Date: Mon, 5 Apr 2021 11:29:00 +0700 From: Eugene Grosbein <eugen@grosbein.net> To: Dave Cottlehuber <dch@skunkwerks.at>, FreeBSD Stable <freebsd-stable@freebsd.org> Subject: Re: Deprecating base system ftpd? Message-ID: <606A920C.8030502@grosbein.net> In-Reply-To: <a353d9fa-c5fc-41c5-99ab-ca810784dd68@www.fastmail.com> References: <CAPyFy2AbP2X339zbemZ9Y8edjNKdyygnR9mH48Q78nxwDtOBAg@mail.gmail.com> <0070fa8d-1e9c-89c7-f0a8-40aace3030d8@quip.cz> <a353d9fa-c5fc-41c5-99ab-ca810784dd68@www.fastmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 05.04.2021 06:25, Dave Cottlehuber wrote: > Eugene mentioned the convenience of ftpd in the same sentence as ipsec. > I'm willing to bet those systems have ports installed too. Ports/packages are great but they are not replacement for solid operating system with bundled software tested and proven with time. > If speed is an issue, HTTP supports pipelining, compression, chunked > encoding, & parallel connections. I'm not sure ftpd is even in the same > game anymore. Compression and various encodings of raw data are not good for speed. sendfile(2) system call used by ftpd to send raw data is good for speed. Unlimited CPU power should not be assumed. > The more code we hang onto in base, the larger the millstone around our > necks when moving forwards. Each individual opportunity to slim down > base *in itself* is not significant, but cumulatively they represent > gridlock. > > For each removal or deprecation, please consider, is this worth holding > the project back for? Our ftpd code does not hold the project back in any way. It's here, it works, it's very good. High quality bundled software is what we love FreeBSD for. Unfortunately, ports tend to rot more quick due to some known reasons.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?606A920C.8030502>