Importance of LVM (was Re: Partitioning criteria revision proposal)

["Jóhann B. Guðmundsson"]

On 10/25/2012 11:24 PM, Robyn Bergeron wrote:
>>
> I am under the impression that we've been testing with/without LVM 
> anyway, both scenarios?

The installer has been defaulting to EXT4 up to this point there is no 
option to hash or unhash lvm as you could do in the oldUI in the newUI 
and custom partitioning has been more or less broken this whole time.

The oldUI rendered ext4 vs lvm argument moot because it was equally 
easy/hard to disable/enable it for both parties.

> In any case, it doesn't seem as earthshaking as other developments - 
> it's just making the default be what it's been for some time, and 
> given that there exists documentation for the "lvm enabled case" and 
> not much otherwise it seems like a reasonable thing to do.  I would 
> almost make the case that disabling LVM by default - were it a feature 
> - would require a lot of that backup documentation and info that isn't 
> really there....

I think you should focus on getting the feature process to actually 
define what it considers as an actual feature before you propose 
removing lvm as "feature" .

 From that same argumental stand point removing the functionality of 
disabling lvm as easily as it was in F17 should have been mentioned on 
the newUI feature page.

JBG