[whatwg] some feedback on Web Apps 1.0 client-side storage

[Jim Ley]

On 9/30/05, Robert O'Callahan <rocallahan at gmail.com> wrote:
> On 01/10/05, Ian Hickson <ian at hixie.ch> wrote:
> > On Fri, 9 Sep 2005, Robert O'Callahan wrote:
> > > Then the UA might have to terminate any running script(s) (perhaps
> > > after warning the user and giving the user an option to cancel the
> > > emptying).
> >
> > Do we really want to require this? I agree that a user agent may wish
> > to do this, but it seems that the interaction of the user deleting
> > data and running scripts accessing that data would be something that
> > could be safely left undefined, since it doesn't directly impact
> > interoperability.

I agree, I see no reason to define it, users (and by extension their
tools) should be free to delete any data on their machine at any time
they want.  It is also the sort of thing that it would be trivial for
the user to do (using any of the many methods of executing script e.g.
userJS greasemonkey etc.)

> I think some sort of guarantee that data can't just disappear while a script
> is running would be very helpful to developers, and of course they'd be more
> able to rely on it if it was in the spec.

Authors don't currently have this safety - most of them just ignore
the problem, and I don't see any reason for it - what's the worst that
can happen - good authors should be coping with just about any
situation, that's the environment the script is executing in.

Jim.