On Sun, May 02, 2004 at 02:56:33AM +0200, Thiemo Seufer wrote: > I guess you used partman (as parted frontend). Currently, this is the > default on all m68k, but rather due to a bug in partman which I fixed > today. According to the source, partitioner/partconf is supposed to > run instead, with a comment nearby "[partman is] slow and doesn't > support the atari subarch". > > So I have now some questions: > - Should partman generally be used for m68k? This reqires some work on > on partman for atari, or atari has to be dropped. I think that it should. I plan to work on the atari port to parted at some point if no one else does. At least a couple of people said they might work on it also, so I think it'll get done sooner or later. > - Should partman be used for all subarchitectures except atari? This > needs a subarch check in partman for m68k like it is done in the > bootloader stuff. Yes. Could isinstallable handle this? > - Should partitioner/partconf generally be used? It's faster, but has > some long standing bugs, a less flexible layout, and most developers > won't care much about it. I'd rather not. I think long term all the archs should be going the same route. Using something different is a long term maintenance problem. m68k has enough of those. > And WRT beta4: > - Should the fixed version of partman be uploaded for a beta4 update? > It changes behaviour on m68k, mips and s390; only m86k is in beta4 > so far. (For mips, this update is needed for SGI aubarchitectures, > that is, all currently supported ones.) No I'd like to stick with partman on m68k. I'd rather try to handle any complaints about partman within partman if possible. We'll whip atari into shape at some point, until then hopefully partitioner/partconf can be made to work in the beta 5 timeframe. Thanks, Stephen -- Stephen R. Marenka If life's not fun, you're not doing it right! <stephen@marenka.net>
Attachment:
signature.asc
Description: Digital signature