Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 19 Mar 1997 00:03:00 -0600
From:      Richard Wackerbarth <rkw@dataplex.net>
To:        "That Doug Guy" <tiller@connectnet.com>
Cc:        "stable@FreeBSD.ORG" <stable@freebsd.org>
Subject:   Re: -current and -stable mailing lists
Message-ID:  <l03010d06af55265ccc46@[208.2.87.4]>
In-Reply-To: <199703190302.TAA22405@smtp.connectnet.com>

next in thread | previous in thread | raw e-mail | index | archive | help
At 9:01 PM -0600 3/18/97, That Doug Guy wrote:
>On 18 Mar 97 14:42:38 -0600, Richard Wackerbarth wrote:

>>I think this reflects a lack of understanding of the CVS process.
>
>	I made that explicit in my letter, in fact you quoted me. :)
I was simply trying to point out that I agreed that your admitted lack
of knowledge of the details of CVS was the likely source of misunderstanding.
I then proceed to describe some of that detail in an effort to acquaint
you with the workings of the system.

>>within the FreeBSD tree, there are various identifiable branches, 2.1, 2.2,
>>and the "head" branch being the ones of interest. Along any branch labels
>>can be attached; for example - 2.2-RELEASE.  It is possible to retrieve the
>>source code by date or label and from any branch.  From a practical point,
>>2.2-RELEASE is dead. There is a slightly newer version
>>of the 2.2 branch which is the active point on that branch.
>
>	My system allows for retrieving (for example) 2.2.1-Release, or
>2.2-Development (which if I understand you correctly is the "active point
>on that branch."  In regards to using my system for cvs tags, in what way
>is it deficient?
There is only one tag which represents the active point of the 2.2 branch.
That tag must remain invariant as the status of the branch changes through the
various phases of product life cycle.


>>I think the concensus is that there needs to be an additional one to
>>separate 2.2 from 3.x.
>
>	Frankly I'm still fuzzy as to why 2.2 questions can't/shouldn't go
>to -questions, and 3.0 questions can't/shouldn't go to -current.

>>Rather that use list names which change over time, I suggest that we simply
>>create a list for FreeBSD-2.1, a list for FreeBSD-2.2, and a third list for
>>the head of the development branch.
>
>	Can you explain your thinking behind this?
Each branch gets its own unique discussion list. We don't have the problem
which
would occur because the appropriate list changes simply because the FreeBSD
release engineer did some more work.

As you note, "3.0 go to -current" is fine today. Next week (month/year) your
3.0 question going to -current would get a response to the effect "this list is
for discussion of 4.2. Please redirect your question to the xxx list.

>	The path I indicated has 2.1 dying off in 3 months, so a new list
>for it would not really be useful, and will eventually have to be
>re-directed to -questions anyway.

Certainly true. However, for the sake of consistency you should move the
list even if it is going to last for only a few months. You don't want the
needless confusion created because the rule would be "Questions about version
A.B.C go to the 'FreeBSD-A_B list EXCEPT 2.1.7 which is still on
FreeBSD-stable"

>	With respect, I think that you are trying to tie things together
>that are not necessarily related.

>At least, my thesis here is that in
>order for FreeBSD to succeed, the perspective that only programmers are
>welcome at the table has to change.  Please note that I am not accusing
>Richard of this, I don't know him or his views.
I think that my views are much closer to this than those of most of the
FreeBSD development team.

>  What I'm trying to say is
>that the current system is confusing, and in my opinion needlessly so.

>The fact that once you've been around a while and have twiddled with the
>source tree some, it does have some vague sort of relevance is not a
>convincing argument for keeping it.  :)
This does not really have anything to do with the source tree, per se.
I has to do with the fact that various incarnations of FreeBSD have functional
characteristics which are related to the system release rather than its
maturity.

If you want parts for your 1997 Ford, would you go to the "New Car" dealer
or to
the "Used Car" dealer? What if I ask that question in October, 1997?
The "New Car" dealer would say "Sorry, that is last year's model. You must
see the 'Used Car' department" :-(

I think that you are placing too much emphasis on the temporal quality rather
than a permanent one.





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?l03010d06af55265ccc46>