afaict the seabios distfile is only required on arm or !system-seabios; similar for ovmf.
seabios distfile is only required on arm. Hmm, well I' curious to see what dlan has to say on this. He has been working arm hardware lately and is familiar with the layout. Aside from that I spent around 4 months solid last year getting arm keyworded and this rather suggests I missed something completely, unless this has changed in recent versions.
Note: I didn't actually test bypassing the unpack; this is just what I think based on the ebuild.
Huh, looks like the dep changed from !arm to amd64. I will test disabling the unpack later today.
ovmf, seabios SRC_URI should be USE-conditional is to dictate (to us) what it SHOULD be. The use of the word should carries so much baggage. dlan and I have worked these ebuilds extensively long ago. The choice of whether to use a use flag is an arbitrary one we made then. There are comments by long standing devs in the ML threads and bugs that query the merit of making patches or fetching of tarballs conditional on a use flag. There was never a clear cut case for either state to be 'right' or 'best'. You should not tell us what the final form should be. From dlan; not simply adjust SRC_URI, also will touch ebuild (patch, configure) which says the form of the ebuild is fine as is. He also plans to look at (patch, configure) but let's let him expand on that in his own time. I added arm support to xen just under a year ago with full support and input from dlan. We did not do so without routine intensive testing and editing. Test as you will
