Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-18 Thread Ian Jackson
Marco d'Itri writes (Re: packaging a tiny/trivial blob in a DFSG-clean way?): This is an old issue which is well established, widely agreed and does not really need to be discussed again. The facts are: - if the source and the tools to build them are in Debian, then there is no DFSG issue

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-15 Thread Bernhard R. Link
* Michael Tokarev m...@tls.msk.ru [120914 23:12]: But I'm still a bit, well, uncomfortable to ship the blobs, even if the source is available and it is verified on buildds during package build on corresponding architectures. Allowing such packaging may act as a bad example in the future.

packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Michael Tokarev
In qemu package there's a source file, pc-bios/spapr-rtas/spapr-rtas.S, which is an PPC assembly file with exactly 5 instructions: #define KVMPPC_HCALL_BASE 0xf000 #define KVMPPC_H_RTAS (KVMPPC_HCALL_BASE + 0x0) .globl _start _start: mr 4,3 lis

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Bastian Blank
On Sat, Sep 15, 2012 at 12:15:51AM +0400, Michael Tokarev wrote: This file is also included in the upstream source in compiled form, as pc-bios/spapr-rtas.bin. This is because it needs ppc assembler to compile, and not every system out there has this tool ready available. binutils-multiarch?

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Michael Tokarev
On 15.09.2012 00:31, Bastian Blank wrote: On Sat, Sep 15, 2012 at 12:15:51AM +0400, Michael Tokarev wrote: This file is also included in the upstream source in compiled form, as pc-bios/spapr-rtas.bin. This is because it needs ppc assembler to compile, and not every system out there has this

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Kurt Roeckx
On Sat, Sep 15, 2012 at 12:15:51AM +0400, Michael Tokarev wrote: So we have the following options: 1) package just this single file, of 20 bytes long, in a separate Arch:all package, in it's own separate source. 2) drop ppc support where this file is required. 3) Just ship the

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Michael Tokarev
On 15.09.2012 01:03, Kurt Roeckx wrote: On Sat, Sep 15, 2012 at 12:15:51AM +0400, Michael Tokarev wrote: So we have the following options: 1) package just this single file, of 20 bytes long, in a separate Arch:all package, in it's own separate source. 2) drop ppc support where this

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Kurt Roeckx
On Sat, Sep 15, 2012 at 01:11:47AM +0400, Michael Tokarev wrote: On 15.09.2012 01:03, Kurt Roeckx wrote: On Sat, Sep 15, 2012 at 12:15:51AM +0400, Michael Tokarev wrote: So we have the following options: 1) package just this single file, of 20 bytes long, in a separate Arch:all

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Marco d'Itri
On Sep 14, Michael Tokarev m...@tls.msk.ru wrote: Well, in that case we can ship alot more .bin files from qemu sources too, and build these on corresponding architecturs like already mentioned (to verify the result is still the same). Additional x86 ROMs, sparc ROMs, this PPC ROM, ... The

Re: packaging a tiny/trivial blob in a DFSG-clean way?

2012-09-14 Thread Ben Hutchings
On Fri, 2012-09-14 at 23:49 +0200, Marco d'Itri wrote: On Sep 14, Michael Tokarev m...@tls.msk.ru wrote: Well, in that case we can ship alot more .bin files from qemu sources too, and build these on corresponding architecturs like already mentioned (to verify the result is still the