Re: [PATCH v2 00/31] refactor rebase

From: Junio C Hamano <gitster@pobox.com>
Date: Feb 15, 2011, 00:36 UTC
Message-ID: <7v62smqgkb.fsf@alter.siamese.dyndns.org>
In-Reply-To: <alpine.DEB.2.00.1102132047500.4253@debian>

Martin von Zweigbergk <martin.von.zweigbergk@gmail.com> writes:

> Improve the situation by allowing options to be combined with sub
> commands both before and after the sub command itself. Persist the new
> value to make sure it is remembered across interruptions.

I somehow suspected that it would be much more messy, but with your refactoring, things do not look that bad as I feared ;-)

The only use case it might be useful, however, is to start without -v, realize that a verbose output may help you resolve conflicts better, and continuing with the verbose option, which conceptually should be helpful.

But in practice I doubt the ability to ask for -v output in the middle would be so helpful, so doing this may not be worth it.

I am not a regular rebaser anymore, so I wouldn't be a good judge beyond just my gut feeling. Comments from heavy rebase users who often see conflicts are appreciated.