And I got the idea to check for some CD apps with source on both BeOS
WWW sites:
typedef struct {
uchar toc_data[804]; /* table of contents data (see SCSI standard) */
} scsi_toc;
The comment raised the question what standard was actually meant.
And you mention it too:
> It's pretty primitive - via ioctl. No further documentation (so far) -
> "refer to the SCSI standard".
So I checked my headers and it turns out that CAM seems to be meant.
Very Interesting.
So under FreeBSD I got a "FreeBSD" CD driver that
covers Mitsumi, ATAPI, Sony and SCSI plus a SCSI CAM driver that is not
specific to the OS.
Guess this means that you wrote CAM code and that FreeBSD CAM source code
ist very likely to run under BeOS - if, if one uses the correct
header file and correct library option for each platform and sticks
to SCSI drives.
> The APIs are identical, but the endiannesses are different. The code
> I've got already takes care of that (there are macros/functions in the
> system to convert whatever-endian to host-endian).
As we use such stuff - do you refer to an pre 1.1.0 version?
> I haven't read the BSD licence closely; I undestand it's more lenient
> than the GPL. I didn't know it was applicable to libraries. I'd like
> to go through it first - read the fine print, etc. Up till now I was
> fine with the LGPL/GPL for the library/client.
Slashdot had the BSD/GPL discussion again yesterday.. feel free to dig
through it.. :)
> (Did anyone register when I suggested directory.mozilla.org/
> licence.html as a possible model for the cdindex server-content
> licence?)
We should clarify our license before we carry significant amounts
of data and then waste energy.
Take the reference client again.
It was started under GPL more or less out of habit and now this
license kinda sticks. And this is rather easy code.
I would understand a need for license discussion for complex packages,
but not for this stuff. (sigh)
So I change nothing today and wait for your diifs - ok?
Regards,
Marc