Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 2 Jan 1999 14:37:28 -0500
From:      Christopher Masto <chris@netmonger.net>
To:        current@FreeBSD.ORG
Subject:   Re: wanton Atticizing is bad
Message-ID:  <19990102143728.A23411@netmonger.net>
In-Reply-To: <19981227163320.A998@tidalwave.net>; from Lee Cremeans on Sun, Dec 27, 1998 at 04:33:20PM -0500
References:  <Pine.BSF.4.05.9812271249290.6613-100000@janus.syracuse.net> <68606.914783086@critter.freebsd.dk> <19981227163320.A998@tidalwave.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Dec 27, 1998 at 04:33:20PM -0500, Lee Cremeans wrote:
> On Sun, Dec 27, 1998 at 07:24:46PM +0100, Poul-Henning Kamp wrote:
> > In message <Pine.BSF.4.05.9812271249290.6613-100000@janus.syracuse.net>, Brian 
> > Feldman writes:
> > >What was the reasoning behind ft though?
> > 
> > It's badly implemented and many capable persons have failed to find a
> > way to solve the rather serious problems it suffers from.  ATAPI tapes
> > seems to be the way of the future for cheap tapes.
> 
> If ft were as nice as the SCSI tape driver, then I'd say keep it, but ft is
> extremely crocky (I know, I tried using it once) and very, _very_ slow since
> it has to use the floppy drive controller.

IMHO, "slow" is not the problem.  Yes, floppy tape can range from
abysmally slow to annoyingly slow, but I'm sure there are a lot of
people who don't care - pop a tape in the drive and when it's done,
it's done.

I looked at "fixing" the ft driver at one point.  I was going to make
it look just like a regular SCSI tape to the outside world.  It's
doable, but I have yet to accumulate any free time for the task.  In
any event, it would probably not keep _any_ of the existing ft code.
If I ever do produce such a driver, I will attempt to use whatever
module mechanism we have by then.

Now I want to comment on the cleanups, though I suspect this may fall
on deaf ears.  FreeBSD is a wonderful thing.  I have been using it
extensively in the Internet service business for several years.
Stable and current branches are installed on around 10 servers at my
current company, doing everything from web serving, to usenet news, to
faxing and paging, to temperature monitoring.  I have installed it on
my personal machine at home, and on my office workstation.  I know
many people who use FreeBSD after seeing how successful we have been
with it.  I have personally converted a nearby to FreeBSD after their
NT-based web, e-mail, and file sharing systems disappointed them time
after time.. and since the change, they have not had a single failure
or complaint.  I actively remind people who make the error that Linux,
while certainly deserving of praise, is not the only free Unix out
there.  I have found bugs, submitted PRs and patches, written drivers,
and managed to get the occasional submitted-by in CVS somewhere.  In
short, I am a BSD devotee.

That said, I may soon be forced to jump ship and switch to Linux.

I conceed that many of the things being vaporized are unimportant on
my big server machines that need little more than high-end SCSI
drivers and a stable kernel.  But that's not the only place I use
FreeBSD.  For the past six months, I have had an exclusively
FreeBSD-based desktop, at work and at home.  I do all of my
computer-related work AND play on FreeBSD.  That means that things
like sound cards and x-10 and pccard and other "toys" are very
important to me.  And I am a big advocate of free, open-source
software, so I'm not going to go out and buy OSS.  I would hate to see
FreeBSD move to a model where virtually all of the drivers were
proprietary loadable modules.  I would also hate to see
FreeBSD-for-the-desktop die.  I already spent enough time working
around the linuxisms in things like GNOME, but I've always felt that
it's worth the effort to have a choice.  But I only have so much time
to futz with this stuff.. it becomes more and more tempting to just
use that other free unix that already has support for my toys.

What I'm getting at is that if we want to stand a chance of being able
to increase FreeBSD's visibility, it has to be _useful_ to _more_
people.  It seems like some of these drivers are being removed simply
because nobody on the core team uses them.  Hell, I don't use most of
them either, and I'd love to have less to comment out when I configure
my kernel, but I've already had the experience of having to say
"If you upgrade to 3.0, you will lose support for your SCSI card."

In summary, please don't remove working drivers that have no
replacement yet just because they're old.  At least go through a
release or two with a comment next to them saying "We need a
maintainer for this or it may go away".  Give it a chance.

Since this is way too long and rambling to have been read through, and
it doesn't belong as an appendix to a comment on the ft driver, I
think I'll stop here.
-- 
Christopher Masto        Director of Operations      NetMonger Communications
chris@netmonger.net        info@netmonger.net        http://www.netmonger.net

    "Good tools allow users to do stupid things." -- Clay Shirky

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990102143728.A23411>