Re: [PATCH v3 0/1] CRLF-Handling: bug fix around ce_compare_data()

From: Junio C Hamano <gitster@pobox.com>
Date: May 16, 2016, 16:13 UTC
Message-ID: <xmqqshxi2d80.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <1463413889-12490-1-git-send-email-tboegi@web.de>

tboegi@web.de writes:

Show 13 quoted lines

> From: Torsten Bögershausen <tboegi@web.de>
>
> Changes since v2:
>
> - Only 1 patch, the t6038 needs to go as a separate "series"
>
> has_cr_in_index() uses either
> 	data = read_sha1_file(sha1, &type, &sz);
> or
> 	data = read_blob_data_from_cache(path, &sz);
>
> (I like v2 better, since there is a single code to get a sha object
> into memory, which later will be replaced by a streaming object)

Wait a minute, please. I only asked the reason why you did it that way and mentioned that the end result seemed equivalent. "The end result seems equivalent" does not automatically mean "therefore you must avoid changing the code."

If you still prefer the original code, and your preference is backed by a solid reasoning, don't change the code to a less preferrable version. You may have to explain what you wrote in () above clearly in an updated log message to save other readers from asking the same question as I asked, though.

Thanks.