Date: Mon, 01 Feb 1999 04:26:44 +0900 From: "Daniel C. Sobral" <dcs@newsguy.com> To: Michael Hope <michaelh@earthling.net> Cc: Mike Smith <mike@smith.net.au>, Warner Losh <imp@village.org>, small@FreeBSD.ORG Subject: Re: compressed executables in low memory environements Message-ID: <36B4AE74.D74B14EB@newsguy.com> References: <Pine.LNX.3.96.990201074335.312B-100000@goldie.nest.pcmedia.co.nz>
next in thread | previous in thread | raw e-mail | index | archive | help
Michael Hope wrote: > > You might be interested in the LZO library: > > http://wildsau.idv.uni-linz.ac.at/mfx/lzo.html > > The compression isnt that great (files are, say, 15% bigger when > compressed than the equivalent file compressed with gzip), but the > decompression routine is _incredibly_ fast and doesnt require a dictonary. Err... having a dictionary was a plus... :-) We want a "fixed-block" compression. For example, so that I could decompress any 512-bytes block of it (aligned on 512, of course :) without context. Well, statistical compression algorithms could be used too, but the ones giving decent compress ratios depend on context. The dictionary based ones are easy to adapt, because all you need to do is dump the dictionary and begin again each time you reach a block boundary. Incidentally, RAR has something similar. It adds "context" data every so and so bytes, so that you can recover data from partically damaged files. Or that's what I think they do, anyway... :-) -- Daniel C. Sobral (8-DCS) dcs@newsguy.com Would you mind not shooting at the thermonuclear weapons? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-small" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?36B4AE74.D74B14EB>