Re: CVS is the `released version'

[David Kastrup]

"Robert J. Chassell" <address@hidden> writes:

>     It has worked over the last months and years quite well, but do you
>     really want to recommend people to use CVS Emacs when heavy development
>     is under way? Say, when those unicode-2 branch gets merged?
>
> Then I recommend people use an older version.  That is what the
>
>     cvs update -D "10 days ago"
>
> command is for.

Replacing one random date with another is not helping any.  It takes
expertise to know which might be the best date.  A release is
something where the developers as a set of experts agrees.  This
developer knowledge can't be reliably replaced with dice.

It is also hampering development if developers feel compelled to keep
Emacs CVS trunk in a workable state at all times.

> David Kastrup <address@hidden> says 
>
>    ... A release is a point of stability, one where one tries to
>    make reasonably sure that the overall consistency (of packages
>    working together and with the core, and of code and
>    documentation) is in reasonable shape, for whoever happens to use
>    the software.
>
> That presumes most people are not going to contribute, which may
> well be true.

I don't see this at all and don't have the slightest idea why you
would think such a thing.

> The view may not be a presumption, it may be an accurate
> description.  But the issue may be one of morals, not evidence.  The
> argument may be that people should find it easy to contribute.

And they can't contribute if there contributions never end up on a
user's computer.

> In any case, David Kastrup's view does not reflect the experience
> those who I have been calling (to myself) `the ancients'.  RMS is a
> wonderful example of them, since he first wrote Emacs in the 1970s
> and does not change habits if he can keep them.

You should be careful about your examples.  RMS has (quite prudently)
changed his working style on Emacs and the development style several
times.  It is now maintained in CVS, and the CVS has been made open to
public, too.  And he is about to change the habits again, as can be
witnessed by his call for a new Emacs maintainer.

So Richard is quite aware himself when it becomes infeasible to keep
his habits.

>    But never releasing anything for which one has at least some
>    inclination to stand behind it and call it "this is as good as it
>    gets right now" is not a good idea either, in my book.
>
> The point I am trying to make is that some people think Emacs is
> released every day.  That is the opposite of `never releasing
> anything'.  Often enough, but not always, you can use those
> releases.

Because they aren't releases.

> But others do not think of those updates as releases.

Word games don't help in the current situation.

-- 
David Kastrup