From: Hitoshi Mitake Date: Thu, 11 Feb 2010 14:38:13 GMT Subject: Re: [PATCH 4/4] git-imap-send: Add method to convert from LF to CRLF Message-ID: <4B741655.6090107@dcl.info.waseda.ac.jp> In-Reply-To: <7vbpfye0e5.fsf@alter.siamese.dyndns.org> (2010年02月10日 03:20), Junio C Hamano wrote: > Linus Torvalds writes: > >> So I suspect that CRLF should be unconditional. > > That matches my reading of the RFC. Thanks. OK. I'll make converting LF to CRLF default behaviour. > > Hitoshi, can we have a modified version of 4/4 as a separate patch, so > that we can apply it independently from the rest of the series? Yeah, but as I describe below, I'll rewrite my patch series and these only has 2 parts. > > As to the MD5 implementation, I am somewhat torn. > > Even if your md5 implementation were vastly superiour and faster than > OpenSSL one (I don't know), the use of MD5 is not performance critical > like SHA-1 is (for which we uniformly use Linus's block SHA-1 these days); > the only thing it would be buying us to have our own implementation is one > less dependency for people who do want to use imap-send with CRAM-MD5 but > without SSL support. How common is that combination? Yes, using OpenSSL's md5 is better implementation. But git-imap-send.c also asuumes non existence of OpenSSL. And in theory, CRAM-MD5 and STARTTLS are independent things. So I import MD5 things from libsylph. But, as you told, CRAM-MD5 without STARTTLS is not so common situation. If you permit, I'd like to use MD5 functions of OpenSSL. It seems that OpenSSL also provides base64 stuffs, so I will be able to reduce patch series into 2 part. I'll send these later. Please discard my previous series.