Btrfs and Squashfs merged into Linux kernel

I'm glad brtfs was merged, that's going to be a really nice filesystem, and I think it may just render the zfs question a moot point.

Not so fast, herzeleid. Btrfs is far from being production ready.

It isn't even ready for the average home user. But, yes, it is a really nice file system. Looks very promising.

> Not so fast, herzeleid. Btrfs is far from being production ready.

Eh? who said it was production ready? I'm interested because I'll eventually need to move off of reiserfs. And yes, I'm ready to start testing it (as well as ext4) as soon as it's available.

Ext4 is *much* more mature.

For Btrfs you should wait until http://bugzilla.kernel.org/show_bug.cgi?id=12435

is /really/ fixed. A file system should have a proper ENOSPC handling (as the bare minimum).

multics,

Please tell me how it gets fixed if you talk everyone out of testing it?

Are you just a consumer that likes the 0 (dollar, pound, euro, yen, ...) price tag? That is, you want but you do not give anything back. It, btrfs, was merged to encourage testing so that it could replace Ext4.

[Sorry all too serious]

> For Btrfs you should wait until

Nah, I'm going to devote a test partition to it, and dive in head first.

> Ext4 is *much* more mature

No argument there.

TxtEdMacs, I just wanted to lower herzeleids expectations. It seemd to me that it is better to warn before people are frustrated.

It is difficult to fix. I would love to do it myself, but I can't. It is a complex project (Chris Mason himself said that) + it is kernel code. I simply don't have the time.

If I would be a consumer who "likes the zero euro price tag" I wouldn't be posting here, nor would I even know that Btrfs is a file system for Linux.

*Everyone* trips over the ENOSPC problem. I thought it is worth mentioning. I would have waited with putting Btrfs into mainline before that one is fixed.

Since ENOSPC was already supposed to be done according to the btrfs timeline, I would not be surprised it it actually is done by the tine 2.6.29 is released.

Read the logic for its inclusion in the main branch (I suggest checking lwn.net on why this move makes sense). The reality is: it's not tested seriously until it's merged.

multics, I beg to differ, freeloaders (credit © to TC) are everywhere and LXer is NOT immune.