Unreleased improvements lingering on master

Unreleased improvements lingering on master

[Bjartur Thorlacius]

[-- Attachment #1: Type: text/plain, Size: 859 bytes --]

Dear Emacs developers

Thanks for developing the most extensible editor. I noticed an improvement committed to master on March 2018 that has neither been merged to the emacs-26 branch nor released. Namely encoding .emacs.desktop in UTF-8 rather than mule (commit 5bebacc476259f854d3735a6dc5a639016f1673a by Stefan "le dieux" Monnier). In fact, 2148 repository files differ between master and emacs-26 according to <https://repo.or.cz/emacs.git/commitdiff/refs/heads/emacs-26?hp=master#patch658>. Both master and emacs-26 are actively developed, yet as far as I can see only emacs-26 is actively being released. Is it time to release master under another major version number, say 27? Or are users which want to enjoy the fruits of the development on master from the past years expected to build GNU Emacs themselves?

Best regards,
Bjartur Thorlacius


[-- Attachment #2: Type: text/html, Size: 3044 bytes --]

Re: Unreleased improvements lingering on master

[Paul Eggert]

On 10/18/19 9:20 AM, Bjartur Thorlacius wrote:
> Both master and emacs-26 are actively developed, yet as far as I can see 
> only emacs-26 is actively being released. Is it time to release master 
> under another major version number, say 27?

Yes, we plan to do that soon (for some value of "soon" :-). emacs-26 
development is effectively done.

Re: Unreleased improvements lingering on master

[Phil Sainty]

Hi Bjartur,

On 19/10/19 5:20 AM, Bjartur Thorlacius wrote:
> Both master and emacs-26 are actively developed, yet as far as I can
> see only emacs-26 is actively being released. Is it time to release
> master under another major version number, say 27?

You are simply observing the normal development cycle for GNU Emacs.

The master branch contains what will eventually become the next major
release, which will indeed be version 27 (with 27.1 being the initial
stable release version).


> Or are users which want to enjoy the fruits of the development on
> master from the past years expected to build GNU Emacs themselves?

Exactly so.  Or at least, OS distributions are generally not in the
habit of packaging unstable versions for public consumption (you may
or may not be able to find such packages from third-parties for your
own OS, but official OS packages are unlikely).

In general users can (and do) build Emacs themselves from the master
branch in order to participate in the development and testing process
or, in some cases, merely to gain access to the new improvements -- at
the obvious greater risk of encountering bugs.

In due course there will be one or more 'pretest' releases of Emacs 27
intended for wider consumption by end-users, and a full release tends
to follow within the following few weeks or months, depending on the
outcome of the testing.

There is no fixed schedule for this (that I am aware of, at any rate),
and I don't know the maintainers' current intentions, but I would guess
that we'll see the 27.1 release in 2020.


-Phil