bug#20322: 25.0.50; indent-tabs-mode should default to nil

[John Wiegley <jwiegley@gmail.com>]

>>>>> Dmitry Gutov <dgutov@yandex.ru> writes:

> I can't imagine it happening otherwise. What else? Tab-fueled users'
> rebellion on Emacs mailing lists?

You're forgetting one thing: One cannot, by casual observation, tell that this
change has been made, since both spaces and tabs are invisible.

I'll tell you a little story. I'm making it up, but it's more probable than
you might think.

Imagine a programmer: we'll call him Joe. Joe works in a Java shop that
mandates the use of tabs in source files. They exist. I've worked in such
shops.

Joe is fine with this, because Emacs inserts tab characters; so he happily
goes about his task. He then sees that Emacs 25.1 comes out, and upgrades.
After all, a newer Emacs just means more good stuff, right?

However, without any indication (who reads the entire NEWS file? I don't), his
TAB key now inserts spaces. Joe doesn't know this, because everything appears
exactly the same.

But the head engineer who reads every diff notices. He storms into Joe's
office, wondering why he's suddenly ignoring the coding standard. Joe has no
idea way, and everyone is mystified for a moment. A quick Google explains the
reason.

Now Joe has lost face, and he's lost some trust in Emacs, because why did we
make such a change? Can't people who don't want tabs just configure *their*
Emacs, he thinks to himself?

And most of all, he'll be upset with me for letting this happen on my watch.

Everyone who uses Emacs and would be made happy by this change, has already
configured their Emacs not to use tabs. So they won't notice. Everyone who
doesn't use Emacs and would be made happy by this change, won't know that
things were different before. So they won't notice.

The people who will, finally, notice, will be everyone like Joe who came to
depend on their environment not changing long-held defaults without notice.

Therefore, I'm closing this as "wontfix". Feel free to re-open when the next
maintainer arrives. Take this as also establishing a general policy of not
changing long-held defaults without compelling reasons, rather than what we
think people will want.

-- 
John Wiegley                  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com                          60E1 46C4 BD1A 7AC1 4BA2