Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 22 Jan 1997 14:29:24 -0700
From:      Warner Losh <imp@village.org>
To:        "Jordan K. Hubbard" <jkh@time.cdrom.com>
Cc:        config@freebsd.org
Subject:   Re: Commerical applications 
Message-ID:  <E0vnAEX-0002Fm-00@rover.village.org>
In-Reply-To: Your message of "Wed, 22 Jan 1997 12:42:31 PST." <25361.853965751@time.cdrom.com> 
References:  <25361.853965751@time.cdrom.com>  

next in thread | previous in thread | raw e-mail | index | archive | help
In message <25361.853965751@time.cdrom.com> "Jordan K. Hubbard" writes:
: You're forgetting serial consoles and machines with monochrome
: adaptors so funky that I doubt any reasonable X would run on them
: (don't laugh, I think freefall has one of these cards in it - it's a
: server machine with no need of a good gfx card).

Those machines are very low in numbers, could be managed remotely, and
should not set the direction for the nice looking project.  While it
would be nice to support them, I've seen enough cool MS installations
to think that they should be second class citizens.  I guess my
comment was more by way of pointing out most systems can run X, so
we should put the bulk of the efforts into that.  This will likely be
how people judge the product, so it should look the best it can.

However, that said, there will be a need for something smaller.  X is
too big for floppies, and many small memory machines.  There needs to
be an alternative for these machines.  That's why I was pushing the
meta-data idea...

: write a single CLI based tool which implements an argument set from
: hell for frobbing DNS.  Then everything else fronts for that tool and
: insulates the users from its devilish argument syntax.

Yes.  Or have it just read from stdin rather than take args from the
command line.

: We could put aside our GUI selection problems for the time being
: and just pile in on writing little `pw'-like tools for all the stuff
: we want to bring into the first version of the system.  Once critical
: mass has been achieved there, I suspect that the front-end folks would
: start coming out of the woodwork and, in the short term, I could write
: some libdialog based stuff to handle our immediate proof-of-concept
: needs.

I like this idea.  You'll likely want to have as much parsing done as
possible in a common routine as well.  99% of system admin is database
frobbing, abstracted out.  You are adding something, or changing
something or deleting something.  A regular base to build from should
recognize this and build from there....

: P.S. Is it time for us to move this over solely to config?
: I'm not sure if this is -chat material anymore.

Yes.

Warner



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E0vnAEX-0002Fm-00>