Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!husc6!cmcl2!brl-adm!adm!Postmaster%FINFUN.BITNET@WISCVM.WISC.EDU From: Postmaster%FINFUN.BITNET@WISCVM.WISC.EDU (PMDF Mail Server) Newsgroups: comp.unix.questions Subject: Undeliverable mail Message-ID: <9729@brl-adm.ARPA> Date: Sat, 10-Oct-87 15:55:20 EDT Article-I.D.: brl-adm.9729 Posted: Sat Oct 10 15:55:20 1987 Date-Received: Mon, 12-Oct-87 07:10:46 EDT Sender: news@brl-adm.ARPA Lines: 356 The message could not be delivered to: Addressee: pietilainen Reason: %MAIL-E-NOSUCHUSR, no such user PIETILAINEN at node OPMVAX ---------------------------------------- Received: from FINFUN.BITNET by OPMVAX.KPO.FI; Sat, 10 Oct 87 21:43 O Received: From WISCVM(SMTP) by FINFUN with RSCS id 4666 for PIETILAI@FINFUN; Sat, 10 Oct 87 21:43 O Received: from SEM.BRL.MIL by WISCVM.WISC.EDU ; Sat, 10 Oct 87 14:22:28 CDT Received: by SEM.BRL.ARPA id ab04419; 30 Sep 87 6:54 EDT Received: from SEM.BRL.MIL by SEM.brl.ARPA id ab03529; 30 Sep 87 3:02 EDT Received: from sem.brl.mil by SEM.BRL.ARPA id aa03505; 30 Sep 87 2:45 EDT Date: Wed, 30 Sep 87 02:45:40 EST From: The Moderator Subject: INFO-UNIX Digest V4#030 To: INFO-UNIX@BRL.ARPA Reply-to: INFO-UNIX@BRL.ARPA Message-ID: <8709300245.aa03505@SEM.BRL.ARPA> INFO-UNIX Digest Wed, 30 Sep 1987 V4#030 Today's Topics: Re: Looking for a high-quality digital voice synthesiser Re: AWK Re: Info about ratfor and/or efl Eyacc Invisible Ascii with VI Exclusive file creation and NFS. csh environment variables and special commands... variables set in sh while loop don't get set? ps -s /dev/null kills system? ----------------------------------------------------------------- From: Greg Hackney Subject: Re: Looking for a high-quality digital voice synthesiser Date: 25 Sep 87 18:01:19 GMT To: info-unix@brl-sem.arpa In article <2138@pbhyf.UUCP>, tlh@pbhyf.UUCP (Lee Hounshell) writes: >I'm looking for a high-quality digital voice box that can be >used with any System V Unix machine.. through an RS-232 interface. >Lee Hounshell - ptsfa!pbhyf!tlh 2600 Camino Ramon, #4E700 >415-823-2432 San Ramon, CA. 94583 You might look at Digital Equipment Corporations DecTalk Model DT-01 voice synthesizer. We have 3 of them that are used for calling in to get voice mail on System 5 Unix machines. There are several standard male and female voices, or you can design your own. The input is RS232 plain old ascii text, or you can input phonetically. DecTalk has an internal speaker, and an external speaker jack. It also has jacks to attach to a Touch-Tone phone line. It can answer the line, prompt for touch tone responses, etc. It can also originate a call and prompt for responses under computer control. The manual gives examples of C programs necessary to drive the interactive calls. Or, you can plug a dumb terminal into it and just type sentences for fun. Price...I think around $2,500. I wrote some C programs that interface the standard unix mail files so users can get their mail via the phone from home, but I have always toyed with the idea of hooking one to a home alarm system via a PC, you know, something like, "Greg! There is a car coming up the driveway". I can hardly wait for the voice recognition technology to be better developed. -- Greg Hackney Southwestern Bell Telephone Co. ihnp4!killer!jgh ----------------------------- From: Guy Harris Subject: Re: AWK Date: 26 Sep 87 04:32:11 GMT Sender: news@sun.uucp Keywords: awk, unix S5R3.1, 4.3BSD To: info-unix@brl-sem.arpa > Does anyone know what the new features are in this "latest" version > of awk ?? For example, are functions included ?? > It gets frustrating to keep encountering all of awk's new features > in the CACM's "Programming Pearls" column, but not be able to get > hold of the program !! If you have the requisite 1) money and 2) source license, you need suffer no more; the new "awk" is, indeed, the one used by Bentley and does have functions, along with some new built-in functions and variables. It also supports 8-bit characters. It is *not*, however, 100% compatible with the old "awk". The S5R3.1 release notes list these incompatibilites. The new "awk" is "nawk" in S5R3.1; "awk" is still the old one. According to the release notes, in the next major S5 release the new "awk" will become "awk" and the old one will be renamed "oawk". > Also, Guy Harris noted about a week ago that the version of awk > supplied with S5R2 was "faster" than the 4.3 BSD version, so I went > ahead and compiled the S5 version on our BSD VAX. You may want to apply some of the bug fixes made in the 4.3BSD version to the S5R2 version. I'm reposting a list of bug fixes to the S5R2 version to "comp.bugs.sys5"; they include those fixes, as well as some null-pointer-dereference fixes. I think the new "awk" has all those fixes. > The S5 version turned out to be almost 3 times faster for a simple program > (print $1) and 30% faster for a long (200 line) awk script. Unfortunately, the new "awk" loses some of that speed advantage, at least in the tests I ran. Guy Harris {ihnp4, decvax, seismo, decwrl, ...}!sun!guy guy@sun.com ----------------------------- From: Patrick Pesch Subject: Re: Info about ratfor and/or efl Date: 26 Sep 87 01:34:57 GMT Keywords: f77 preprocessors structure To: info-unix@SEM.BRL.MIL In article <12@vanvleck.MATH.WISC.EDU> kailhofe@vanvleck.MATH.WISC.EDU (kailhofe) writes: >I'm looking for information about the two fortran preprocessors efl(1) and >ratfor(1), preferring info about efl. All I have are the man pages, >which, as always, could use a bit of fleshing out. > Sorry, no efl info... BUT I think ratfor is used/discussed in the "Software Tools" by Kernighan and Plaugher(sp?). If I remember correctly, a ratfor pre-processor implementation is given in the book. Pat Pesch NCR Corporation ----------------------------- From: Stefan Olausson Subject: Eyacc Date: 24 Sep 87 13:04:32 GMT To: info-unix@brl-sem.arpa Does anyone have documentation (or pointer to documentation) for "eyacc"? Eyacc is an enhanced version of yacc that comes with Berkeley's Pascal distribution, which supports error recovery etc. Please respond by e-mail. Stefan Olausson ========================================================================= UUCP: {seismo,philabs,decvax}!mcvax!enea!chalmers!olausson CSnet: olausson@chalmers.csnet Mail: Informations-behandling, Chalmers University, S-412 96 Goteborg, SWEDEN ----------------------------- From: Andy Clews Subject: Invisible Ascii with VI Date: 23 Sep 87 13:30:37 GMT To: info-unix@SEM.BRL.MIL Does anyone out there know if it's possible to insert/search/replace non-printing ascii characters? Specifically: I sometimes get files from a net source that have an "extra" CR tacked on the end of the line, and VI shows them as ^M on the screen. For various reasons these are a pain. I'd like to be able to get rid of them with, say, a global substitute. I've tried 1,$s/^V^M$// but this does not seem to do what I want. I'd appreciate email replies if any. Andy Clews andy@uk.ac.sussex.unx1 ----------------------------- From: John Sperbeck Subject: Exclusive file creation and NFS. Date: 25 Sep 87 15:33:28 GMT Keywords: NFS exclusive file creation To: info-unix@SEM.BRL.MIL Does NFS support exclusive file creation? That is, does the call: fd = open("t.tmp", O_CREAT | O_EXCL) always fail if t.tmp already exists on an NFS file system? I have heard from two people that it usually works, but there is a race condition that can occur if two client processes try to do the open on the same non-existent file. When the server does the work of creating the file, both client processes receive a message that the file was created successfully. Can this really happen? I ask because we have /usr/spool/mail mounted on many machines and I am under the impression that the programs that read and deliver mail use the exclusive file creation for locking. By the way, the machines I am dealing with right now are Ultrix 2.0, but I would be interested in whether or not other implementations of NFS have the same problem. -- John Sperbeck ...{hplabs|decwrl|amdcad|qantel|oliveb}!intelca!mipos3!sperbeck ----------------------------- From: Rob DeMarco Subject: csh environment variables and special commands... Date: 26 Sep 87 15:14:28 GMT To: info-unix@brl-sem.arpa Hello, I am looking for some CSH environmental variables and SPECIAL csh commands. That is, i'm not interested in "if" commands, but special commands like "rehash" and environmental variables like "ignoreof". Thanks, [> Rd -- North Coast Computer Resources(ncoast) - 216-781-6201 (or 781-6202) UUCP:decvax!mandrill!ncoast!robertd Sysop: NEODG (login "sbbs") ----------------------------- From: Thomas Mellman Subject: variables set in sh while loop don't get set? Date: 24 Sep 87 15:22:46 GMT Keywords: sh, ksh scripts, loops To: info-unix@SEM.BRL.MIL When I set a variable in a while loop in a sh script, in some scripts it doesn't seem to get set. For example, a common thing is to do a pipeline | while read proc do and then to test for some condition in a case statement. Twice now, I have written scripts where a variable set in the case loop (or anywhere in the while loop, for that matter) won't have been changed from its initial value after the while loop terminates. Can anyone explain to me why this is so. For example, I have a little script that will log me off of my 630 terminal by nohup-ing the login process - as long as only insignificant processes are still running: # # This script will shutdown my account # PATH=$HOME/bin:/bin:/usr/bin:/usr/lbin options="xd" arglist="" optlist="" set -- `getopt $options $*` if [ $? != 0 ]; then echo "Usage: $0 [$options]" exit -1 fi for arg in $*; do case $arg in --) shift break ;; -d) execute=2 export execute shift ;; -x) execute=1 export execute shift ;; esac done ps -u$LOGNAME | awk '{print $4}' | while read proc do execute=3 case $proc in sysmon | awk | ksh | ps | layers | COMMAND) ;; "") echo '\007\007' defunct process? execute=0 export execute ;; *) echo '\007\007' $proc still running execute=0 export execute ;; esac done echo execute is $execute if [ $? -eq 0 ]; then if [ ${execute:-0} -eq 1 ]; then kill -1 ${fatherpid:-"fatherpid not set"} else echo ${fatherpid:-"fatherpid not set"} fi fi But, at the echo, it can be seen that execute never got set during the while loop. It's driving me crazy. Any help would be appreciated. ----------------------------- From: Thomas Mellman Subject: ps -s /dev/null kills system? Date: 24 Sep 87 15:54:19 GMT Keywords: system V, examining core files To: info-unix@brl-sem.arpa ps(1) says, about the -s option: "Use the file *swapdev* in place of -/dev/swap-. This is useful when examining a *corefile*; a *swapdev* of -/dev/null- will cause the user block to be zeroed out." Is this for real? What is the *user block*, anyway? Presumably, if it's a disk block, you have to be the super-user to have write access to disk, but even then... ----------------------------- End of INFO-UNIX Digest ***********************