On 1/28/07, Ulf Lamping <ulf.lamping@xxxxxx> wrote:
> [snip]
Well, first of all, what problem do you try to solve here?
the lack of an uniform way to deal with tables, they are dealt
differently when they could be dealt in a single uniform way.
I have four diferent examples I wrote myself:
four user encap "protocols", the k12 config file, the snmpv3 users
file and lastly the df_macros table
but there's more candidates in there, not written by me:
ssl uses a single splitted line of text for the various decryption
keys, it works but a table will do better.
ipsec has 4 (always 4 and only 4) replicate preference sets for associations.
and if I dive in the code I bet I'll find many more.
What I want is a single uniform way (although not a very
straightforward one I admit) to deal with tables of "settings"
accessible to the user.
--
This information is top security. When you have read it, destroy yourself.
-- Marshall McLuhan
