Date: Mon, 13 Mar 1995 16:22:59 -0800 (PST) From: Steven G Kargl <kargl@troutmask.apl.washington.edu> To: phk@ref.tfs.com (Poul-Henning Kamp) Cc: freebsd-hackers@freefall.cdrom.com (FreeBSD) Subject: Re: install compressed binary patch Message-ID: <199503140023.QAA15989@troutmask.apl.washington.edu> In-Reply-To: <199503140005.QAA01137@ref.tfs.com> from "Poul-Henning Kamp" at Mar 13, 95 04:05:34 pm
next in thread | previous in thread | raw e-mail | index | archive | help
According to Poul-Henning Kamp: > > 1. The patch. You shouldn't need to fork a process to make the > symlink, but then again, you shouldn't need the symlink in the > first place ? Whoops, I have misunderstood how your psuedo-device has worked since, well..., since I've used it. Damn, I better look closer at the system. I was under the impression that a symlink was mandatory to tell the kernel that the binary was compressed. Well, I guess I can remove the sym link part. > 2. I'm not sure it makes much sense to do it in the first place. > > Let me explain what I mean. On any machine capable of doing a make world, > you should not need to used gzip bin's on a large scale. My 350 MB FreeBSD partition was at one time capable of make world, but lately I have to remove /usr/local before a build. We seem to be adding more to FreeBSD than removing:) > > Wouldn't you gain more diskspace if you told cc(1) about ".gz" files for > instance ? source compress better than binary I'd expect... Actually, make world was a (poor?) example. But, consider the installation on a production machine of some of the ports. The binary for Octave was over 4 MB before compression. With `gzip -9', the binary is around 750 KB. I get similar compression for other large binaries. The `-z' would be useful perhaps for XFree86 where the site.def(?) file allows one to specify the install program and install flags (if i recall correctly). Then, you can automatically have X built with compressed binaries. > Now THERE'S a project: > > Take a copy of the "nullfs", and call it "gzipfs", and it's perfectly OK > if it can only mount read-only. > The only difference from nullfs should be that if the file > when read contains a "gzip" header, you uncompress it on the fly... > The "inflate" code is already in the kernel... > > Poul-Henning, > (trying to lure another innocent victim into kernel-hacking...) More like naive victim. This is my first attempt at actually missing up my src tree. I read -hacker, -current, and cvs-all, but I am a complete idiot when it comes to kernel stuff. Maybe someday. -- Steven G. Kargl | Phone: 206-685-4677 | Applied Physics Laboratory | Fax: 206-543-6785 | University of Washington |---------------------| 1013 NE 40th St | FreeBSD 2.1-current | Seattle, WA 98105 |---------------------|
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199503140023.QAA15989>