Hey Dave, I put out a beta release:
http://ftp.gluster.com/pub/freeipmi/qa-release/freeipmi-0.8.6.beta0.tar.gz That should compile the sunbmc driver automatically even if bmc_intf.h is not found (and assuming other build conditions are met). Could you give it a shot and LMK if it works? Al On Mon, 2010-05-03 at 13:12 -0700, Al Chu wrote: > Hey Dave, > > On Mon, 2010-05-03 at 12:52 -0700, Dave Love wrote: > > Al Chu <[email protected]> writes: > > > > > GPL and the CDDL (the Sun license) have incompatibilities with each > > > other (see http://**www.**gnu.org/licenses/license-list.html), which is > > > why > > > I did it this way. However, it appears the incompatibilities may exist > > > only w/ linking and not with code. So presumably I could copy the code > > > in. But I'm not sure. I've been trying to figure this out, but as you > > > can expect, licenses are hard to understand :-) I'll ask around, see if > > > I can tweak this or not. > > > > Sure if it was CDDL code, you couldn't distribute it to combine with > > GPL'ed code unless the copyright holders on all the GPL code allowed an > > exception. However, as I understand it, a header with no code isn't a > > problem. Anyway, what I pointed to is BSD-ish, not CDDL, which is why I > > think there's no problem -- you'd just need to add the copyright notice > > at most. That version is trivially different to the one in OpenSolaris. > > Ahh! That was a subtlety I missed before. There is the bmc_intf.h from > OpenSolaris and the bmc_intf.h from ipmitool. I had always assumed > ipmitool copied it from OpenSolaris and thus both were CDDL. But that > is not the case. Ipmitool's bmc_intf.h appears to be "forked" from an > earlier version, so I could freely use that version. > > Thanks for the catch. I'll look into it a bit more deeply. Assuming no > snags, I can copy that file in. I'll make a beta, and if you have the > chance, could you try out the beta for me? > > Al > > > > Hmmm. I think the reason it fails like that on Linux is b/c /dev/bmc > > > doesn't exist. This is in contrast to Solaris where /dev/bmc does > > > exist, but something doesn't allow the code to work with it (thus > > > "device not found" would not be a good error code). Looking through the > > > error code list, IPMI_ERR_DEVICE_NOT_SUPPORTED is probably the better > > > one. Sound like a good error code to change to? > > > > Right, that sounds right -- definitely better than an internal error. > > If you can just distribute the header, it's moot anyway. > > > > For what it's worth, here's the meat of the licence on bmc_intf.h from > > ipmitool: > > > > * Redistribution and use in source and binary forms, with or without > > * modification, are permitted provided that the following conditions > > * are met: > > * > > * Redistribution of source code must retain the above copyright > > * notice, this list of conditions and the following disclaimer. > > * > > * Redistribution in binary form must reproduce the above copyright > > * notice, this list of conditions and the following disclaimer in the > > * documentation and/or other materials provided with the distribution. > > -- Albert Chu [email protected] Computer Scientist High Performance Systems Division Lawrence Livermore National Laboratory _______________________________________________ Freeipmi-users mailing list [email protected] http://lists.gnu.org/mailman/listinfo/freeipmi-users
