From mboxrd@z Thu Jan 1 00:00:00 1970 From: erik quanstrom Date: Mon, 24 Jun 2013 09:03:18 -0400 To: 9fans@9fans.net Message-ID: <7d1c080e58d9aaf4d40fff29ec1bc1e0@lyons.quanstro.net> In-Reply-To: References: <96413B2B-5F5B-4C27-B2B5-483FB9B811D2@ar.aichi-u.ac.jp> <0fd845b9ba300f41b4f0607ede960285@lyons.quanstro.net> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Subject: Re: [9fans] long filenames in cwfs Topicbox-Message-UUID: 67eb16f6-ead8-11e9-9d60-3106f5b1d025 On Sun Jun 23 18:01:11 EDT 2013, skip.tavakkolian@gmail.com wrote: > with 8K names, using base64, one could encode 6111 bytes of data in the > name. i just did a quick inventory[*] of my $home; 74% of my files have > less than 6112 bytes of data. i was proposing a pointer, which could be nil. i would put the pointer where the file name should be=E2=80=94perhaps by setting the first 8 bytes to non-unicode magic and the next 8 bytes to a pointer to the direct name block. so why not change nothing. the first 8180 bytes go in the first direct block anyway. btw, i have far fewer large files than you, but still 25% of the ones bigger than 6112 bytes are smaller than 8180 ; du -a .>[2=3D]|awk '$1>8180 {s++}END{print s/NR}' 0.00325015 ; du -a .>[2=3D]|awk '$1>6112 {s++}END{print s/NR}' 0.00416286 - erik