Somehow I got the impression with some equipment that the configuration file was nothing more than a direct copy in plain text of command line commands. This seemed logical and useful. Then those commands are automatically run at boot time. Then other commands can be added during run time and made permanent (or not).
(I don't appreciate those configuration files that are somehow encrypted and can't be read at all except by loading them into the device.)
I'm working with a Juniper SRX lately. The configuration files are readable so that's good. They are also editable via the J-Web GUI and that's good. And there's a CLI interface built within the GUI yet serious players seem to favor PuTTY.
It suddenly it occurs to me:
There is a configuration file that is readable that fits some "structured" format BUT while the structured format can be edited like code, its relationship to the CLI commands is rather obscure. Is that a reasonable way of looking at it?
That is, the CLI commands get "compiled or interpreted" into the configuration structure (?).
Did I miss reading some critical high-level description that talks about this?
I note that the CLI interface gives lots of help in structuring commands such as:
set security ?
will tell you what choices come next.
But that's not exactly the same as learning the CLI command set.
And doing that seems to imply learning the configuration file structure as well??
And why the departure from a serious GUI (with respect to the SSG line)? By serious I mean that it's complete and correct within reason and that most serious things can be done within it. Have the 'nix heads taken over?