Sun-Spots-Request@Rice.edu (William LeFebvre) (10/25/88)
SUN-SPOTS DIGEST Monday, 24 October 1988 Volume 6 : Issue 273 Today's Topics: Re: UUCP, bug or feature (3) Re: netstat -r is broken with sub-netted Class B nets Re: ftp problem Re: srecs from a.out Re: SunOS 4.0 Suntools speedup Re: multiple NFS mounts under 4.0 Re: NeXT Bitnet mangles source files Remote booting a Sun-3/280 -- failure! NFS file inconsistency bug Dvi previewer to run under suntools? Frame Maker to (di)troff translator? detecting use of fpa in executables? TTY Windows in Sunview? PD DATABASES? Send contributions to: sun-spots@rice.edu Send subscription add/delete requests to: sun-spots-request@rice.edu Bitnet readers can subscribe directly with the CMS command: TELL LISTSERV AT RICE SUBSCRIBE SUNSPOTS My Full Name Recent backissues are available via anonymous FTP from "titan.rice.edu". For volume X, issue Y, "get sun-spots/vXnY". They are also accessible through the archive server: mail the request "send sun-spots vXnY" to "archive-server@rice.edu" or mail the word "help" to the same address for more information. ---------------------------------------------------------------------- Date: Thu, 20 Oct 88 22:26:32 EDT From: Preston Mullen <mullen@css.nrl.navy.mil> Subject: Re: UUCP, bug or feature (1) Reference: v6n266 To get around uucp's setting the modem's autoanswer bit, patch the dial string in /usr/lib/uucp/uucico using adb, GNU emacs, or even textedit (if you then use something else to delete the newline textedit adds at the end). ATV0Q0E0S0=1S2=255S12=255 ^ change 1 to 0 You can likewise patch tip. Adding the S0=0 to the phone number in /etc/remote works fine for tip unless you want to exploit /etc/phones by putting a "@" in the pn string: the "@" isn't recognized as special if there is other stuff in the pn string. On Gould systems running UTX 2.0, you can specify the interaction of tip and uucp with the modem in /etc/modcap (analogous to termcap). This supposedly lets you adapt tip and uucp to just about any modem. Presumably, other systems have something like this too; I wonder why SunOS doesn't. ------------------------------ Date: 21 Oct 88 02:29:56 PDT (Fri) From: daver!mfgfoc!exodus@sun.com (Greg Onufer) Subject: Re: UUCP, bug or feature (2) > What can you suggest? 1) Patch the uucico binary (I personally do not recommend this, but it works _very_ well). As long as you copy it first you should be okay. 2) Use the following shell script: (call it uureset or whatever) #!/bin/sh # Change baudrate and anything else system dependent cu -s 1200 S0=0 >/dev/null 2>&1 exit 0 Call it in your /usr/lib/uucp/uucp.* scripts at the very end (so it gets called after all of the polls) and execute it by hand after each manual uucp (ie. not started by cron via uucp.* scripts). I have better work-arounds for TrailBlazers... to get auto-answer to turn on and off via modem control signals. Tricky, but it seems to work. TB's are more difficult since they are nice about resetting to default/EEPROM values when DTR drops. But they also will not answer while DTR is dropped and therein lies the trick. Also, since TBs do not have an Auto-Answer indicator, this method gives visual feedback to the answer status--- necessary for my home machine on my normal telephone line! --Greg ------------------------------ Date: Fri, 21 Oct 88 09:22:54 EDT From: Sergei A. Gourevitch <asg@space.mit.edu> Subject: Re: UUCP, bug or feature (3) Didn't you have to specify the dialing sequence in /usr/lib/uucp/L.sys? You can do the same thing there that you did with tip; reprogram your modem. ------------------------------ Date: Thu, 20 Oct 88 11:31:53 EST From: trinkle@purdue.edu Subject: Re: netstat -r is broken with sub-netted Class B nets It is not really netstat that is broken. Use "netstat -n -r" to see what it really gives you. Netstat uses getnetbyaddr() to convert the net number (128.118.6.0 in your case) to a name. In turn getnetbyaddr() uses a routine called nettoa() to truncate the net number strings ("128.118.6.0") at the appropriate "." before looking it up in /etc/networks or with YP. The nettoa() routine is very dumb. It just truncates based on the class of the network number. It knows nothing of subnets. At this level I am not sure what else it could do. Maybe getnetbyaddr() should have an optional subnet mask argument. I (almost) always use the "-n" option to netstat because it is much faster and when things are broken (which is usually the case if I am using netstat), it does not hang trying to look up names. Also, gateways that have multiple interfaces but one official name do not show up as meaningful information. Besides, you should have all you IP addresses memorized (:-) Daniel Trinkle trinkle@cs.purdue.edu ARPA Department of Computer Sciences trinkle%purdue.edu@relay.cs.net CSNET Purdue University {ucbvax,decvax}!purdue!trinkle UUCP West Lafayette, IN 47907 (317) 494-7844 PHONE ------------------------------ Date: Thu, 20 Oct 88 11:04:47 EDT From: steve@umiacs.umd.edu (Steven D. Miller) Subject: Re: ftp problem You should check your ifconfigs in /etc/rc.boot to be sure they're really happening. (Best check: type b -sb at the monitor prompt, then do the stuff in rc.boot by hand, including the ifconfig.) I've seen (and reported to Sun) a problem where diskless nodes will seem to work OK if they learn their IP address purely via reverse ARP, but the address in the ifnet structure has garbage where it should have zeroes, and binding local addresses fails when that garbage gets compared to zeroes somewhere else. The ifconfig stuffs the address into the interface structure properly. I suspect that this problem is in all SunOSes up to (but possibly not including) 4.0. I'll also bet that the ifconfig is failing on the clients because of some interaction with YP and the nameserver. You might try putting the client IP addresses directly into their ifconfig lines in rc.boot and see if that helps. (Does it sound like this has happened to me before?) -Steve Spoken: Steve Miller Domain: steve@mimsy.umd.edu UUCP: uunet!mimsy!steve Phone: +1-301-454-1808 USPS: UMIACS, Univ. of Maryland, College Park, MD 20742 ------------------------------ Date: Fri, 21 Oct 88 08:41:28 EDT From: brad%cayman@harvard.harvard.edu (Brad Parker) Subject: Re: srecs from a.out Malcolm says: > I am looking for a program to compile Motorola S-Records from standard > 68000 assembler code.... I have such a beast, it will even "split" srec files for burning into eproms. It's derivative of "dl68" from Stanford. Send mail and I'll send it. Brad Parker ------------------------------ Date: Thu, 20 Oct 88 10:47:59 PDT From: tsd@sun.com (Terry Donaldson) Subject: Re: SunOS 4.0 Suntools speedup Reference: v6n263 >>From: dg-rtp!meissner@mcnc.org (Michael Meissner) >>... >>the patch: >>Edit /etc/syslog.conf and enter the following line before the first >>non-comment line, starting in column 1 (note the comma space): >>define(LOGFILE, 1) >>... The correct entry for /etc/syslog.conf is: define(LOGHOST,1) ------------------------------ Date: Thu, 20 Oct 88 19:47:37 EDT From: Fuat C. Baran <fuat@cunixc.cc.columbia.edu> Subject: Re: multiple NFS mounts under 4.0 Reference: v6n265 ekrell@ulysses.att.com writes: >I've got used to type "mount -a" to try to mount all the file systems in >/etc/fstab. Under 4.0, however, (on a 3/60) the NFS file systems which >were already mounted get mounted AGAIN each time I do this. I looked at >/etc/mtab and, sure enough, the entries are duplicated. Yes, we've seen this too. We are running SunOS 4.0 on a Sun 4/280, with NFS mounts to other Sun 4/280's and Ultrix 2.0 on a VAX 8700. This happened a lot when we first got the machines and were setting them up. Now we don't notice this, since things are relatively stable and file systems only get mounted at boot, and stay that way till we crash or shutdown. I don't think we ever reported this to Sun, though we probably should have. I quick look at the "Software Technical Bulletin"s (index in Sept. 1988 issue) did not reveal anything on this. --Fuat INTERNET: fuat@columbia.edu U.S. MAIL: Columbia University BITNET: fuat@cunixc.cc.columbia.edu Center for Computing Activities USENET: ...!rutgers!columbia!cunixc!fuat 712 Watson Labs, 612 W115th St. PHONE: (212) 280-5128 New York, NY 10025 ------------------------------ Date: Thu, 20 Oct 88 09:50:58 PDT From: franz!snooze!layer@ucbarpa.berkeley.edu (Kevin Layer) Subject: Re: NeXT Reference: v6n265 A small correction to David NeXT summary: > From: frame!troy!drf@sun.com (David Fuchs) > ... > Sybase, Mathematica, Franz, WriteNow will all be built-in. `Franz' is the name of the company, Allegro CL is the name of the Common Lisp. Kevin Layer Franz Inc. ------------------------------ Date: 20 Oct 88 08:22:27 PDT (Thu) From: coraki!pratt@sun.com Subject: Bitnet mangles source files While I hadn't run into ^ as a problem for uuencoded files through bitnet (such nodes needs to be spoken to firmly!) I had run into the trailing-space-deletion problem on several occasions. Although deleting trailing spaces is enough to upset uudecode, it is not an information-losing operation because uuencode produces fixed-length lines and hence deleted trailing spaces can be reconstructed. So either put another filter ahead of uudecode to restore trailing spaces, or if you have the uudecode source hack that up appropriately. The advantage of this approach is that you don't have to wait for the rest of the world to change to a new uuencode convention. Vaughan Pratt ------------------------------ Date: Thu, 20 Oct 88 14:11:47 EDT From: rang@cpsin3.cps.msu.edu (Anton Rang) Subject: Remote booting a Sun-3/280 -- failure! We have a new Sun-3/280, and are trying to get it set up. It doesn't have a local tape, so we are following the procedures in "Installing UNIX on tapeless workstations". We've been through this before, but haven't encountered this particular problem before (at least, as far as anyone still here can remember). First, a little background. We're on a class-A network (35.xxx), so we need to boot a little differently than on class-C. The command we used to start copy was: >b ie(0,80c65,0)stand/copy <-- 80c65 ==> .8.12.101 (server addr). This works fine. (BTW, the address of the new machine is 35.8.12.60.) The problem is that when standalone copy is running, we try: From: ie(0,80c65,0)minifs and it doesn't work. Using 'traffic' to check the Ethernet reveals that the new machine is sending ND requests to the server, but the server isn't sending anything back. Suggestions? Any help would be appreciated. Anton Rang rang@cpswh.cps.msu.edu ------------------------------ Date: Thu, 20 Oct 88 11:49:37 PDT From: pixar!rta@sun.com (Rick Ace) Subject: NFS file inconsistency bug [This may be a re-hash of a previous discussion, but I haven't seen a "reproduce by doing this" description yet, so here is one.] Environment: Host A - Sun-4/280 server running SunOS 4.0 Host B - Sun-3/60 diskless running SunOS 4.0 Symptom: Filesystem is local to host A and NFS-mounted by host B. A file in that filesystem appears to have different contents when viewed from hosts A and B. Reproduce-by: 1. Run this program on host B. main(argc, argv) int argc; char **argv; { int fd; (void) unlink("testfile"); fd = creat("testfile", 0777); if (fd < 0) { perror("cannot create testfile"); exit(1); } write(fd, "foo\n", 4); fsync(fd); close(fd); } 2. Give this C-shell command on host A: echo 'foo bar' > testfile 3. Give this command on hosts A and B: od -c testfile Compare the output from host A to the output from host B. We have observed that removing the fsync() call from the C program will cause the problem to go away. Rick Ace Pixar 3240 Kerner Blvd, San Rafael CA 94901 ...!{sun,ucbvax}!pixar!rta ------------------------------ Date: Thu, 20 Oct 88 11:20:42 edt From: rjc@cs.brown.edu Subject: Dvi previewer to run under suntools? I am looking for a dvi previewer that will run under suntools. I recently tried to use dvipage, but we don't have the necessary pxl fonts here, and apparently no one is willing to make them. Does anyone know where I can ftp a set of the fonts (at least enough of them to test out the program)? Even better, does anyone know where I can get a previewer that doesn't use pxl at all? ==> Randy <== uucp: ...!{ihnp4,decvax}!brunix!rjc Randy Calistri bitnet: rjc@browncs.bitnet Box 1910 arpa: rjc%cs.brown.edu@relay.cs.net Department of Computer Science csnet: rjc@cs.brown.edu Brown University ------------------------------ Date: Thu, 20 Oct 88 12:54:11 EDT From: beareq!quants5!fitz@wheaties.ai.mit.edu (David Fitzgerald) Subject: Frame Maker to (di)troff translator? Chuck Musciano's plug for Frame Maker reminded me of a product enhancement I requested from Frame Technology. Under the Document.save options I would like to see a dump to (di)troff format. So I could create documentation using Maker and display it on a ascii terminal (with the 'man', troff, nroff or other related command). I know I can just dump the text but that is kind of crude. Unfortunately Frame said that although they have had several requests of this nature, they have no plans to implement this in the near future. >From looking at the Maker Interchange Format (MIF) it does not appear too difficult to make the translation of text (graphics is another story). But, before I dive in, has any one else out there considered this? i.e. done any work in that direction. I would prefur to translate to ditroff macros but my Sun only has troff and troff doc so that will be my first choice. Any comments suggestions or the like would be appreciated. Thanks! Dave FitzGerald beareq!fitz ------------------------------ Date: Thu, 20 Oct 88 16:57:09 EDT From: kcb@macaw.jhuapl.edu (Kevin C Brown x4700 1-e136) Subject: detecting use of fpa in executables? I have recently purchased software from a vendor which consists of an executable and some object files. Is it possible, by examining these files, to assure myself that the code was compiled with the fpa option? Kevin Brown Applied Physics Laboratory Laurel, MD 20707 (301) 953-5000 ARPA: kcb@macaw.jhuapl.edu ------------------------------ Date: Thu, 20 Oct 88 18:04:09 EDT From: "Paul R. Jordan" <pj2f@uvaee.ee.virginia.edu> Subject: TTY Windows in Sunview? First, thanks to those who've helped with the font problem I mentioned earlier with Sunview (it was mostly my error in mis-reading the docs...). Next, to a question which I sort-of tried to ask earlier: On our Sun 3/50, an application which sets up subwindows using a TTY process in one of these windows will only execute a startup process in this TTY subwindow ONCE (i.e. I have a program that runs another program which is forked in this TTY). Is there a hard limitation on the number of TTY windows in one application ? and Why, when this TTY program is finished and the TTY window destroyed, can't another TTY window be created in its place ? I've experimented with several combinations of subwindows and TTYs, and have found only ~20 subwindows can be created with 1 TTY, 15 with 2 TTYs, etc. And if the TTY window is destroyed and then recreated, the number of subwindows decreases by 2*(#TTY's). Any advise or ways around the earlier problem ? Thanks in advance, Paul J. (pj2f) ------------------------------ Date: Thu, 20 Oct 88 12:04:04 -0400 From: kraussW@batman.moravian.edu Subject: PD DATABASES? I would be interested in any Public Domain, multi-user databases that are available for Sun 4/280's or Sun 3/50's. Any information anyone could provide would really be appreciated. Please send the info to me direct. Thanks, Bill Krauss UUCP : ..!sun!liberty!moravian!kraussW CSNET: kraussW@moravian.edu INET : kraussW%batman.moravian.edu@relay.cs.net ------------------------------ End of SUN-Spots Digest ***********************