There are problems with the requirement to format volumes in anaconda (either autopart or Manual Partitioning.) 1. To properly install to hardware RAID 0, 10, 5, 6, I effectively have to learn kickstart. I can properly create the volumes myself, informing the file system at make time what the chunk and stripe width are. The anaconda requirement to reformat in this case is unreasonable. 2. It makes Fedora 18 unfriendly with existing, and its own Btrfs volumes. Existing btrfs volumes, including Fedora 18 installs, legitimately place home in a subvolume. There are other legitimate uses for subvolumes including installing entirely different distributions, user and OS snapshots, etc. The reformat policy demands either that I obliterate all of these, or learn kickstart, and I find this short term hostile and long term untenable for Btrfs. 3. The policy complicates the UI/UX of the Installation Destination spoke. Both basic and advanced users can benefit from a simplified point and shoot UI: select a pre-formatted volume, click an install button. Two clicks and installation commences. This is how 95% of the world's desktop OS's work by default, yet it cannot be done in anaconda at all. Unless I learn kickstart. I am new to oldUI and newUI, I prefer newUI by a lot except for Installation Destination which I think is unnecessarily complicated and hindered by the lack of a point and shoot UI. I should be able to pick volumes by default, not disks, when there are only one or two disks available on the whole system. It's legitimate to disqualify a volume as an install target if it has an invalid file system that doesn't pass fsck, and some other reasonable parameters. But the present behavior I find weird and unreasonable because in effect anaconda is saying "no you can't use other tools and have me install there, only through me do you get any kind of installation." Chris Murphy _______________________________________________ Anaconda-devel-list mailing list Anaconda-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/anaconda-devel-list
