persistent property store - first try

[Kay Sievers]

On Sat, 2004-06-12 at 20:58 +0200, David Zeuthen wrote:
> On Fri, 2004-06-11 at 17:21, Kay Sievers wrote:
> > > > On Fri, 2004-06-11 at 00:06 +0200, Kay Sievers wrote:
> > 
> > Here is a small update with a few more checks for NULL parameters
> > and a simpler device.c integration.
> > 
> > Hmm, should we convert:
> >   hal_device_property_set_persistence(d, "info.test", TRUE)
> > 
> > to a generic method, so we may use it for other possible attributes too?
> > Something like this:
> >   hal_device_property_set_attribute(d, "info.test", HAL_PROP_ATTR_CALLOUT, FALSE)
> >   hal_device_property_set_attribute(d, "info.test", HAL_PROP_ATTR_PERSISTENCE, TRUE)
> > 
> > Or if we have bits only:
> >   hal_device_property_set_attribute(d, "info.test", HAL_PROP_ATTR_CALLOUT |)
> >                                                     HAL_PROP_ATTR_PERSISTENCE)
> > 
> 
> I'd say go for the former, but that's more a personal preference. We
> probably want to look at what attributes we should have other than
> these. PER_USER and READONLY comes to mind. 

Yes, I'm not a big fan of bitfields too :)

> By default we should probably go with exactly CALLOUT, PERSISTENCE,
> READONLY. And we should probably hardcode things such that any property
> whose keyname starts with 'user.' is always PER_USER, !READONLY and
> !CALLOUT. Per-user probably requires a bit rework on hald_dbus.c.

Sounds good.

> One comment on this patch
> 
> > +#define HAL_PSTORE_SYSTEM_PATH      "/var/lib/hal"
> > +
> 
> This should probably be PACKAGE_LOCALSTATE_DIR so it works with
> installing in /usr/local.

Yeah, I just didn't know where to put it.


If the checkin is still impossible and you have a recent patch with your
changes, let me know :)

Thanks,
Kay


_______________________________________________
hal mailing list
hal at freedesktop.org
http://freedesktop.org/mailman/listinfo/hal