E-Maculation

Thanks for posting the compiled binary!

After downloading and loading the revised XCode project on Catalina/XCode 11.3.1, I still get the same 3 warnings and 2 errors (although if I re-run the release build, I then get 176 warnings). The error relates to the CPU part again. I navigated to the MacOSX folder and ran "make -f Makefile.gencpu" from the shell (which is the shell command in \Run Script\Build Phase) and it says the following:

make: *** No rule to make target `/../uae_cpu/build68k.c', needed by `/gencpu_output/build68k'. Stop.

"/Users/racini/Library/Developer/Xcode/DerivedData/BasiliskII-eedamcnnvfgqgnecudaqxokswvxz/Build/Products/Release/gencpu_output/" as a folder exists but is empty.

If I add "all" to the command, it doesn't matter. I'm not sure why it continues to barf on building the CPU core unless others who have made successful builds have something else installed or configured that I don't.

Rich

I do not know if there is a software problem on your Mac or if it is something you do.

It should be very simple. This how I do it:
1. Make sure SDL2.framework version 2.0.10 is installed in /Library/Frameworks/
2. Empty the folder ~/Library/Developer/Xcode/DerivedData
3. Download the zipped source from https://github.com/kanjitalk755/macemu and expand it
4. Open the file macemu-master/BasiliskII/src/MacOSX/BasiliskII.xcodeproj
5. In the icon bar in Xcode window choose BasiliskII > Edit Scheme... and make sure the Build Configuration is set to "Release" for all five.
6. Choose Build from the Product menu.

Ronald P. Regensburg wrote:I do not know if there is a software problem on your Mac or if it is something you do.

It should be very simple. This how I do it:
1. Make sure SDL2.framework version 2.0.10 is installed in /Library/Frameworks/
2. Empty the folder ~/Library/Developer/Xcode/DerivedData
3. Download the zipped source from https://github.com/kanjitalk755/macemu and expand it
4. Open the file macemu-master/BasiliskII/src/MacOSX/BasiliskII.xcodeproj
5. In the icon bar in Xcode window choose BasiliskII > Edit Scheme... and make sure the Build Configuration is set to "Release" for all five.
6. Choose Build from the Product menu.

Ok, starting over:
1. Confirmed
2. Done
3/4. Re-done
5. I only had "Release" on the Run one, not Test and Analyze. Profile and Archive were already at Release
6. Done

184 warnings and the same 2 errors. The two errors still relate to the CPU.

I decided to re-try the above on my Parallels VM that has Sierra and XCode 8.3.3. It produces 1 warning (on the depreciated function in etherhelpertool.c) and the same two errors on the CPU. I also tried running the shell script manually and it also barfs, same as before.

Looking at the output, I thought that maybe I had the files buried too deeply in the file system (I had it in Documents/XCode Projects) which is something I've seen with Visual Studio on Windows. So, I moved the macemu-master to my home folder. In Sierra and XCode 8, it now get 500+ warnings but no errors, and the build completes. I then retried it with Catalina and XCode 11 and I get 4497 warnings (!!) and no errors and the build completes. I moved the build to my Basilisk folder and ran it, and it runs. So that's good.

It would be interesting to see where in the file system other builders have their files and if that might be part of it.

Rich

RichCini wrote:I had it in Documents/XCode Projects

You are supposed to have a Documents folder in your Home folder. Did you create another Documents folder at root level on your System disk? Or where?

It would be interesting to see where in the file system other builders have their files and if that might be part of it.

I simply have the source on my desktop. After builds were successful, I trash it. There is no point in keeping it unless one wants to study and maybe edit the code.

Ronald P. Regensburg wrote:

RichCini wrote:I had it in Documents/XCode Projects

You are supposed to have a Documents folder in your Home folder. Did you create another Documents folder at root level on your System disk? Or where?

No. The files were originally in /Users/racini/Documents/XCode Projects/. Now they're at /Users/racini.

Ronald P. Regensburg wrote:5. In the icon bar in Xcode window choose BasiliskII > Edit Scheme... and make sure the Build Configuration is set to "Release" for all five.
6. Choose Build from the Product menu.

The most recent build had just only one scheme (Kanijatalk just confirmed that.
The version with 5 schemes, I am also able to build. But more recent versions with just one scheme fail with same error messages as with RichCini‘s builds.

I feel the merger to just one scheme might have triggered something, effecting not all Xcode IDEs.
My other Mac builds OK, but now I am on the road and rely on your compilations.

So all good. - I am so glad, that AppleTalk is supported that I couldn‘t stop smiling all Sunday! 8D

KnuddlMac wrote:The most recent build had just only one scheme (Kanijatalk just confirmed that.

I didn't say that there are five schemes. There is only one scheme "BasiliskII" and when you choose "Edit Scheme..." you will see five items that can be set to either Debug or Release.

Ronald P. Regensburg wrote:

KnuddlMac wrote:The most recent build had just only one scheme (Kanijatalk just confirmed that.

I didn't say that there are five schemes. There is only one scheme "BasiliskII" and when you choose "Edit Scheme..." you will see five items that can be set to either Debug or Release.

All good, we‘re on the same page.
On one machine all compiles OK. On the other, reports from RichCini match 100%.
I‘m on the road again for this week and can‘t drive detailed comparisons. Plus on the bad side, I don‘t have the system in my luggage, that compiles OK. - But I have a atalkd Linux machine with me to also get my MacBook familiar with the new bytes being offered.

Still interesting why the compilation fails. But I did work with different MacOS SDK versions compiling KEXTs over the past months. Just interesting, that RichChini observed an identical experience.

Looking into similar error reports, the ocean is wide. - I think it‘s not worth drilling to the details.
Right now we have the necessary binary at hand and that counts.

Regarding Thunderbolt and bridge0 -- has anyone attempted to toss tap0 into bridge0 along with en0? I figure that Thunderbolt is addressable even though most people don't connect over it, but that should at least allow local instances to talk to each other over the tap device.

Of course, if you actually use ethernet-over-tb, this means it'll now be bridged to these other virtual devices -- but that shouldn't be a problem for most people.

adespoton wrote:Regarding Thunderbolt and bridge0 -- has anyone attempted to toss tap0 into bridge0 along with en0? I figure that Thunderbolt is addressable even though most people don't connect over it, but that should at least allow local instances to talk to each other over the tap device.

Yes, I tried it and it works.

adespoton wrote: Of course, if you actually use ethernet-over-tb, this means it'll now be bridged to these other virtual devices -- but that shouldn't be a problem for most people.

Might be a bit forward and back through the tunnel and finally onto the IP stack, but I don't complain. The bits are fast enough and I enjoy the added connectivity. J

Elyus provided an interesting article (page 4 of this post) about OpenVPN and how traffic could be tunneled for using WiFi.
https://www.emaculation.com/doku.php/br ... rver_setup
That would be my next thing to be solved, as I am mostly running around with a wireless MacBook. But the instructions are exhaustive and maybe a project for silent Christmas days. - So a long time to go...

I just continued my work setting-up a Catalina MBP with TAP support.

As the latest Tunnelblick driver failed loading, I desperately tried the good old TUN/TAP from http://tuntaposx.sourceforge.net and to my surprise it works again on Catalina. Still no WiFi (without tunneling), but the Thunderbolt Ethernet adapter works.

Thanks to Cat_7's comments on page 4 of this thread, I managed to load kernel extensions also after reboot. Just the syntax did slightly change as the driver now has a different name:  net.sf.tuntaposx.tap

Did someone already make experience with a beta of Big Sur and associated support of TAP or whatever com layer will be offered?
There is an interesting outlook on https://tunnelblick.net/cBigSur.html

we should just be able to lift the notarized kext from tunnelblick and distribute it in a BII/SS-friendly package, shouldn't we? We can make an installer that drops the launchdaemon config if TunnelBlick doesn't exist already

putrzop wrote: ↑Fri Oct 21, 2022 10:33 am we should just be able to lift the notarized kext from tunnelblick and distribute it in a BII/SS-friendly package, shouldn't we? We can make an installer that drops the launchdaemon config if TunnelBlick doesn't exist already

I've had that same thought myself, but not the time to do it. Tunnelblick is GPLv2, so including the binary and the license with a link to the project should be enough?

I've had that same thought myself, but not the time to do it.

That‘s also my problem. Last time I did invest my time into BA II connectivity, feedback was low.
I was questioning, if anyone still makes use of the AppleTalk Ethernet connection.

I stopped all my Motorola 68k Hardware renovation projects and could survive past years with just one brave NetaTalk Server on Samsung silicon.

So is still someone making use of that Ethernet connectivity part out there?

Best,
-Pit

i'm not sure what to do now tech guides get-mobdrovip.com

Right now I’m servicing my LaserWriter 4/600 PS. But without Mac OS speaking AppleTalk I couldn’t have ran Apple Printer Utility to find out what the printer’s error message is. Also, to my knowledge, it’s the only tool that allows me to amend my printer’s contrast after installing a new toner.

KnuddlMac wrote: ↑Sat Oct 22, 2022 8:06 am I was questioning, if anyone still makes use of the AppleTalk Ethernet connection.

[…]

So is still someone making use of that Ethernet connectivity part out there?

So yes, I do.

Hey mabam, you made my day!
It is good to know, that sometimes invested 'System 7 time' is being spent in a useful way. -

Right now, I can't simply grab my Xcode and compile something.
But let's continue the discussion here and see, if there are possibilities with input from putrzop, adespoton and others.

From what I learned at Tunnelblick, Apple announced to discontinue Kext support without further notice.
But that status was maybe discussed ~2 years ago?

KnuddlMac wrote: ↑Mon Oct 24, 2022 11:54 am Hey mabam, you made my day!
It is good to know, that sometimes invested 'System 7 time' is being spent in a useful way. -

Right now, I can't simply grab my Xcode and compile something.
But let's continue the discussion here and see, if there are possibilities with input from putrzop, adespoton and others.

From what I learned at Tunnelblick, Apple announced to discontinue Kext support without further notice.
But that status was maybe discussed ~2 years ago?

Kexts are deprecated, but still used by Apple themselves, as well as by Tunnelblick and MacFUSE.

Personally, I'd like to see Apple integrate FUSE directly into their filesystem framework, since both frameworks are essentially creating a userland space for storage drivers.

Likewise, Apple could integrate Tun directly into the network stack, with a userland interface for interacting with tunnels, like they've done with VPN interfaces on iOS.

In both cases, most of the work to transition to userland drivers is already done; Apple just has a few more steps to take.

And honestly, this is what's kept me from putting any spare time into the kext installers over the past few years -- I assumed they'd be totally broken by now and new userland interfaces available.

Last year I tried Cat_7’s test build with etherhelper. But he reported that, opposed to a connection via tap, it gets the same IP address as the host.
So I guess for a guest system using AppleTalk via DDP only, it should be fine.

Back then I made a test build of Basilisk II with vde (viewtopic.php?t=11466). The build launched but I didn’t succeed in establishing a network connection. Possibly it was just me failing due to wrong network settings. I don’t have much knowledge of that stuff. But it’s also possible that vde in Basilisk II needs fixes first.

Possible options I see are:
1) vde,
2) kext installers for tuntap,
3) avoid using kexts by following ZeroTier’s approach (discussed at page 2 of this thread).

But I can’t judge whether any of these are feasible/persistent.


(On a different note:

mabam wrote: ↑Mon Oct 24, 2022 10:27 am Right now I’m servicing my LaserWriter 4/600 PS. But without Mac OS speaking AppleTalk I couldn’t have ran Apple Printer Utility to find out what the printer’s error message is. […]

According to Apple Printer Utility there was something wrong with the fuser of my LaserWriter 4/600 PS. But after replacement it still threw the same error message. So I plan to replace all capacitors on the controller board and hope that will fix it … )

mabam wrote: ↑Mon Oct 31, 2022 12:35 am Back then I made a test build of Basilisk II with vde (viewtopic.php?t=11466). […]

Just in case anyone’d like to test: The download link to the test build provided in that other thread (in this post to be precise) is active again now.

mabam wrote: ↑Mon Oct 31, 2022 12:35 am (On a different note:

mabam wrote: ↑Mon Oct 24, 2022 10:27 am Right now I’m servicing my LaserWriter 4/600 PS. But without Mac OS speaking AppleTalk I couldn’t have ran Apple Printer Utility to find out what the printer’s error message is. […]

According to Apple Printer Utility there was something wrong with the fuser of my LaserWriter 4/600 PS. But after replacement it still threw the same error message. So I plan to replace all capacitors on the controller board and hope that will fix it … )

Well, the new capacitors cured my good old printer!

AppleTalk lives …

mabam wrote: ↑Mon Nov 28, 2022 11:21 pm

mabam wrote: ↑Mon Oct 31, 2022 12:35 am (On a different note:

mabam wrote: ↑Mon Oct 24, 2022 10:27 am Right now I’m servicing my LaserWriter 4/600 PS. But without Mac OS speaking AppleTalk I couldn’t have ran Apple Printer Utility to find out what the printer’s error message is. […]

According to Apple Printer Utility there was something wrong with the fuser of my LaserWriter 4/600 PS. But after replacement it still threw the same error message. So I plan to replace all capacitors on the controller board and hope that will fix it … )

Well, the new capacitors cured my good old printer!

AppleTalk lives …

Hooray! And now you don't have to worry about the board being etched by electrolytes, either!

That was before (still the old capacitors on this picture).

mabam wrote: ↑Tue Nov 29, 2022 12:44 am



That was before (still the old capacitors on this picture).

Wow... good thing you got to it before that 400V one blew!

Yeah, now the fan is blowing again instead …