Re: fs-independent quotas

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Fri, Oct 21, 2011 at 03:37:43PM +0200, Michal Suchanek wrote:
> Except that tools not aware of the new value or wider id will now see
> non-unique entries or wrong values due to conversion wrap. I don't
> think even this is workable.

No, if there is an id that doesn't fit in 32bits they will get an error when
doing the convertion.

> 
> >
> >
> > "numerous" is a strong word here. There are convertions in the kernel that
> > would happen anyway (you have to convert from the syscall API to the
> > filesystem API), and I think the userland tools could be doing less 
> > convertions
> > than they're doing now.
> 
> 
> They will be hidden somewhere inside proplib but they still do exist.
> So long as you can assume quota calls are never performance critical
> may be ok I guess. Still it adds code bloat and execution time bloat
> and and data bloat for no obvious advantage. Conversion and versioning
> has to happen one way or another.

When you consider versionning and convertion required for ABI compat of
binary structures, I'm not sure problib is that much bloat.

> > Nor in the tree-based dictionnary, or in the multidimentionnal array.
> 
> No, in an array the unused locations do exist.

I don't understand this. If you have a 2-dimention array
quota[id][type], and quota[class=group] doesn't exist for this filesytem,
you have quota[class=group]=NULL and no memory associated with it.

> 
> > What a table can't easily express is that a line of this table can be
> > of variable lenght, and some of its fields are optionnal.
> 
> Tables can deal with optional fields just fine. Prior art exists. And
> even with optional values they are still way more efficient than XML.

There are text formats more efficient than xml too.

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--