Hi Joseph, On Thu, Dec 02, 2010 at 10:57:59AM -0500, Joseph Jezak wrote: > Sorry for not playing a larger role in getting this taken care of, I've > been really busy with other things. Sure, no problem. > The code that I posted was not finished, I know. When I posted it to the It is an excellent starting point though. > yaboot mailing list, I was looking for testing and some feedback as to > the approach I took. I have not seen a report of the code breaking where > the old ofpath worked or a good reason to keep the old ofpath code > around besides the fear of breaking things. However, I do also > understand the desire to not rock the boat, especially from Debian's > perspective. > > In any case, if I were to integrate my code with the old ofpath, are > there enough people to test it with any confidence over the new version > I wrote? Does anyone actually want this? What would be the best solution? I actually think it is a good idea, removing the cruft is always possible later. And using the ideas behind your code in an attempt to add sysfs support to the current ofpath seems to be working great so far on the G5, G4, and G3 PowerMacs I have access to. I am more worried about IBM machines though. I could test on Pegasos and Efika, but is anyone really using Yaboot on those? Because, in the past, Sven made SmartFirmware directly load Linux and D-I followed that spirit. I actually think Milan has a great idea to ask for tarballs of device-trees and sysfs info. It would be nice if those could be uploaded to the BTS for public reference and perhaps regression tests. Cheers, -- .''`. Aurélien GÉRÔME : :' : `. `'` Debian Developer `- Unix Sys & Net Admin
Attachment:
signature.asc
Description: Digital signature