On Fri, 8 Jul 2016, Óscar Fuentes wrote:

> Just the second commit on the log output:
>
>  Remove just input mark
>
>  * lisp/ibuffer.el (ibuffer-unmark-all): When MARK is not ?\r remove
>  just MARK.
>
> no hint about why it was necessary.

Sorry for not being more clear in my commit, and thanks for mentioned it,
so i can try do it better next time.

IMO, from this commit one emacs developper should conclude:

1) Its a change in file: lisp/ibuffer.el
2) Affects a func. to unmark buffers.
3) Affects the behaviour when an input arg (MARK) satisfy one condition
   (not equal ?\r): it will remove (unmark) just buffers with mark being 
MARK.
*) Fix? Enhancement? Unclear.

After that, a reader interested in the commit could check the diffs: i 
agree that in this case the diffs are not helpful to understand the 
necessity of the change.

Then, a really interested reader could easily put the point in the func.
name between parens and do: F1-f RET
(That automatically confirms 2).

After my commit the function spread 29 lines: it fully matches
within my laptop screen (with average screen size).
The first 3 lines help to understand the thing:
'(defun ibuffer-unmark-all (mark)
   "Unmark all buffers with mark MARK."
   (interactive "cRemove marks (RET means all):")
'
Basically, a command unmarking buffers marked with MARK; if MARK equals 
?\r
then all marked buffers are unmarked.  This is what is written in the 
commit, so the reader should conclude that this is a fix instead of an 
enhancement.

I don't know how much average time could take this for an experience emacs
developer; negligible i would say: probably Eli can do this almost while
taking a nap.

I got the write permissions in this project last month and i am still
a newby in git.  For my lack of experience my commits are far to be
perfect.  Sorry.   I am sure they will be better and better just with the 
practice.

Tino