Mailing List Archive

On Thu, 17 Aug 1995, Robert S. Thau wrote:
> Ooops... I forgot to mention where to pick up my 10x spin. As with
> other stuff which hasn't yet passed group muster, it's on ftp.ai.mit.edu,
> in ftp://ftp.ai.mit.edu/pub/users/rst (with the obvious filenames therein).
>
> Also, if anyone wants to keep score, then, having tested these two
> additional patches fairly extensively, I'm +1 for a new beta with this
> set of bug fixes, so users see some improvement while we continue to work
> on the rest of the bugs...

+1 as well (compiled and running on hyperreal). It's been worth the wait,
but we should wait no longer.

Brian

--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
brian@organic.com brian@hyperreal.com http://www.[hyperreal,organic].com/

On Thu, 17 Aug 1995, Rob Hartill wrote:
> > > in ftp://ftp.ai.mit.edu/pub/users/rst (with the obvious filenames therein).
>
> It has the same old Makefile. Is that right ?. What happened to David's
> new one ?

I thought we agreed that 0.8.x's would be *bugfixes*, 0.9 would be a
batch of new features (with 0.9.x bugfixes), and finally Apache 1.0 at
some point.

Brian

--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
brian@organic.com brian@hyperreal.com http://www.[hyperreal,organic].com/

> > in ftp://ftp.ai.mit.edu/pub/users/rst (with the obvious filenames therein).

It has the same old Makefile. Is that right ?. What happened to David's
new one ?

rob

> On Thu, 17 Aug 1995, Rob Hartill wrote:
> > > > in ftp://ftp.ai.mit.edu/pub/users/rst (with the obvious filenames therein).
> >
> > It has the same old Makefile. Is that right ?. What happened to David's
> > new one ?
>
> I thought we agreed that 0.8.x's would be *bugfixes*, 0.9 would be a
> batch of new features (with 0.9.x bugfixes), and finally Apache 1.0 at
> some point.

Okay, if that's what we agreed.

0.8.10x is installed on my HP. No problems to report.

I'll put it on Cardiff soon.


rob

Ooops... I forgot to mention where to pick up my 10x spin. As with
other stuff which hasn't yet passed group muster, it's on ftp.ai.mit.edu,
in ftp://ftp.ai.mit.edu/pub/users/rst (with the obvious filenames therein).

Also, if anyone wants to keep score, then, having tested these two
additional patches fairly extensively, I'm +1 for a new beta with this
set of bug fixes, so users see some improvement while we continue to work
on the rest of the bugs...

rst

>(the DEFAULT_CC hack is roughly what I had in mind, btw, with the overriding
>working, if the user has selected something else in the Makefile, by
>generating
> CC= whatever
>instead of
> CC= $(DEFAULT_CC)

--- this is a few lines of awk in building Makefile.go).
No it's not! Simply put that code at the top of Makefile. No awk required.

David.

> Date: Thu, 17 Aug 1995 09:32:05 +48000
> From: Brian Behlendorf <brian@organic.com>
>
> I thought we agreed that 0.8.x's would be *bugfixes*, 0.9 would be a
> batch of new features (with 0.9.x bugfixes), and finally Apache 1.0 at
> some point.
>
> A note on nomenclature...
>
> I've gotten the suggestion, from someone at the W3C, that when we have
> a stable 0.8.x, we might be better off calling it 1.0, and calling the
> experimental release series with the new features we're planning 1.1.x
> instead of 0.9.x, to give the impression that *we* believe that the
> final 0.8.x thing is a stable, usable product (there may be people who
> could use it, but are currently scared off by a 0.x version number).
>
> This wouldn't be a change in plans, just a change in names.
> Just a thought...
>
> rst
>

I whole heartedly agree with this. I think it's time for 1.0.
It is a major change from existing code, and the Apache groups
first product.

+1

Date: Thu, 17 Aug 1995 09:32:05 +48000
From: Brian Behlendorf <brian@organic.com>

I thought we agreed that 0.8.x's would be *bugfixes*, 0.9 would be a
batch of new features (with 0.9.x bugfixes), and finally Apache 1.0 at
some point.

A note on nomenclature...

I've gotten the suggestion, from someone at the W3C, that when we have
a stable 0.8.x, we might be better off calling it 1.0, and calling the
experimental release series with the new features we're planning 1.1.x
instead of 0.9.x, to give the impression that *we* believe that the
final 0.8.x thing is a stable, usable product (there may be people who
could use it, but are currently scared off by a 0.x version number).

This wouldn't be a change in plans, just a change in names.
Just a thought...

rst

In reply to Robert S. Thau who said
>
> Date: Thu, 17 Aug 1995 09:32:05 +48000
> From: Brian Behlendorf <brian@organic.com>
>
> I thought we agreed that 0.8.x's would be *bugfixes*, 0.9 would be a
> batch of new features (with 0.9.x bugfixes), and finally Apache 1.0 at
> some point.
>
> A note on nomenclature...
>
> I've gotten the suggestion, from someone at the W3C, that when we have
> a stable 0.8.x, we might be better off calling it 1.0, and calling the
> experimental release series with the new features we're planning 1.1.x
> instead of 0.9.x, to give the impression that *we* believe that the
> final 0.8.x thing is a stable, usable product (there may be people who
> could use it, but are currently scared off by a 0.x version number).

This is psychology of course but you have to take things like this into
account when you're trying to win people to your cause :-)

FreeBSD did something similar. When we were ready for a release
that we felt was actually useable it was called 1.0, even though
there were millions of things left to do, you'll never reach 1.0 if you
keep trying to get everything working. The general end-users won't
pick up 0.* code since they just assume it's not ready for general
use yet and they'll wait until until you do a "real" release that's
1.x.

On the numbering scheme, think about this now, we're still confusing
ourselves in FreeBSD because we never got this right (we might finally have
this sorted now after two years, the technical problems are nothing
compared to the organisational one's).

If you're planning long period between "official" releases then you might
want to release bug-fix updates. In that case you'll need version number
space so you can do releases, like 1.0.1, bug fixes to 1.0 etc.

You want you're next true release to be 1.1, since that's what the general
public expect (more psychology). If you're going to to interim, development
releases you'll need a numbering scheme. I'd recommend something like
1.1-dev-x.x which shhows that it's a precursor to the 1.1 release and is
snapshot x.x, you could use dates and thigns too, it's up to you but
you want to get this right because when Apache gets a large user base you'll
start getting queries like, "I'm seeing this bug, I've got Apache X.x" and
unless X.x is clearly identifiable as a particular snapshot of the code
you'll have a hard time working out whether it's an old bug that been
quashed or something that needs looking at.

--
Paul Richards, Bluebird Computer Systems. FreeBSD core team member.
Internet: paul@FreeBSD.org, http://www.freebsd.org/~paul
Phone: 0370 462071 (Mobile), +44 1222 457651 (home)