Re: The argument for fs assistance in handling archives

[Tonnerre]

Salut,

On Thu, Sep 02, 2004 at 08:35:58PM +0000, Erik Hensema wrote:
> In this case it's hard to imagine what advantages a multistream file
> will bring, simply  because we aren't used to  working with them. Or
> maybe there  aren't any advantages at  all. Point is:  we don't know
> yet. Maybe  we have to  toy with the  idea for some years  before we
> really learn why multistream files are so great.

Thus I recommend doing the same thing everything does to get his stuff
implemented.

How did User Mode Linux get accepted into the kernel?

In the beginning,  people couldn't imagine any advantage  of it. Then,
someone  started coding  it and  provided patches.   People  tried it,
improved it (important!  It got  to work on many platforms while being
out of the main tree!) and some day we noticed that it was really cool
for kernel hacking and virtualization. 

So when it approached a mature level where we finally saw it was cool,
it went into the mainline kernel, where it is now.

For multistream files, what is that development cycle going to be? Get
the idea, flame, flame, flame... ?

Code talks, bullshit walks.

			    Tonnerre

Attachment: signature.asc
Description: Digital signature