perrin at apotheon.com
Thu Jan 19 23:14:50 UTC 2012
On Thu, Jan 19, 2012 at 02:36:29PM -0800, Devin Teske wrote:
> > From: Frank Shute
> > The new installer will get better with time.
> The new installer is buggy, and the above maxim is something I'd rather not have
> to deal with when downloading RELEASE software.
I do not dispute that the new installer is "buggy", nor do I agree that
it is "buggy". I have used it twice, without any bugs biting me. That
may just be good luck. I have, however, discovered that usability is no
better than sysinstall; it's just *different*. In fact, in some
respects, it feels more limiting. I suspect some of my issues with it
will be resolved by simple familiarity -- but then, some of those issues
are not due solely to differences between bsdinstall and sysinstall; they
are also due to differences between bsdinstall and *every* console-based
piece of software with that general curses-style appearance. Maybe I'll
never get to quite *that* level of familiarity with bsdinstall,
considering I use a lot of other console-based applications, too.
I do not recall running into any bugs in sysinstall, either, by the way.
Considering how many more times I have used it, I think it is far less
likely that I was just lucky. Perhaps it has bugs, but it must have bugs
primarily with features for which I have (so far) had no use.
If the fact sysinstall does not support some functionality needed for
installation of new versions of FreeBSD (I believe someone has suggested
this is the case) while bsdinstall does is a result of sysinstall's
architecture being insufficiently well organized for the addition of this
functionality to be a reasonable alternative to writing a new installer
instead, I can understand the desire to create and propagate the use of
bsdinstall. In that case, great: I'm glad we're moving forward. If it
is functionality that not everyone needs, I think it might be nice to
offer both installers as options (perhaps bsdinstall as the default, if
we must). As someone who has never really looked into the code used to
handle starting the installation process, I do not know how feasiable
that is, and would appreciate someone who knows from first-hand
experience enlightening me as to whether it's a good idea. It is likely
that many people will not need the new functionality that bsdinstall
would support, if it relates to things like ZFS support, after all.
If the reason it was decided to create bsdinstall and replace sysinstall
was simply to do something new, without particular interest in
maintaining the benefits provided by sysinstall, and without any actual
technical requirement for the new installer, I have a somewhat different
opinion -- one normally reserved for ludicrous exercises of neophilia
like those rampant in the Ubuntu community in particular and the Linux
community in general, breaking all the old ways of doing things just
because someone decided to write some code one day. Did you know that
ifconfig is no longer guaranteed to work as a tool for restarting
networking on Linux-based systems? Are you aware of the Cthulhoid
tentacular horror of the Linux sound architecture, especially with
PulseAudio thrown into the mix? Have you seen the filesystem and shell
environment clutter that is the XDG Base Directory Specification?
Please, let the reasons behind bsdinstall be better than for all of those
I'm inclined to believe that the motives for bsdinstall are good motives,
knowing what I do of the FreeBSD developers' philosophy (maybe not a lot,
but enough to know it tends to eschew such radical changes for change's
sake, in my experience). It may have moved slightly too quickly, but it
may be a movement in the right direction nonetheless, and I hope it is.
I'd just like to know more about the whys and wherefores than statements
(from people who have not indicated where I can see it that they actually
know anything about it first-hand) that sysinstall is "buggy" because the
manpage says so and bsdinstall is not because it's not sysinstall.
> RELEASE software shouldn't be released under the statement "it will get better
> with time". Releasing feature-INcomplete software that is known to be broken
> hurts the FreeBSD impression far more than sysinstall ever could/did. I feel
> your argument is an attempt to justify the egregious offense of foisting
> premature software on the community when in-fact it does NOT replicate even a
> fraction of the abilities of sysinstall.
I also think it's worthwhile to give people the benefit of the doubt, at
least at first. Perhaps the rhetoric can be scaled back a little bit in
this case. Has there been some response to your complaints that I have
not seen that justifies this level of heat?
Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]
More information about the freebsd-questions