Thread (1 message) 1 message, 1 author, 2016-06-15

Re: [PATCH] Switch receive.denyCurrentBranch to "refuse"

From: Junio C Hamano <hidden>
Date: 2016-06-15 22:46:04

Johannes Schindelin [off-list ref] writes:
at the cost of annoying

- a few oldtimers
- now, instead of later
Nobody seems to have realized this, but suddenly changing the default to
refuse without giving people enough advance warning to adjust will hurt
not just the old-timers (a rough definition is people who are from the
kernel circle and have been using git since summer of 2005), but people
who picked up a recipe from various how-to web pages to push to a live
repository and updating the checkout that is otherwise never touched by
the humans with its post-update hook running "reset --hard".  Old timers
may be savvy enough to know what has changed and may be able to grudgingly
react, but what is your plans for these recipe following kids?

How many times do I have to repeat that it is much worse to break a
working setup of people without advance warning and sound transition
guidance than having a known breakage that users can be trained to avoid?

And realize that I am not saying we need to keep the known breakage
forever.

The only thing I am saying is that you need to have a smooth transition
plan for changing the default, and a mechanism to guide people in place.

I'll ignore you if you keep repeating "all it takes is for old timers to
flip a switch".  Such an argument shows that you didn't learn a thing
after the 1.6.0 fallout.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help