[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

'unstable' (was Re: potato late, goals for woody (IMHO))



> > i think that the ideal word is one that doesn't unduly worry
> > non-technical ppl, but still provides a sense of warning to
> > those who seek a conservative release.
> > 
> > i suggest 'latest'.

> then 'latest' could be 'stable' while 'in-progress' could be 'unstable'... I
> believe that it is just a matter of taste probably. While should we continue on
> this subject?

should we continue with this subject? i don't know. i will post
one more message and see if there is any interest; if not, i
will give up.

imo, this issue is sufficiently orthogonal to the discussions
about changing the release process, that it warrants separate
consideration.

this message is amazingly long for such a seemingly trivial isssue.
that's because i think the *effect* of this issue is very
significant, even if the cause and a solution are fairly trivial.

--

ok, here goes.

use of the term 'unstable' creates serious, negative, unintended
consequences for debian.

i can imagine that the term 'unstable' was deliberately picked
to imply complete unreliability, regardless of any technical
evaluation of a *particular* snapshot, to force those who wish
to be at all conservative to use 'stable' or a non-Debian distro.
if so, it is effective.

i could also imagine that the term 'unstable' was picked merely
to convey that it *might* be unreliable, and that one should do
a technical evaluation of a particular snapshot to determine its
suitability for a given situation.

or, it might merely imply it's changing, with no intended overtones
regarding reliability.

if either of the latter two interpretations are intended, or
something like them, then there is a problem, imo, in that the
actual effect is a long way from the one intended.

--

based on anecdotal evidence (including several messages posted
to this list), for many, possibly most, of those who need to
install a system for others, especially in a business context,
and possibly mostly in an english-as-primary-language context
(which influences interpretation of 'unstable'), selecting a
system called 'unstable' represents an unacceptable, or at
least problematic, risk to their career/client relationship.

to emphasize: *regardless of technical and situational merit*,
i would not risk using a system which was called 'highly
dangerous, do not install', for a system intended to be used in
a non-test situation, if i knew there was a good chance that
the client would come to know that that was the system's name.

--

given that words matter, what about the word 'unstable'?

i would guess that many, probably the majority, of management
staff whose first language is english, would interprest 'unstable'
in a computer context to mean, more or less, 'frequently crashes'.
*i* understand it really refers to 'frequently changes', which is
not necessarily the same thing, but that's irrelevant. i would
look stupid if it crashes, regardless of technical and situational
merit.

assuming this is so, does anyone *not* accept that calling a
release 'frequently crashes', when it is the only practical
choice for many installations, will lead to significantly
diminished use of that release, and therefore, distribution?
does anyone *not* think that such reduced use is a problem
worth considering?

--

i'll not list any suggested alternatives this time. the first
step is to establish consensus about whether there is or is
not a problem that needs to be considered.

-rdjm


Reply to: