Re: GConf design goals.

[Colm Smyth <Colm Smyth sun com>]

>From: Dietmar Maurer <dietmar ximian com>
>X-Accept-Language: en
>MIME-Version: 1.0
>To: Havoc Pennington <hp redhat com>
>CC: Colm Smyth <Colm Smyth sun com>, bje apnic net, gconf-list gnome org
>Subject: Re: GConf design goals.
>Content-Transfer-Encoding: 7bit
>
>Havoc Pennington wrote:
>
>> Dietmar Maurer <dietmar ximian com> writes:
>> >
>> > I consider the possibility to store CORBA_any as more flexible, since I
>> > do not make any restrictions to the data which can be stored.
>> >
>> > Your argument is to protect users from doing strange things. I simply think
>> > you can't protect someone from making errors. Maybe you can make I a
>> > bit more unlikely. On the other side there are many usage scenarios where
>> > structures are perfectly well, and I want to allow the use of structures
>> > where
>> > it is useful.
>> >
>>
>> If you just want to use structures, you can just use a wrapper such as
>> bonobo-conf, or just a utility function such as
>> corba_any_to_string(). I don't see what difference that makes for
>> GConf in any deep sense.
>>
>
>The difference is that only bonobo-conf programs can access those
>values if I decode everything into a string. Else every program can
>access/modify those values.

Wrong. Only programs that know about the data-structure you are
encoding and the encoding method can access those values.

Colm.

>- Dietmar
>