[PATCH] revert: simplify help message, better error message on reverting a merge

Matt Mackall mpm at selenic.com
Mon Jun 6 16:52:53 CDT 2011


On Mon, 2011-06-06 at 23:39 +0200, Adrian Buehlmann wrote:
> On 2011-06-06 23:13, Matt Mackall wrote:
> > On Mon, 2011-06-06 at 22:52 +0200, Adrian Buehlmann wrote:
> > 
> >> (now I'll shut up on this, since Matt wants to keep that line as it is)
> > 
> > I was -completely- open to changing this line provided the replacement
> > actually ADDRESSED THE EXPRESSED INTENT. Otherwise, I would not have
> > spent on the order of two hours corresponding with you on the subject. I
> > would have just said no.
> 
> Yeah. It's indeed amazing how much we wrote about this line. And I don't
> think we made any progress at all.
> 
> > This is not some mysterious personality quirk of mine, it's how software
> > engineering is done. We don't drop imperfect bug fixes, we make them
> > better.
> 
> I really don't know what you are talking about here.

What I object to is:

a) simply dropping the line - this line is an attempted bug fix

You don't drop existing bug fixes on the ground without a replacement.
That's bad engineering.

b) replacing the line with something that doesn't address the bug

If the problem is "the UI is confusing", throwing jargon at it won't
help. And that's just obvious.

-- 
Mathematics is the supreme nostalgia of our time.




More information about the Mercurial-devel mailing list