32 Bit FreeBasic OS

[Kevin]

ho thanks, btw, i think it should be easy to port it in freebasic

That's possible, of course, at the cost of making it harder to keep both versions in sync. But then, updates are probably infrequent enough that it could be managable to just port later changes, too.

[panoramic]

the hard drive image (for qemu)
http://www.codinfinity.net/NewOS_hd.zip 4M

This file contains hd.img

If I try to burn a CD-ROM, with Windows or with Imgburn I get : "Invalid or unsupported image file format".
Therefore I cannot boot my computer on a CD with your OS.
Is it possible to boot on a CD to be independent of another OS (Windows) to test your OS ?

Do you intend to release a LiveCD ?

[trolly]

the hard drive image (for qemu)
http://www.codinfinity.net/NewOS_hd.zip 4M

This file contains hd.img

If I try to burn a CD-ROM, with Windows or with Imgburn I get : "Invalid or unsupported image file format".
Therefore I cannot boot my computer on a CD with your OS.
Is it possible to boot on a CD to be independent of another OS (Windows) to test your OS ?

Do you intend to release a LiveCD ?

the hd.img, is and hard drive image, i didn't implemented cdrom driver and cd file system yet, so livecd is not yet suppored, but feel free to contribute if you want

[sernico]

-- removed --

[trolly]

Hi all

I have Lost m'y hard Drive and i do note have any backup

Does someone of you have downloaded a copy of the sources i shared?

[bzt]

btw, i dont have git, how do i dowload the sources tree?

I have Lost m'y hard Drive and i do note have any backup

That's bad. This is exactly why you should have used git (preferably with a remote repo for backup). Nice looking interface btw, reminds me of MWM and CDE from my old Ultrix and VMS days. Well done! Hope you manage to get the source back.

If you want to use some drivers from me (and others), you can just implement the header files of the CDI interface.

I've tried to take a look, but all doc and source links lead to HTTP 404 (and the doc.h only contains comments, no prototypes at all). It does not have a proper page in our wiki, just a stub with more dead links (lowlevel site might have some info, but that's in German only, and the provided source link gives HTTP 404). I was unable to find anything with websearch either (well I have found lots and lots of links, but I doubt they're what you were referring to, like core device interface, container device interface and even Dell's networking interface). It looks like the acronym CDI is already heavily overused. And how is CDI different to EDI for example? EDI seems to have a proper specification and example drivers too. Why not revive that instead? Was CDI designed with microkernel or monolithic kernel in mind?

Cheers,
bzt

[Kevin]

I've tried to take a look, but all doc and source links lead to HTTP 404

It's been a while... Yes, looks like after moving to a different server, the generated documentation wasn't put online any more. You can just clone the git repository and run "make doc", though, if you have doxygen installed.

Essentially what it boils down to is: A driver is basically a struct containing function pointers for callbacks, and the OS calls these callbacks to make requests to drivers. Drivers may call the functions defined in the CDI headers and a subset of the C standard library to implement the functionality. The implementation for these functions must be provided by the OS.

Was CDI designed with microkernel or monolithic kernel in mind?

Both, really. týndur is a microkernel system, but CDI has been used in monolithic kernels, too, pretty much from the start.

And how is CDI different to EDI for example? EDI seems to have a proper specification and example drivers too. Why not revive that instead?

I don't remember the details, but wasn't EDI that very abstract framework, making things a bit more complicated than they had to be? I'm sure I've discussed the differences somewhere on this forum some years back. And did EDI actually ever get more than an example driver?

Anyway, it wasn't on our radar at the time. CDI was started in 2007, so EDI isn't even much older. And today switching to a different interface that has never gained traction sounds like a bad use of the very little time that I still spend with OS development. But of course this shouldn't stop anyone else from doing reviving EDI if they think it's useful.