Hacking Official Corbenik - Another CFW for advanced users (with bytecode patches!)

[chaoskagami]

I always update every file of corbenik, it still gave me that strange reaction, but skeith worked well (Maybe it was a compile problem(?) but it didn't show any error when comiling)

If you compiled on windows, then there's a known issue where for whatever reason the paths in the binary get fucked up somehow and end up msys paths like z:\msys/corbenik/... So don't compile on windows. I've said this many times. I will not go out of my way to fix windows either. Use a VM.

What the locale folder have? What it does?

Readme.

The folder `/corbenik/locale` is automatically generated language emulation files from 3dsdb for games that only specify one region and one language. Games which support more than one language are not generated, because there's no 'correct' language. You can remove this if the number of files unnerves you. It isn't required. You can also add new files if you have specific needs.

Can someone test a thing for me? Install Super ftpd II Turbo .cia on your 3DS and see if corbenik shows the app icon
For some reason it does not appears on mine, but appears with Cakes and Luma, see if you can reproduce this bug(?)

Nothing to do with corbenik, I can say that much.

Also, y'all are gonna hate me for this, but I'm dropping the weird folder names in favor of a unix-like prefix structure. Next release will require a manual reinstall.

 

[Gray_Jack]

If you compiled on windows, then there's a known issue where for whatever reason the paths in the binary get fucked up somehow and end up msys paths like z:\msys/corbenik/... So don't compile on windows. I've said this many times. I will not go out of my way to fix windows either. Use a VM.

I'll never use Windows as my primary OS

Thanks for the info

 

[chaoskagami]

I'll never use Windows as my primary OS

Thanks for the info

Assuming you didn't compile on windows, I'm afraid I have no clue as to the problem. Sorry.

 

[Gray_Jack]

Assuming you didn't compile on windows, I'm afraid I have no clue as to the problem. Sorry.

No problem, I'll use skeith instead

 

[gnmmarechal]

Can someone test a thing for me? Install Super ftpd II Turbo .cia on your 3DS and see if corbenik shows the app icon
For some reason it does not appears on mine, but appears with Cakes and Luma, see if you can reproduce this bug(?)

I do have it installed and I do use Corbenik.

 

[The Catboy]

Been a while, any progress on the dim background fix? Anything I can break?

 

[gnmmarechal]

Been a while, any progress on the dim background fix? Anything I can break?

Chaoskagami is writing stuff on a different branch atm, I think. He's not going to release any new nightlies or so before merging with master, I guess.

 

[chaoskagami]

Been a while, any progress on the dim background fix? Anything I can break?

Chaoskagami is writing stuff on a different branch atm, I think. He's not going to release any new nightlies or so before merging with master, I guess.

Porting it to autotools, and not only does it not quite work, building is currently half-broken. Until I fix the build process, I can't merge that branch back to master in good faith (and therefore, no skeith builds until, you know, builds work.) Aside from that, the skeith build scripts will need to be fixed up (read: easy.) And, despite the progress on github in that branch - I have quite a few commits locally that haven't been pushed, mainly because some configure-time flags aren't behaving (as in, --disable-chainloader)

Since nothing has has really happened in said branch aside from swapping out the build system and changing paths - you're really not missing much (aside from frustration), @Crystal the Glaceon. Nothing new to test for now.

I've also been actually playing games lately since I've needed a break. I'm trying to pace myself at this point since I need to relax and watch anime, play games, lurk /a/ or whatever.

I'm also attempting to do something else 3DS-related which is proving a royal pain without any proper way to debug it, and has a decidedly un-fun testing cycle which neither the unbootable XDS, hardware-inaccurate citra, Luma-dev, nor Corbenik can help with (at the moment, at least.) I think at this point my only testing route might be DHS-Cakes (ew)

 

[The Catboy]

Porting it to autotools, and not only does it not quite work, building is currently half-broken. Until I fix the build process, I can't merge that branch back to master in good faith (and therefore, no skeith builds until, you know, builds work.) Aside from that, the skeith build scripts will need to be fixed up (read: easy.) And, despite the progress on github in that branch - I have quite a few commits locally that haven't been pushed, mainly because some configure-time flags aren't behaving (as in, --disable-chainloader)

Since nothing has has really happened in said branch aside from swapping out the build system and changing paths - you're really not missing much (aside from frustration), @Crystal the Glaceon. Nothing new to test for now.

I've also been actually playing games lately since I've needed a break. I'm trying to pace myself at this point since I need to relax and watch anime, play games, lurk /a/ or whatever.

I'm also attempting to do something else 3DS-related which is proving a royal pain without any proper way to debug it, and has a decidedly un-fun testing cycle which neither the unbootable XDS, hardware-inaccurate citra, Luma-dev, nor Corbenik can help with (at the moment, at least.) I think at this point my only testing route might be DHS-Cakes (ew)

Fair enough, take your time and I trust you.
Just hated seeing this thread so dead, gets me a little worried.

 

[Wolfvak]

Fair enough, take your time and I trust you.
Just hated seeing this thread so dead, gets me a little worried.

The fact that a thread is "dead" is kind of a good thing in my eyes: means users don't have issues and the software is considered stable.
Of course, this only counts when the software is actually finished.

 

[The Catboy]

The fact that a thread is "dead" is kind of a good thing in my eyes: means users don't have issues and the software is considered stable.
Of course, this only counts when the software is actually finished.

Little worried.
But yeah, you're right. I've been using the latest skeith from 5 days ago and just ended up enjoying Monster Hunter Generations because I couldn't break anything.

 

[chaoskagami]

Fair enough, take your time and I trust you.
Just hated seeing this thread so dead, gets me a little worried.

Not dead. Just slower. I've seen the Luma topic drop to like page five and get revived before, so it's not all that abnormal.

The fact that a thread is "dead" is kind of a good thing in my eyes: means users don't have issues and the software is considered stable.
Of course, this only counts when the software is actually finished.

I suppose. Not like corbenik is "finished" but it's certainly in a very stable state - much of what needs to be done is going to take a long time. The only reason why there was so many releases initially was that some stuff was outright broken.

Either way, I might fork off master, cherry pick the important stuff and do a release. It's been a month. Most people use skeith at this point from what I can tell - that's not a good thing.

 

[Sonic Angel Knight]

Little worried.
But yeah, you're right. I've been using the latest skeith from 5 days ago and just ended up enjoying Monster Hunter Generations because I couldn't break anything.

Stop breaking things, you should know better

 

[Joel16]

Kinda off topic but do you plan on implementing something similar to PSP CFW's plugins? I know NTR cfw had something sort of like that. Closed source can be a pain, you're probably not even gonna think of reverse engineering that haha. Maybe if you can gather a team like uOFW (we're trying to reverse engineer the psp's kernel modules to gain more flexibility. Progress is slow as usual, due to the lack or support and motivation ) it would be possible. Just a suggestion.


(Just so you know, I'm well aware that this program uses patches similar to cakes)

 

[Gray_Jack]

Kinda off topic but do you plan on implementing something similar to PSP CFW's plugins? I know NTR cfw had something sort of like that. Closed source can be a pain, you're probably not even gonna think of reverse engineering that haha. Maybe if you can gather a team like uOFW (we're trying to reverse engineer the psp's kernel modules to gain more flexibility. Progress is slow as usual, due to the lack or support and motivation ) it would be possible. Just a suggestion.


(Just so you know, I'm well aware that this program uses patches similar to cakes)

If I'm not wrong, 3DS CFW is kernel9 only, something similar to NTR would need kernel11, so it would need an app to do so.
Kernel9 CFWs can replace or add a new button someware to do a kernel11 exploit and start this new CFW (but I think this would need something similar to the loader replacement, but a replacement of other things then the loader)

Please, correct me I'm wrong about it

 

[chaoskagami]

Kinda off topic but do you plan on implementing something similar to PSP CFW's plugins? I know NTR cfw had something sort of like that. Closed source can be a pain, you're probably not even gonna think of reverse engineering that haha. Maybe if you can gather a team like uOFW (we're trying to reverse engineer the psp's kernel modules to gain more flexibility. Progress is slow as usual, due to the lack or support and motivation ) it would be possible. Just a suggestion.

(Just so you know, I'm well aware that this program uses patches similar to cakes)

A bit off topic, sure. I'll reply anyways, because it's an interesting question on both fronts.

Whenever I get around to implementing a thread in old pre-loader reinand style at very least I'll support the same bytecode patches. No timetable, because I'm lazy and busy getting RL stuff in order - I'm moving pretty soon.

I agree, closed source is a royal pain. There's quite a few people who would like to see something like uOFW (and in fact, I've been watching uOFW for years now, and run the uOFW modules that are 100% replicated.) However, I don't think I'm the right person to gather a team for it. The 3DS' OS is also gigantic compared to the contents of flash0. Quite a few people have been working off and on - some people are working on parts of the OS aside from loader at least. It's also likely much more reasonable to aim for a replacement FIRM than a whole OS.

If I'm not wrong, 3DS CFW is kernel9 only, something similar to NTR would need kernel11, so it would need an app to do so.
Kernel9 CFWs can replace or add a new button someware to do a kernel11 exploit and start this new CFW (but I think this would need something similar to the loader replacement, but a replacement of other things then the loader)

Please, correct me I'm wrong about it

You're close to 90% wrong. NTR can almost certainly be injected from loader automatically. I just have no interest in this. The problem isn't one of kernel vs user, since it's trivial to call svcBackdoor from loader (which gets one ARM11 kernel.)

The problem is finding a mostly-clean place to stick the code in memory that stays running all the time. HOME menu will cut it for NTR, but the limitations of that are many - you can't patch system modules until HOME is running, for example. As for 'add buttons' that's nonsense. Process9 has nothing to do with anything in userland (in official software, at least.)

 

[Gray_Jack]

You're close to 90% wrong. NTR can almost certainly be injected from loader automatically. I just have no interest in this. The problem isn't one of kernel vs user, since it's trivial to call svcBackdoor from loader (which gets one ARM11 kernel.)

The problem is finding a mostly-clean place to stick the code in memory that stays running all the time. HOME menu will cut it for NTR, but the limitations of that are many - you can't patch system modules until HOME is running, for example. As for 'add buttons' that's nonsense. Process9 has nothing to do with anything in userland (in official software, at least.)

Thanks for correcting me :3
I'll read more to complete understand and learn everything that I was wrong

 

[chaoskagami]

Anyways, autotools branch got fixed and is merged to master. I'm going to wait on @gnmmarechal to restart my buildbot so his updater is in place, but I'm also probably going to stable release at some point today since people seem to think I'm dead. 78 commits since last release is not dead. Again, the only reason things were so quick initially is because it was broken.

It'll be up before end-of-day, but please do expect month-long stable releases from now on. Also, I'm changing my policy on nightlies; please do report bugs on them, since they're fairly stable built off master. Don't report bugs against branches - those are probably known broken or WIP.

 

[gnmmarechal]

Ok, Skeith Updater *should* work, hopefully.

 

[chaoskagami]

Well, Skeith was built and pushed. Time to test how the updater handles it.