Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!purdue!haven!adm!xadmx!postmaster@"[unknown]" From: postmaster@[unknown] (SMTP MAILER) Newsgroups: comp.unix.questions Subject: Mail Delivery Problem Message-ID: <18902@adm.BRL.MIL> Date: 31 Mar 89 22:15:05 GMT Sender: news@adm.BRL.MIL Lines: 1182 ----Reason for mail failure follows---- Sending mail to host drum : Unrecognized host name or address. ----Transcript of message follows---- Date: 31 Mar 89 14:26:00 WET From: info-unix@BRL.MIL Subject: INFO-UNIX Digest V7#036 To: "baynes" Return-Path: Received: from SEM.BRL.MIL by nusc.arpa with SMTP ; Fri, 31 Mar 89 14:25:08 EST Received: from SEM.BRL.MIL by SEM.brl.MIL id ab08972; 31 Mar 89 3:00 EST Received: from sem.brl.mil by SEM.BRL.MIL id aa08968; 31 Mar 89 2:46 EST Date: Fri, 31 Mar 89 02:46:32 EST From: The Moderator (Mike Muuss) To: INFO-UNIX@BRL.MIL Reply-To: INFO-UNIX@BRL.MIL Subject: INFO-UNIX Digest V7#036 Message-ID: <8903310246.aa08968@SEM.BRL.MIL> INFO-UNIX Digest Fri, 31 Mar 1989 V7#036 Today's Topics: What does "munch" in C++ and/or Concurrent C do? Re: UNIX prompts (-ksh) Return from exit? Re: Some csh how-to, please Need csh alias to match patterns in history Re: Future at Berzerkeley Re: slicing the date Re: Future at Berzerkeley Re: VMS vs. UNIX s/w development tools - query Re: Some csh how-to, please Re: Re^2: Getting UNIX prompt to display current directory Driver close system call Re: csh: still trying to read file Re: VMS vs. UNIX s/w development tools - query Re: Future at Berzerkeley Re: slicing the date csh - now how about pseudo-arrays? Re: Need csh alias to match patterns in history Re: slicing the date uucp machine aliasing nroff page length Re: slicing the date Help with FTP? Re: C bug causes double fault Re: Future at Berzerkeley Japanese language and computer, help requested Re: Future at Berzerkeley Re: Getting UNIX prompt ; easy questions Re: Driver close system call Re: Implications of large memory systems Re: Need csh alias to match patterns in history Re: slicing the date Re: Future at Berzerkeley Re: Some csh how-to, please another way (was Re: slicing the date) ----------------------------------------------------------------- From: Scott Hammond Subject: What does "munch" in C++ and/or Concurrent C do? Date: 28 Mar 89 18:01:59 GMT Followup-To: comp.lang.c++ To: info-unix@sem.brl.mil Can someone enlighten me as to what "munch" does for C++ and/or for Concurrent C? I know it performs some sort of manipulations on the object file but that's all I know. It also seems to add a lot to compile time and I want to know if I can do without it. I'm actually specifically interested in Concurrent C, but since it uses nearly the same parser as C++, I gather that "munch" is doing the same thing for both. Thanks for any information. My apologies if this is a naive question. -- Scott Hammond, R & D Associates, Marina del Rey, CA (213) 822-1715 : {ksuvax1,zardoz}!rdahp!scott : rdahp!scott@ksuvax1.cis.ksu.edu : scott@harris.cis.ksu.edu ----------------------------- From: "Bruce T. Harvey" Subject: Re: UNIX prompts (-ksh) Date: 28 Mar 89 17:30:48 GMT To: info-unix@sem.brl.mil In article <4549@vpk4.UUCP>, hjespers@vpk4.UUCP (Hans Jespersen) writes: > In article <2391@buengc.BU.EDU> bph@buengc.bu.edu (Blair P. Houghton) writes: > >In article <11080@well.UUCP> tneff@well.UUCP (Tom Neff) writes: > >> export PS1='$PWD> ' > Absolutely correct. I think Tom ment > export PS1=`$PWD> ` Nope. Tom meant '$PWD' when he said it. The reason for the single quotes is so that the shell where the definition is being made does not interpret $PWD. If it did, the current working directory of _that_shell_ would _always_ display. By quoting it, four characters ($, P, W, and D) are exported and not interpreted by ksh _until_displayed_. I don't think sh- does any interpretation of displayed variables in this manner, which is why it doesn't work this way using sh-. If you use back-quotes, as in above, you end up trying to execute a directory. (" ... Hmm. I wonder if that's why directories must be x-ecutable to be used ...") -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Bruce T. Harvey (B-}> | ... cp1!sarin!wb3ffv!idsssd!idssup!bruce (Title depends on day) | ... ctnews!idsssd!idssup!bruce (301) 584-1960 | Convergent Route Distribution Sys. - Hunt Valley, MD ----------------------------- From: "Frank I. Reiter" Subject: Return from exit? Date: 28 Mar 89 16:28:26 GMT Keywords: curses exit signal To: info-unix@sem.brl.mil I have a small monitoring program which uses curses to update it's display every 15 seconds. It has been in use and worked fine for a couple of years with one exception: It appears that sometimes it returns from an exit! SIGINT is set to be handled by this routine: void quit() { signalson(NO); /* Leave us alone while we cleanup */ close(sysfd); clear(); refresh(); endwin(); detachutil(); /* Log us out of shared memory */ puts("watch: terminated"); exit(0); } signalson(flag) /* Enable/disable sig handling routines */ int flag; { int i; static int lastflag=YES; if(flag==YES) { signal(SIGINT,quit); signal(SIGTERM,quit); } else { signal(SIGINT,SIG_IGN); signal(SIGTERM,SIG_IGN); } i=lastflag; lastflag=flag; return(i); } When the interrupt key is sent to the process sometimes the screen will clear, the exit message will be printed, and then a few seconds later the screen is refreshed with the latest statistics. After that happens the interrupt key is ignored and we must use SIGQUIT to kill the process. Can anybody suggest why this might be happening? -- _____________________________________________________________________________ Frank I. Reiter UUCP: {uunet,ubc-cs}!van-bc!rsoft!frank Reiter Software Inc. frank@rsoft.bc.ca, a2@mindlink.UUCP Langley, British Columbia BBS: Mind Link @ (604)533-2312, login as Guest ----------------------------- From: Gregory Bollella Subject: Re: Some csh how-to, please Date: 29 Mar 89 12:58:48 GMT Sender: news@thorin.cs.unc.edu Keywords: csh C-shell shell programming unix read To: info-unix@sem.brl.mil In article <7467@thorin.cs.unc.edu> white@white.cs.unc.edu (Brian T. White) writes: >In article <2127@pikes.Colorado.EDU>, pklammer@pikes.Colorado.EDU (Peter Klammer) writes: >> Could someone please tell me how to read a file line at a time >> into the C shell? Better yet, can you refer me to any good C-shell >> text. > >Try "The UNIX C Shell Field Guide" by Anderson and Anderson. To read >a file one line at a time, try > >foreach p ( "`cat file`" ) > set line=`echo $p` > (commands acting on $line) >end I tried the above on a large file and it did not work. csh reported Too many words from ``. The file was 118583 bytes. Does anybody have any ideas on how to get around this limit and still get one line at a time? Gregory Bolella bollella@cs.unc.edu Department of Computer Science CB# 3175 Sitterson Hall University of North Carolina Chapel Hill, NC 27599-3175 ----------------------------- From: Naim Abdullah Subject: Need csh alias to match patterns in history Date: 24 Mar 89 02:46:14 GMT To: info-unix@sem.brl.mil A friend of mine wants a csh alias that will print all lines in the history list that match a pattern if one is supplied, otherwise just print the history. He asked me and I suggested the following csh alias: alias h ' \ if ( "X\!*" == "X" ) then \ history \ else \ history | grep \!* \ endif' For some reason, this does not work. Can some kind soul explain why and supply a working answer to this problem ? Thanks. Naim Abdullah Dept. of EECS, Northwestern University Internet: naim@eecs.nwu.edu Uucp: {oddjob, chinet, att}!nucsrl!naim ----------------------------- From: Doug Gwyn Subject: Re: Future at Berzerkeley Date: 29 Mar 89 17:07:04 GMT To: info-unix@sem.brl.mil In article <3095@stpstn.UUCP> aad@stepstsone.com writes: -In article <4572@pt.cs.cmu.edu> jps@cat.cmu.edu (James Salsman) writes: ->a long time, but SysV will always be more stable. -Let's see: svr1,svr2,svr3,svr3.2,svr4... Seems like they come out with -versions a lot more often. He said "stable", not "stagnant". ----------------------------- From: "R. E. Mackey" Subject: Re: slicing the date Date: 29 Mar 89 15:25:14 GMT To: info-unix@sem.brl.mil In article <216000010@s.cs.uiuc.edu>, carroll@s.cs.uiuc.edu writes: > [......] The question is, how can I cut up the output of > date to put the hours, minutes, and seconds into seperate variables? Thanks! The line eval `date '+hr=%H;min=%M;sec=%S'` will set hr to the current hour, min to the current minute, and sec to the current second. You're welcome. -- Ron Mackey - AT&T Bell Laboratories UUCP: att!ihlpb!mackey 2000 N. Naperville Road WORK: (312) 979-2140 Naperville, Illinois 60566 OFFICE: IH 2C-423 ----------------------------- From: Chuck Karish Subject: Re: Future at Berzerkeley Date: 29 Mar 89 16:02:17 GMT Sender: USENET News System Keywords: Leading Edge == Bottom Line To: info-unix@sem.brl.mil In article <3095@stpstn.UUCP> aad@stepstsone.com wrote: >In article <4572@pt.cs.cmu.edu> jps@cat.cmu.edu (James Salsman) writes: >>a long time, but SysV will always be more stable. >Let's see: svr1,svr2,svr3,svr3.2,svr4... Seems like they come out with >versions a lot more often. Yes, but it's a stable development environment because AT&T specifies what backward compatibility they'll provide in new versions. This may be carried a bit too far, to the extent that they're reluctant to fix misfeatures for fear of breaking customers' work-arounds. Any developer who's happy with the old functionality can ship products for the newer version of the OS without having to port to it. Chuck Karish hplabs!hpda!mindcrf!karish (415) 493-7277 karish@forel.stanford.edu ----------------------------- From: Todd Merriman Subject: Re: VMS vs. UNIX s/w development tools - query Date: 29 Mar 89 12:47:08 GMT Keywords: VMS UNIX tools To: info-unix@sem.brl.mil In article <401@bionov.UUCP> rayl@bionov.UUCP (Ray Lillard) writes: >I am a software consultant working with a client company with lots >of code developed for RSX-11. We are casting about for a new >system (hardware and software) to replace the existing imbedded >target system......... I do cross development between Unix, VMS, and MSDOS; and I have a few opinions on the subject. I consider Unix to be a superior *development* environment, but VMS to be a superior *production* environment. The variety of tools and power of the Shell make Unix an ideal choice for technical types and programmers. The sophistication of VMS, combined with its excellent documentation and reliability make it the clear winner in production. ...!gatech!stiatl!todd Todd Merriman * 404-377-TOFU * Atlanta, GA Note: I have no idea what my employer's views on the subject are. ----------------------------- From: Seth Robertson Subject: Re: Some csh how-to, please Date: 29 Mar 89 18:03:54 GMT Keywords: csh C-shell shell programming unix read To: info-unix@sem.brl.mil In article <7485@thorin.cs.unc.edu> bollella@ra.UUCP (Gregory Bollella) writes: >>In article <2127@pikes.Colorado.EDU>, pklammer@pikes.Colorado.EDU (Peter Klammer) writes: >>> Could someone please tell me how to read a file line at a time >>> into the C shell? Better yet, can you refer me to any good C-shell >>> text. >> [Method deleted] >I tried the above on a large file and it did not work. csh reported > Too many words from ``. >The file was 118583 bytes. Does anybody have any ideas on how to >get around this limit and still get one line at a time? I don't know how to do it in csh, but the method that works for (k)sh is: #! /bin/sh exec 3<&0 < /usr/local/lib/ctrsuns while read test do echo $test rsh $test -n "$1" done exec 0<&3 3<&- This gives the error missing name for redirect. You might be able to fool around with it and get it working... -Seth seth@ctr.columbia.edu ----------------------------- From: Michael Morrell Subject: Re: Re^2: Getting UNIX prompt to display current directory Date: 28 Mar 89 21:09:29 GMT To: info-unix@sem.brl.mil / hpsal2:comp.unix.questions / lbn@ksuvax1.cis.ksu.edu (Lars Bo Nielsen) / 7:52 am Mar 27, 1989 / You are right. Here are the modified (quicker) versions: 2) Only name of current dir: alias cd 'cd \!*; set foo=$cwd; set prompt=$foo:t"% "' 3) Include name of machine in prompt alias cd 'cd \!*; set prompt=$host":"$cwd"% "' or alias cd 'cd \!*; set foo=$cwd; set prompt=$host":"$foo:t"% "' ---------- I don't think you need the extra variable foo. "set prompt=$cwd:t" works just fine. Michael ----------------------------- From: "Wm. Tseng" Subject: Driver close system call Date: 29 Mar 89 19:59:35 GMT Sender: news@ncr-sd.sandiego.ncr.com To: info-unix@sem.brl.mil I have a question about driver close routine. Thanks in advance for any answer. If a process exits or terminates before it closes a driver, does kernel close the driver for that process ? In user process, several processes can open the same driver. Are driver open and close routines executed as many time as process calls it ? Does kernel control the close system call, so close is only executed once (when last process closes it)? When system crash, does kernel close all the opened file and driver or just leave them open ? william williamt@babel.sandiego.ncr.com ----------------------------- From: Maarten Litmaath Subject: Re: csh: still trying to read file Date: 29 Mar 89 15:39:44 GMT Keywords: csh C shell read programming unix To: info-unix@sem.brl.mil pklammer@pikes.Colorado.EDU (Peter Klammer) writes: \... what I want is to read a file a line at a time in the \midst of a csh script that is in dialogue; that is, redirecting \stdin is not acceptable. You don't want csh scripts, if you want to avoid problems. Use sh scripts: i=0 exec 3< foo while read bar <&3 do i=`expr $i + 1` echo "line $i: $bar" done \... Another csh mystery (at least until I get my book, maybe): if I prepare \a csh script file (with the mandatory "#" on line 1) and chmod +x it, \and then invoke it from a subdirectory, it is run from my HOME directory! ... Probably there's a cd command in your .cshrc. Furthermore, try if your kernel recognizes `#!/bin/csh -f' on line 1. -- Modeless editors and strong typing: |Maarten Litmaath @ VU Amsterdam: both for people with weak memories. |maart@cs.vu.nl, mcvax!botter!maart ----------------------------- From: Piet van Oostrum Subject: Re: VMS vs. UNIX s/w development tools - query Date: 29 Mar 89 15:46:34 GMT Sender: piet@ruuinf.cs.ruu.nl Followup-To: comp.software-eng Keywords: UNIX tools To: info-unix@sem.brl.mil In article <401@bionov.UUCP>, rayl@bionov (Ray Lillard) writes: I have extolled the virtues of the UNIX software `environment and have been told that ..... My question is: What do people generally consider to be a ``software development environment''. Is it something like the collection of tools that Unix has or is there much more in it, for example tools for the specification and design parts of the software life cycle. What experience do you have with additional tools in the Unix environment? -- Piet van Oostrum, Dept of Computer Science, University of Utrecht Padualaan 14, P.O. Box 80.089, 3508 TB Utrecht, The Netherlands Telephone: +31-30-531806. piet@cs.ruu.nl (mcvax!hp4nl!ruuinf!piet) ----------------------------- From: Guy Harris Subject: Re: Future at Berzerkeley Date: 29 Mar 89 20:25:38 GMT Keywords: Leading Edge == Bottom Line To: info-unix@sem.brl.mil > When bsd and System V are merged, will the holders of 32V licenses >still be able to get source to newer bsd versions? Or will everyone wishing >to get source have to go out and get a SV.4 licence? I think you misunderstand what "when bsd and System V are merged" means. It means that S5R4 will be picking up BSD stuff, not that S5 and BSD will become one and the same thing. As far as I know, Berkeley isn't going to convert to S5R4, so presumably a 32V licence should continue to be sufficient. ----------------------------- From: Dave Turner Subject: Re: slicing the date Date: 29 Mar 89 18:38:27 GMT To: info-unix@sem.brl.mil In article <216000010@s.cs.uiuc.edu> carroll@s.cs.uiuc.edu writes: >Allright, I can't figure this out. I'm trying to put the current system >time in my prompt. Under SysV, I can use 'cut' to pick out the seperate >hours, minutes, and seconds from 'date'. For non-SysV types, 'cut' lets >you pick out fields using either delimiter characters or absolute column You might try: date "+%H %M %S" | read HH MM SS echo $HH echo $MM echo $SS -- Dave Turner 415/542-1299 {att,bellcore,sun,ames,pyramid}!pacbell!dmt ----------------------------- From: Peter Klammer Subject: csh - now how about pseudo-arrays? Date: 29 Mar 89 22:53:01 GMT Keywords: csh C shell UNIX programming arrays subscripts To: info-unix@sem.brl.mil OK, I'm learning, I'm getting better at this... Yes, our .cshrc ended with 'cd', apparently to initiate the pretty prompt via the cd alias. So if I change that to 'cd .' instead, we get the pretty prompt without reverting to HOME in every subshell. Now I would like to know if csh can be warped into array-like subscripting. I am trying to write a table-driven menu in a csh script, so that I can loop through cases instead of adding another explicit paragraph of case code whenever we expand the menu. Here's the rough outline: ------------------------------------------------------------ #csh # menu4 @ n = 0 @ n += 1; set entry$n = (EXIT EXIT) @ n += 1; set entry$n = (ADDUSER ADD a new user account) @ n += 1; set entry$n = (MODUSER MODIFY a user account) @ n += 1; set entry$n = (DELUSER DELETE a user account) @ n += 1; set entry$n = (SHOWUSER SHOW a user account) @ n += 1; set entry$n = (ADDCLASS Add a new CLASS) @ n += 1; set entry$n = (ADDGROUP Add a new GROUP) @ n += 1; set entry$n = (DELGROUP CANCEL a group) @ n += 1; set entry$n = (INSTRUCT INSTRUCTION) #This, for specific case 3, works: set command = $entry3[1] set prompt = "${entry3[2-]}" echo The command $command will $prompt #This, for general case i, fails, with message 'Variable Syntax': @ i = 1 while( $i <= $n ) set command = ${entry$i[1]} set prompt = "${entry$i[2-]}" echo The command $command will $prompt @ i += 1 end ------------------------------------------ That's as close as I have been able to get in a couple days trying without the Anderson book. I'd be happy to consider another approach, but for some other reasons we're kinda bound to doing it in C-shell. (I'd rather do it in awk, but our awk here can't import the arv command line, so it's (snicker) awkward.) ----------------------------- From: Marco Zagha Subject: Re: Need csh alias to match patterns in history Date: 29 Mar 89 20:36:59 GMT To: info-unix@sem.brl.mil In article <3680046@eecs.nwu.edu>, naim@eecs.nwu.edu (Naim Abdullah) writes: > > alias h ' \ > if ( "X\!*" == "X" ) then \ > history \ > else \ > history | grep \!* \ > endif' > > For some reason, this does not work. Can some kind soul explain why > and supply a working answer to this problem ? I think this will do the trick: alias hh 'history | grep "\!*"' This works because 'grep ""' is not the same as 'grep'. Your alias produces strange behavior. Neither the "if" or the "else" condition is ever executed. I don't know why. == Marco (marcoz@cs.cmu.edu) -- ----------------------------- From: Dennis Pelton CSM Contractor x8876 Subject: Re: slicing the date Date: 29 Mar 89 15:57:14 GMT To: info-unix@sem.brl.mil In article <216000010@s.cs.uiuc.edu>, carroll@s.cs.uiuc.edu writes: > > Allright, I can't figure this out. I'm trying to put the current system > time in my prompt. Under SysV, I can use 'cut' to pick out the seperate > hours, minutes, and seconds from 'date'. For non-SysV types, 'cut' lets > you pick out fields using either delimiter characters or absolute column > counts. I found a way to do this using Ksh (with the pattern matchers), > but it is very ugly. The question is, how can I cut up the output of > date to put the hours, minutes, and seconds into seperate variables? Thanks! > > Alan M. Carroll "And then you say, > carroll@s.cs.uiuc.edu We have the Moon, so now the Stars..." - YES > CS Grad / U of Ill @ Urbana ...{ucbvax,pur-ee,convex}!s.cs.uiuc.edu!carroll OK, here's one way: hrs=`date +'%H'` min=`date +'%M'` sec=`date +'%S'` Andhere's another: date +'%H %M %S' >now read hrs min sec Subject: uucp machine aliasing Date: 29 Mar 89 04:25:35 GMT Keywords: uucp HDB To: info-unix@sem.brl.mil Apparently BSD 4.3 uucp has a file of machine aliases (L.aliases). Does HDB uucp have any simular facility? If not, any idea how to do it or how one may want to represent it in the Systems or Permissions files? -- Jonathan Hanna yonatan@dybbuk.uucp ...uunet!attcan!utzoo!yunexus!dybbuk!yonatan ----------------------------- From: Joe Wells Subject: nroff page length Date: 30 Mar 89 01:30:18 GMT Followup-To: comp.unix.questions To: info-unix@sem.brl.mil Dear Wise and Knowlegeable People (with Source Access), It would be very useful to be able to tell nroff that its page length is infinite (or just very very large). This would probably involve modifying one of the files in /usr/lib/term that come with nroff. However, I haven't been able to find documentation on the format of these files. Does anyone know enough about this to help? I would consider editing the binary an adequate solution. -- Joe Wells INTERNET: jbw%bucsf.bu.edu@bu-it.bu.edu IP: [128.197.10.201] UUCP: ...!harvard!bu-cs!bucsf!jbw ----------------------------- From: Doug Gwyn Subject: Re: slicing the date Date: 30 Mar 89 02:57:10 GMT To: info-unix@sem.brl.mil In article <431@ncsc1.AT&T.NCSC> dgp@ncsc1.AT&T.NCSC (Dennis Pelton CSM Contractor x8876) writes: >Why is it that the following does NOT work? > date +'%H %M %S' | read hrs min sec Because the shell built-in "read" runs in a subshell due to the pipeline. You're setting the variables in a subshell, all right, but that isn't what you wanted. ----------------------------- From: Paul Houtz Subject: Help with FTP? Date: 29 Mar 89 22:31:47 GMT To: info-unix@sem.brl.mil I am just starting to use ftp, and I have a few questions: 1. How does it work? I am currently trying to get into some anonymous sites, and am experiencing the following difficulties: a: I will be connected to the site, some line noise or something knocks me off, and then I can no longer connect to that site. I get the following error: ftp: connect: No route to host b: Some sites I attempt connection to give me the error ftp: (site address): Unknown host Once I have been bounced off, I seem to get the first message above no matter what machine I attempt to open. Is it possible that getting bounced screws things up on my machine for some undetermined amount of time (say--to the next boot up)? 2. Is there another method of specifying the address of machine, rather than the Arpanet address (i.e., machine.university.edu)? In some of the site lists, the give a bunch of numbers with periods (302.333.23.2349) that appear to be some kind of address. What is this? Thanks for your time and help! Paul Houtz HP Technology Access Center 10670 N. Tantau Avenue Cupertino, Ca 95014 (408) 725-3864 hplabs!hpda!hpsemc!gph gph%hpsemc@hplabs.HP.COM ----------------------------- From: "M.R.Murphy" Subject: Re: C bug causes double fault Date: 30 Mar 89 02:16:17 GMT To: info-unix@sem.brl.mil In article <660@micropen> dave@micropen (David F. Carlson) writes: ! !The truth is that Microport early versions had the potential to corrupt the !kernel stack on floating point exceptions, which is what this should be. !This was supposedly fixed several versions ago and I never had saw this !again. (It was a showstopper though for a multi-user development machine: !too insecure to use.) ! The program as written doesn't double panic uPort V/AT 2.2.2 with 80287, or 2.3, 2.4 without 80287. With other combinations, your mileage may vary:-) Which, incidentally, points out one of the large problems in getting the bugs out of an operating system which is expected to run in who-knows-how many hardware configurations (CPU,motherboard,disk controller,disk drive, and on and on...). The hardware used in this exhaustive test was no-name clone. -- Mike Murphy Sceard Systems, Inc. 544 South Pacific St. San Marcos, CA 92069 mrm@sceard.UUCP {hp-sdd,nosc,ucsd}!sceard!mrm +1 619 471 0655 ----------------------------- From: Larry Taborek Subject: Re: Future at Berzerkeley Date: 30 Mar 89 12:54:35 GMT To: info-unix@sem.brl.mil My question is, since Berkeley has been paying licsence fees to AT&T for portions of their code, will AT&T pay Berkeley for portions of their code included in SVR4? If not, then will AT&T price SVR4 so that users are not paying for portions of code that AT&T did not buy or develop? Anyone from AT&T or Berkeley wish to comment? -- Larry Taborek ..!uunet!grebyn!macom1!larry Centel Federal Systems larry@macom1.UUCP 11400 Commerce Park Drive Reston, VA 22091-1506 703-758-7000 ----------------------------- From: Francois Wauthier Subject: Japanese language and computer, help requested Date: 29 Mar 89 13:15:34 GMT Keywords: japanese, unix, kana, kanji To: info-unix@sem.brl.mil Hello World, We are planning to port one of our product to .... Japanese. So we need informations about Japanese Unix, Japanese Keyboards, Japanese fonts, Kanji-Kana coding, ... and all these kind of subjects. So if you have informations or references to informations related to these subjects, please email at UUCP.mcvax!prlb2!sunbim!fw or fw@sunbim.be By the way, I am learning Japanese, so if you have any reference to a good Kanji dictionary, it would be helpfull Thanks in advance Sayoonara Francois Wautier --------------------------------------------------------------- Francois Wautier | I like trafic lights (ter) BIM | But only when they're green Mail: fw@sunbim.be | (from a Monty Python's song) UUCP.mcvax!prlb2!sunbim!fw --------------------------------------------------------------- ----------------------------- From: Doug Gwyn Subject: Re: Future at Berzerkeley Date: 30 Mar 89 17:23:16 GMT To: info-unix@sem.brl.mil In article <4814@macom1.UUCP> larry@macom1.UUCP (Larry Taborek) writes: >My question is, since Berkeley has been paying licsence fees to >AT&T for portions of their code, will AT&T pay Berkeley for >portions of their code included in SVR4? If not, then will AT&T >price SVR4 so that users are not paying for portions of code that >AT&T did not buy or develop? I'm sure that AT&T's lawyers have made the necessary contractural arrangements with all providers of software that AT&T will be distributing. I'm also sure that they're not going to disclose the details of any such arrangement just because you're curious about it. The pricing for SVR4 has not yet been announced. The implied accusation of theft that lies behind your questions is, to the best of my knowledge (and I do have some inside knowledge) unfounded and demands an apology. ----------------------------- From: David J Bell Subject: Re: Getting UNIX prompt ; easy questions Date: 29 Mar 89 20:45:53 GMT To: info-unix@sem.brl.mil >< P.S.3: To those people who continue flaming about frequent and repeated >< postings of elementary questions in this news group: Short of starting >< a comp.unix.questions.elementary, if the answer to the >The group comp.unix.questions was started *specifically* because there >were a lot of questions being asked in comp.unix.wizards that weren't >wizard-level questions - remember that new users need to learn things, >too, and (after we've reminded them to RTFM) there are a lot of things >that are not obvious to an inexperienced user - even things that are in >the manuals often need a lot of thought, especially because > > some of the manual pages aren't all that well-written, and > sometimes you have to know where to look before you look there. ******* SOMETIMES???? ******* OFTEN the only way to find something is to have already RT<#!&*@$>FM, and virtually MEMORIZED it... Thank the UNIX gods for online man! Now if man were more like Hypertext... Dave dbell@cup.portal.com ----------------------------- From: Roger Collins Subject: Re: Driver close system call Date: 30 Mar 89 14:04:34 GMT To: info-unix@sem.brl.mil In article <1193@ncr-sd.SanDiego.NCR.COM> williamt@babel.SanDiego.NCR.COM (Wm. Tseng) writes: > > If a process exits or terminates before it closes a driver, does kernel close > the driver for that process ? Yes. All files (includes device nodes, regular files, pipes, etc.) are closed. > In user process, several processes can open the > same driver. Are driver open and close routines executed as many time as process > calls it ? Yes and no. Yes for opens, no for closes. All opens get to the driver. Only the last close gets to the driver. > Does kernel control the close system call, so close is only executed > once (when last process closes it)? You got it, but not to confuse anyone else: You meant to say, "...so *the driver's* close is only executed once..." The kernel does execute close logic for every close system call (decrement the reference counter for example), it just doesn't call the driver's close routine until the reference counter is zero. > When system crash, does kernel close all > the opened file and driver or just leave them open ? No. The kernel leaves everything basically as is (except for syncing the drives). This is probably vendor specific; I know only about the NCR TOWER. -- Roger Collins NCR - E&M Columbia rogerc@ncrcae.Columbia.NCR.COM ----------------------------- From: Roger Collins Subject: Re: Driver close system call Date: 30 Mar 89 14:13:00 GMT To: info-unix@sem.brl.mil In article <4337@ncrcae.Columbia.NCR.COM> I write: > > Yes. All files (includes device nodes, regular files, pipes, etc.) are > closed. Oops. The question was does the kernel close the *driver* when a process terminates? The answer is clear if you read the rest of my posting, but just in case: All the files opened by the process are closed. If the file that is the device node for the driver is closed and the reference count is zero (last close for the device), then... yes, the driver gets closed when the process terminates. Otherwise no. ----------------------------- From: Dan Schlitt Subject: Re: Implications of large memory systems Date: 30 Mar 89 18:53:43 GMT To: info-unix@sem.brl.mil In article <68@crdgw1.crd.ge.com> barnett@crdgw1.crd.ge.com (Bruce Barnett) writes: :Another thing to consider about 1 GByte Memory workstations, is that :when the systems have more potential, the creative researcher finds :a way to use that power. They thought 64K was enough. Then 256K was :enough.... : At the first usenix conference I attended there were questions from the audience about how more than 1 Mbyte could be best put to use on a unix system. They got answers like "use it as a fast disk" :-) Make the memory available and people will find lots of good ways to use it. And not always the ones that you think of beforehand. -- Dan Schlitt Manager, Science Division Computer Facility dan@ccnysci City College of New York dan@ccnysci.bitnet New York, NY 10031 (212)690-6868 ----------------------------- From: Maarten Litmaath Subject: Re: Need csh alias to match patterns in history Date: 30 Mar 89 05:38:28 GMT Keywords: sed, sed, sed, csh, Bill 'bug' Joy, Bugs Bunny To: info-unix@sem.brl.mil naim@eecs.nwu.edu (Naim Abdullah) writes: \A friend of mine wants a csh alias that will print all lines in \the history list that match a pattern if one is supplied, otherwise \just print the history. \He asked me and I suggested the following csh alias: \alias h ' \ \if ( "X\!*" == "X" ) then \ \history \ \else \ \history | grep \!* \ \endif' \For some reason, this does not work. Welcome to csh! Generally it doesn't like foreach, while, etc. in aliases. Solutions: alias h 'set q=(\!*); eval h$#q' alias h0 history alias h1 '(history | sed -n -e \$q -e /"$q"/p)' or alias h 'if ("\!*" == "") set status=1 &&'\ 'history | sed -n -e \$q -e /"\!*"/p ||'\ history or alias h 'if (\!* == "") set status=1 &&'\ 'history | sed -n -e \$q -e /\!*/p ||'\ history The final version causes csh to complain if multiple arguments are given. BTW, it's obvious why I used sed instead of grep. -- Modeless editors and strong typing: |Maarten Litmaath @ VU Amsterdam: both for people with weak memories. |maart@cs.vu.nl, mcvax!botter!maart ----------------------------- From: carroll@s.cs.uiuc.edu Subject: Re: slicing the date Date: 29 Mar 89 17:24:00 GMT Nf-ID: #R:s.cs.uiuc.edu:216000010:s.cs.uiuc.edu:216000011:000:747 Nf-From: s.cs.uiuc.edu!carroll Mar 29 11:24:00 1989 To: info-unix@sem.brl.mil RE: Slicing the date I've gotten several replies, and I need to clarify something - I have tried the '+' notation on the 'date' command to get it to format its output. The OS running here (I believe 4.3BSD) does not accept that. Any argument is assumed to be an attempt to *set* the date, not format it, according to the man pages and actual use. I was moved (against my will) from SysV to BSD, and this is something that I lost - I used to use the date '+%H:%M:%S' format, and it doesn't work anymore. That's why I have to slice the normal output. Alan M. Carroll "And then you say, carroll@s.cs.uiuc.edu We have the Moon, so now the Stars..." - YES CS Grad / U of Ill @ Urbana ...{ucbvax,pur-ee,convex}!s.cs.uiuc.edu!carroll ----------------------------- From: noe@s.cs.uiuc.edu Subject: Re: slicing the date Date: 29 Mar 89 23:01:00 GMT Nf-ID: #R:s.cs.uiuc.edu:216000010:s.cs.uiuc.edu:216000012:000:451 Nf-From: s.cs.uiuc.edu!noe Mar 29 17:01:00 1989 To: info-unix@sem.brl.mil Maybe expr will work? Again, it's going to look sort of hideous, but it should work. For example, if date produces "Wed Mar 29 16:57:30 1989" then NOW=`date` PS1=`expr "$NOW" : '... ... \(..\)'` would set PS1=29, the day of the month. This is Bourne shell, by the way, and would also work in Kourne shell (-: but not C-shell. I assume anyone forcibly removed from a System V environment at least has the good sense to avoid csh like the plague. ----------------------------- From: Eduardo Krell Subject: Re: Future at Berzerkeley Date: 31 Mar 89 00:32:29 GMT Sender: netnews@ulysses.homer.nj.att.com To: info-unix@sem.brl.mil In article <4814@macom1.UUCP> larry@macom1.UUCP (Larry Taborek) writes: >My question is, since Berkeley has been paying licsence fees to >AT&T for portions of their code, will AT&T pay Berkeley for >portions of their code included in SVR4? As far as I know, AT&T is negotiating with Sun to merge SunOS features into SVR4, not with Berkeley. What the financial arrangements between AT&T and Sun are, I have no idea. Does Sun pay Berkeley for using BSD code in SunOS? How about all other vendors who are "BSD derivatives"? Isn't BSD code public domain (that is, BSD code that is not derived from AT&T)? If so, why should AT&T or Sun or anyone else have to pay for getting something that's already in the public domain? Eduardo Krell AT&T Bell Laboratories, Murray Hill, NJ UUCP: {att,decvax,ucbvax}!ulysses!ekrell Internet: ekrell@ulysses.att.com ----------------------------- From: "Randal L. Schwartz @ Stonehenge" Subject: Re: Some csh how-to, please Date: 30 Mar 89 19:29:13 GMT Sender: news@omepd.uucp Keywords: csh C-shell shell programming unix read Posted: Thu Mar 30 11:29:13 1989 To: info-unix@sem.brl.mil In article <1175@Portia.Stanford.EDU>, karish@forel (Chuck Karish) writes: | [ a lot of stuff deleted] | An sh script to do this job would look like | | cat file | | while read fie | do | echo A $fie | done | | No magic cookies needed; 'read fie' fails at EOF. | A pipe or redirect after 'done' receives the entire output of the loop. | | csh is OK as an interactive command line interpreter. As a programming | language, it's not so hot. My advice would be to get a copy of | Kernighan and Pike, and use sh instead. Why do people use 'cat' at every opportunity? Maybe that's why they sell so many multiprocessor machines? :-) :-) Simpler: while read fie do something with $fie done ...!uunet!tektronix!biin!merlyn | \ Cute quote: "Welcome to Oregon... home of the California Raisins!" / ----------------------------- From: "Randal L. Schwartz @ Stonehenge" Subject: another way (was Re: slicing the date) Date: 30 Mar 89 19:41:44 GMT Sender: news@omepd.uucp Posted: Thu Mar 30 11:41:44 1989 To: info-unix@sem.brl.mil In article <1177@Portia.Stanford.EDU>, karish@forel (Chuck Karish) writes: | set `date | tr ':' ' '` | hr=$4 | min=$5 | sec=$6 | | If that's not ugly enough for you, use this instead: | | expr "`date`" \: ".*\([ 0-9][0-9]:..\):.*" For the truly ugly (and no additional processes), try: oldIFS="$IFS" IFS=": "; set - `date`; hr=$4 min=$5 sec=$6 IFS="$oldIFS" Amazing what /bin/sh can do for you if you let it. :-) (I'm trying to write EMACS in /bin/sh... anyone want to help? :-) Just another UNIX hacker, -- / Randal L. Schwartz, Stonehenge Consulting Services (503)777-0095 \ | on contract to BiiN (for now :-) Hillsboro, Oregon, USA. | |<@intel-iwarp.arpa:merlyn@intelob.intel.com> ...!uunet!tektronix!biin!merlyn | \ Cute quote: "Welcome to Oregon... home of the California Raisins!" / ----------------------------- End of INFO-UNIX Digest ***********************