Reinoud Zandijk <reinoud%NetBSD.org@localhost> writes: > We have amongs others, MBR, GPT, disklabels, (sun-)disklabels, LVM, RAID > etc.etc. The current idea is to have wedge discovery creating wedges for each > type. Not my favorite but it usually kind of works. Some of those are to be compatible with other operating systems and ROM-based bootloaders. And BSD disklabel is the traditional way. raidframe isn't really a partitioning scheme. > Now LVM is not a holy grail, but wouldn't it be good if we had a BSD licenced > system compatible with LVM as the only frontend? > > I.e. instead of having formats dictate the wedges, have LVM dictate the > (hidden) backend encoding needed for say booting? While still have all the > configurability like growing/shrinking/moving/joining etc? I really don't see how this would help/work. But I can certainly see the notion that some fancier tools could help people get these things right. For me, having better gpt support (boot blocks, easily nested raid) would be the biggest improvement. It's partially there, but it's still in the "too hard" category, it seems.
Attachment:
pgpoBwa0tS8HY.pgp
Description: PGP signature