Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!sun-barr!cs.utexas.edu!usc!enterpoop.mit.edu!ai-lab!hal.gnu.ai.mit.edu!not-for-mail From: mycroft@hal.gnu.ai.mit.edu (Charles Hannum) Newsgroups: comp.os.386bsd.development Subject: Re: File Truncation Philosophy Date: 4 Apr 1993 14:44:42 -0400 Organization: dis Lines: 21 Message-ID: <1pnaaq$qqv@hal.gnu.ai.mit.edu> References: <C4tJ6C.C17@ns1.nodak.edu> <deeken.733841578@iti.informatik.th-darmstadt.de> <1993Apr3.211204.1723@peavax.mlo.dec.com> NNTP-Posting-Host: hal.ai.mit.edu In article <1993Apr3.211204.1723@peavax.mlo.dec.com> paik@mlo.dec.com (Samuel S. Paik) writes: > > While this is possible for the local case, it is not possible for > NFS. NFS's rather weak semantics allows some other process to write > all over a file that another process is using for an executable... > Fixing the local case is doable, but I've been pushing for a more > general solution. (shared and exclusive file locks) 4.3 on a HP 300 just logs `pid XXXX killed due to text modification', and does the appropriate thing. I don't know how it tells this offhand. If you're going to introduce state (locks) into NFS, why don't you just make ETXTBSY work? -- \ / Charles Hannum, mycroft@ai.mit.edu /\ \ PGP public key available on request. MIME, AMS, NextMail accepted. Scheme White heterosexual atheist male (WHAM) pride!