share on other sites

Guest BuildSmart

Very well done, nice to see you took the time to fix things while adding the enhancements, unfortunately the community doesn't place a high value on your work which is a shame, this makes for a far better code-base to work from than the chameleon source.

 

The loss of features make it less than desirable to most and this is due to their ignorance more than anything else and it is nice to see someone releasing their work without too much of a fuss.

 

netkas states he is going to release his source but after more than a month and a half of waiting it seems it's just more of the same un-kept promises, zef on the other hand does release his work however the work released is so old that trying to examine his current released binary code-base has become nothing short of problematic so fixing little issues without sacrificing any functionality is pretty much impossible.

 

This brings us to David E., while the pioneer in this area, he has made little progress since his initial release but his work is the basis for all that we now have so if anyone deserves credit I would have to say it is him.

 

There is no need to credit chameleon in the work itself, it may be based on the same concepts but the implementation isn't the same, your code is much cleaner.

 

My only objection to this type of work is the use of hard coded system information, as a suggestion, I would like to offer the following concept, test for and load /usr/standalone/i386/info.dat if it exists to force specific system information to be loaded rather than hard-coding, this wont corrupt DMI modified setups and allows for greater flexibility and yes, a system that supplies valid SMBIOS information seem to get all messed up because the extension still tries to load the valid SMBIOS information over the hard-coded.

 

One more thing to note, unless you have a Xeon processor you will never have a MacPro compatible or clone and it has been suggested that this will cause problems with future updates so if you want to hard code system information, try coding something more in line with the hardware it is being run on.

 

Although, it is good for a laugh when someone brags about running leopard on their MacPro clone but has no real clue what hardware comes in a MacPro, and think that running kalyway is the best setup ever.

 

Again, thanks for your contribution and a job well done.

Link to comment
Share on other sites

Andy, great work! I'm looking forward to you adding support for /Extra and possible support for the Munk EFI partition method.

 

Hey Andy,

 

since a while, you are seeling beef from others, in another envelope

pretending to be the great butcher.

 

Chameleon is open source. Anybody can do whatever they like with the code as long as they release the modified code. Andy has done that.

 

So, westwaerts, what exactly have you done for the scene?

Link to comment
Share on other sites

Forget startupfile, and put your boot into your root folder where boot1h got dd'ed to. The startupfile loading method won't be supported with next versions of boot1h, but can load 64k+ boot files in advance.

 

This is what I am truly looking forward to.

Link to comment
Share on other sites

Very well done, nice to see you took the time to fix things while adding the enhancements, unfortunately the community doesn't place a high value on your work which is a shame, this makes for a far better code-base to work from than the chameleon source.

 

The loss of features make it less than desirable to most and this is due to their ignorance more than anything else and it is nice to see someone releasing their work without too much of a fuss.

 

netkas states he is going to release his source but after more than a month and a half of waiting it seems it's just more of the same un-kept promises, zef on the other hand does release his work however the work released is so old that trying to examine his current released binary code-base has become nothing short of problematic so fixing little issues without sacrificing any functionality is pretty much impossible.

 

This brings us to David E., while the pioneer in this area, he has made little progress since his initial release but his work is the basis for all that we now have so if anyone deserves credit I would have to say it is him.

 

There is no need to credit chameleon in the work itself, it may be based on the same concepts but the implementation isn't the same, your code is much cleaner.

 

My only objection to this type of work is the use of hard coded system information, as a suggestion, I would like to offer the following concept, test for and load /usr/standalone/i386/info.dat if it exists to force specific system information to be loaded rather than hard-coding, this wont corrupt DMI modified setups and allows for greater flexibility and yes, a system that supplies valid SMBIOS information seem to get all messed up because the extension still tries to load the valid SMBIOS information over the hard-coded.

 

One more thing to note, unless you have a Xeon processor you will never have a MacPro compatible or clone and it has been suggested that this will cause problems with future updates so if you want to hard code system information, try coding something more in line with the hardware it is being run on.

 

Although, it is good for a laugh when someone brags about running leopard on their MacPro clone but has no real clue what hardware comes in a MacPro, and think that running kalyway is the best setup ever.

 

Again, thanks for your contribution and a job well done.

 

Most people don't have the vision to see that forking of projects in the long term is a bad thing. The chameleon project is extremely active and has a number of developer contributing nearly daily to the sources. Taking Chameleon to the next level has required massive internal restructuring and just breaking the 64k limit was a difficult and long process. The chameleon project now is part of the Voodoo team and other than the kernel work there are a number of other projects in development.

 

The long term goal is to create quality alternative implementations of the key foundations of OS X. This was why the kernel was dealt with first and the boot loader is in the process of being made much more modern and expandable. We also have a few other extensions that will resolve some very outstanding and troublesome issues for users. Also some of the smaller extensions which have been required have become integrated into the boot loader (SMBIOS enablers, EHCI ownership problems etc). We also hope to have PCI probing added to allow device properties to be added (to fix problems such as GMA 950 on some systems as well as networking problems - time machine).

 

While the source code for all these projects will be open to everyone, during development we have found that its better to keep the sources closed. The truth of the matter is that most people have no interest in enhancing and fixing problems with OS X. The kernel sources we put together have had practically no interest (and even with a simple building tool to make kernel compiling easy for those looking to learn). The majority of the advancements in OS X running on normal PCs has been done by a very small group of people. I wish I could say it was a community achievement but that really hasn't been the case at all.

 

While its possible for the community to continue to toss out DVD's with a million different options (just look at the new iPC DVD) practically no user will know what option to choose and end up with a system that doesn't work. Is this the direction we really want to go? Alternative extensions with a quality boot loader and kernel will give a much simpler install and a much higher compatibility. If the OS X community wants to grow up, it has to move in this direction. This is what the Voodoo team is trying to do. We hope others are willing to join us.

Link to comment
Share on other sites

The problem, though, in the end, is that the community will eventually move more and more towards trying to create a "vanilla" experience. While I think this is great, the problem becomes people with older hardware, like me. Already with the 10.5.6 update, my obsolete FX5500 has been phased out by Apple. The new iPC DVD aims to create a distro that allows a user to empower themselves with all the greatness of what the "community" has created, in both Vanilla and non-Vanilla scenarios. For the first time, I personally, after helping devd it, had a system that worked 99% OOTB at first boot, the 1% being caused by my aging GFX card with 10.5.6 running with 10.5.5 drivers. Nothing else has enabled that for me, though, and going the "vanilla" route is not possible for me.

 

We can't just grow up with one set direction, we're not all headed the same way. We have to grow forward in all aspect, with both tools of new and tools of old. I think that a lot of people forget this, even the developers who make such great tools.

Link to comment
Share on other sites

Very well done, nice to see you took the time to fix things while adding the enhancements, unfortunately the community doesn't place a high value on your work which is a shame, this makes for a far better code-base to work from than the chameleon source.

 

I'm wondering about the technical details how AnV's code-base can be a "far better" one?

 

There is no need to credit chameleon in the work itself, it may be based on the same concepts but the implementation isn't the same, your code is much cleaner.

 

While there's no question about dfe's achievments regarding the fake EFI implementation and his overall professional developer kind, but your above statement ignores about 3 months of work in all stages in in the boot loader project. How about hybrid GPT/MBR aware boot0? How about the HFS file system loader code found in boot1h? How about Kabyl's proper fsb detection/cpu support? How about foreign OS support for GPT disks? These are all unique developments what you can't find in any other implementations.

Link to comment
Share on other sites

While its possible for the community to continue to toss out DVD's with a million different options (just look at the new iPC DVD) practically no user will know what option to choose and end up with a system that doesn't work.

 

First of all, yes, I created iPC but I am not going to be biased (maybe the last sentence is a bit) but the rest is general.

 

I agree with a lot of what you've said but the above statement makes no sense. A kernel and a bootloader will not fix every problem we have here, especially with support for older systems. The Voodoo kernel has brought a lot of life back into older and less compatible computers but it still doesn't solve everything. And about the user not knowing what to choose, do we really want a person that doesn't even know their hardware running OS X on a PC? This is the root of all the clutter and the useless questions we have on here. I'm not saying to make it harder, but we don't need to try so hard to make it like installing Windows, where you click a button and everything works (I've tried my fair share of attempts at that in the past but it didn't always work), OSx86 is a hobby and its specialty is what makes it great. "Practically no user" is incorrect, 90% of the people who have reported trying iPC have eventually gotten it to work.

Link to comment
Share on other sites

@Dense and Zef

 

Please stick to the technical details.

 

It is NOT advised trying to discuss things rationally in this thread.

 

The admins and mods here don't take kindly to such discussions, I know first hand from when someone "borrowed" the work of JaS and Diabolik to create his own DVD and Nvidia Installer.

 

Since we cannot control the actions of others in these forums we will take our efforts elsewhere -- to a place with far less rules.

As you know I am 100% behind the VooDoo team -- as I have unprecedented access to the people doing the real work.

I have had very long talks with those ladies and gentlemen regarding future visions, plans, and implementations for things to come.

 

If people could just have some of the basics I have been given they would understand just how amazing the Voodoo Camp has become.

 

The future will be the OLD WITH THE NEW and the ETA is near...people must remain calm and focused.

Let's let Andy have his own followers and continue his own work. If his code is superior as BuildSmart claims, let it be so.

 

People will always have favorites and do what they want -- they apply definitions and interpretations in matters of convenience.

I believe it is time for some Public Relations work to be done for the people of OSX86 community from Voodoo...OUTSIDE of this thread!

Link to comment
Share on other sites

I did not tested your bootloader yet !!! But i'am actualy using PCEFI v9 and booting an Apple Raid Volume.

I've noticed that efi v9 isnt supporting Apple Raid natively as Chameleon 1.0.11 was !

 

So first question : does your bootlaoder support Apple Raid ?

 

PCEFI needs a kernel + /S/L/Extensions.mkext + /L/P/SystemConfiguration/com.apple.Boot.plist + /Extra/Extensions.mkext + /dsdt.aml on the "Boot OSX"

partition to boot properly. Chameleon 1.0.11 can natively use the Apple Raid file system migrated in the "/Boot OSX/com.apple.boot.R" folder by Apple Raid.

I'm actualy using an home made Startupitem script to update files to the right place for PCEFI to use them. That's working fine hopefully !!!

 

So second question : Do you think you can add that to your bootloader ?

 

Thanks.

 

 

ANV Bootloader is working fine with Apple Software Raid. Just copy "boot" to both "OSX BOOT"

Link to comment
Share on other sites

 Share

×
×
  • Create New...