*BSD News Article 80430


Return to BSD News archive

Newsgroups: comp.unix.bsd.freebsd.misc
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.Hawaii.Edu!news.uoregon.edu!arclight.uoregon.edu!feed1.news.erols.com!uunet!in1.uu.net!news.new-york.net!news.put.com!main.put.com!le
From: Louis Epstein <le@main.put.com>
Subject: Re: FBSD Future...
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: usenet@news.put.com (The Root)
Message-ID: <Pine.BSF.3.91.961011121248.28443A-100000@main.put.com>
References: <Dz37x9.7vC@news2.new-york.net> <199610111548.JAA29164@trout.mt.sri.com>
Mime-Version: 1.0
In-Reply-To: <199610111548.JAA29164@trout.mt.sri.com>
X-Nntp-Posting-Host: main.put.com
Date: Fri, 11 Oct 1996 12:30:21 GMT
Lines: 92



On Fri, 11 Oct 1996, Nate Williams wrote:

> In article <Dz37x9.7vC@news2.new-york.net>, Louis Epstein <le@put.com> wrote:
> >Well,I see that the expected release of 2.2 has been pushed back from
> >late summer '96 to first quarter '97,quite a jump,and I wonder what
> >more there is to it past Walnut Creek's desire to keep the CDROMs
> >semi-annual.
> 
> It really has little to do with WC, and more to do with developer
> burnout.  The 2.X cycle has been very stressful given that that we've
> been running two development branches for almost 2 years, so the
> developers decided to give themselves a break and back off on the 2.2
> release.  (Although, -current [pre-2.2] has been relatively stable as of
> late, so it's looking pretty good.)

Will the delay mean any added features that wouldn't have gotten in this
year?

> >I gather that 2.1.6 will be released on the net only,not on CDROM,so
> >how would one handle an upgrade from 2.1?(No point in going to
> >2.1.5 first if 2.1.6 is coming).
> 
> What's this about 2.1.6?  It's the first I've heard about it, and I'm
> one of the developers. :)

It's mentioned on the freebsd.org web pages..."A 2.1.6 release will
probably occur on this branch some time in the next 2-3 months,just to
clean up a last few remaining nits from 2.1.5".
 
> >How long would I have to take my
> >ISP down to go to 2.1.6,and get it in shape to honor the 9-12 character
> >IDs enabled by a kernel recompile for my 2.1??And what would I actually
> >gain by so doing?
> 
> Upgrading to 2.1.5 will buy you stability and improved security.  There
> were some fairly significant bugs/holes that were fixed, plus as an ISP
> you  *really* want some of the BIND fixes that went into 2.1.5 dealing
> with classless addressing which is becoming more prevalent.  There are
> also alot of cleanups done to the system, plus keeping fairly current
> means that IFF you find a bug/security problem it's alot easier to
> upgrade your sources from a more recent release than an older release.

(Of course,this was the reason to wait for 2.1R rather than 2.0.5...)

> Given that this a volunteer outfit, you'll probably find someone running
> a recent release that can help you out vs. telling you to upgrade if you
> have problems.
> 
> As far as upgrading the sources to have longer ID's, I don't see what
> it'll buy you IMHO, but I'm an old-timer who prefers short email/account
> names. :)

It'll buy me the ability to have someone used to being 
longusername@elsewhere.com not have to be lngusrnm@put.com.
Since I have a number of long IDs in use already,I have to make
sure they don't lose access in the upgrade.I enjoy having a
very short ID(le@put.com) myself,but flexibility counts.

> It took me about 4 hours to upgrade each system to 2.1.5 from 2.0,
> although I could have done it much quicker if I wasn't so paranoid or I
> tried to automate it more.
> 
> >Which SNAPs of 2.2 will be on the CDROM for SNAP subscribers?
> >501 was,626 wasn't,were 801 and 1006?
> 
> I *think* 1006++ will be on a SNAP CD.  Jordan fixed a couple bugs and
> will probably re-roll the 1006 SNAP before putting it on CD.
> 
> >For a got-to-stay-up business like an ISP,it might be better to wait
> >for a 2.2.x after 2.2R gets its bug reports dealt with...I guess
> >the first such release is at least 9 months off.
> 
> I disagree completely.  I have critical systems, and I upgraded every
> one of them to 2.1.5 b/c of stability and security.  2.2 will contain
> alot more 'experimental' stuff (even # release), and as such will
> probably not be as stable as 2.1.5.

You're misunderstanding me here...I'm saying that I would want the most
mature release of the -stable branch,but would hold off on the 2.2.x
branch until x is no longer 0 so that its stability is increased.

> >(64-bit 3.x is presumably years away still).
> 
> 64-bit is probably more like 6.x or something.

Hmm,isn't SCO(which the non-commercial UNIXes are supposed to be years
ahead of) already doing a 64-bit version?

So what major enhancement would you think will precipitate renumbering
from 2.x to 3.x?