Re: Emacs 26.1 release branch created

[João Távora]

Eli Zaretskii <address@hidden> writes:

> I'm confused: why are there parts of this work on emacs-26, and other
> parts on a scratch branch?

Sorry, I'm afraid this isn't easy to understand, though I believe John
has already and the matter is solved. Let me try again. As I explained
in the part that you cut out:

>> More than a month ago, I asked here whether this work, that I
>> prematurely and mistankenly pushed to the then-master branch in error,
>> should be reverted.

These are the two commits that I prematurely pushed to what was then master:

* 13993c46a2..: João Távora 2017-08-17 Add flymake-backends defcustom
* eb34f7f5a2..: João Távora 2017-08-17 Split flymake.el into flymake-proc.el 
and flymake-ui.el

Some days later, on Aug. 21, I noticed the mistake and wrote here
(https://lists.gnu.org/archive/html/emacs-devel/2017-08/msg00460.html)

>> Unfortunately, I seemed to have jumped the gun and inadvertently
>> pushed these commits into master instead of keeping them in the
>> scratch/flymake-refactor branch until I got some reviews, which
>> naturally was my intention. Sorry about that.
>> [... They don't break
>> anything, but] since they were pushed in error, I can revert them if
>> someone feels that's necessary.

The only reply that I got to this was Stefan's, on Aug. 22 who said the
work was going "in the right direction", so I shouldn't revert it. And I
didn't.

In the meantime I was sidetracked and you created emacs-26. So this is
why there is (rather "was", read below) this half-baked,
going-in-the-right-direction work on the emacs-26.

So a few days ago, the work was in both emacs-26 and on
master. Following John's opinion in this thread (and my first choice) I
reverted it on emacs-26 and marked it "don't merge". I took care that
this reversion kept minor bugfixes by Paul and Sam commited on top of it
in the meantime. The 'git revert' commits are:

* ce540f8a68..: João Távora 2017-09-27 Revert "Split flymake.el into 
flymake-proc.el and flymake-ui.el"
* 7cf59c6635..: João Távora 2017-09-27 Revert "Add flymake-backends defcustom"

Besides the title, the commit messages also explain the situation in
detail. So now, the half-baked work is only on master.

> Does the latter continue where the former
> left off? or is it an entirely different code?

It continues where the former left off, but very heavily expands on it,
and changes some early design decisions.

> Also, how long (calendar-wise) do you envision it to take you to get
> to the point where the code is not "half-baked"?

The code is 90% baked right now.

I requested comments yesterday, and Stefan's review (in the "Flymake
refactored" thread in this list) was very speedy and very productive
(naturally I would very much appreciate other opinions). I'm more than
halfway through integrating Stefan's suggestions, and none seem very
controversial. I'm very interested in getting this onto master ASAP,
since I have other pressing matters to attend to.

If you are considering opening an exception and letting it go to
emacs-26, I have this to say: the code is stable in my testing,
backwards compatible to both user and third-party lisp (with few
exceptions that I can detail). I have expanded the
automated test coverage quite a bit, fixed longstanding bugs, and
also started rewriting the documentation.

So I'd say:

* "one week" to master (where, AFAIU, doc and NEWS aren't showstoppers)
* "three weeks" to emacs-26 (NEWS, doc and polishing)

Obviously, less if someone helps out.

João