Re: [PATCH v6] safecrlf: Add mechanism to warn about irreversible crlf conversions

From: Junio C Hamano <gitster@pobox.com>
Date: Feb 3, 2008, 10:50 UTC
Message-ID: <7vbq6yuxeh.fsf@gitster.siamese.dyndns.org>
In-Reply-To: <12019685081997-git-send-email-prohaska@zib.de>

Steffen Prohaska <prohaska@zib.de> writes:

> CRLF conversion bears a slight chance of corrupting data.
> ...
> thing to do, while for binary file it corrupts data.

The above 25-line or so are well written and deserve to be in the end user documentation somewhere, I think, to explain why it is a good idea to have these warnings to them..

> This commit modifies git apply to fail even if safecrlf=warn,
> because git apply writes its changes back to the work tree
> immediately.  The user would not have a chance to backup the old
> version of the file if only a warning was printed.

I do not get this logic at all.

The whole point of git-apply is to apply the patch. If you say --whitespace=fix and some contents (say one of the testsuite files in our t/ directory) needed to keep trailing newline, you obviously are left with a broken result, and you would recover by checking it out from index or HEAD and reapply. Why shouldn't the same principle hold here?

I haven't looked at the code of this round yet, but I promise I will.