Sun-Spots-Request@RICE.EDU (William LeFebvre) (08/23/88)
SUN-SPOTS DIGEST Sunday, 21 August 1988 Volume 6 : Issue 193 Today's Topics: Finding which process is accessing a filesystem `bmx' - Bitmap Conversion priority vs. niceness? /usr/lib/calendar: determining if a file is NFS mounted? automounting ... why bother? Plot10 for Sun's? how reliable is the Sony color monitor? a WORM for the 3/160? Transputer products on Sun? Anti-aliased lines? Trying to locate ibmtape and ansitape Re: Format of a ".o" file 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: Fri, 22 Jul 88 20:16:27 EDT From: dupuy@columbia.edu (Alexander Dupuy) Subject: Finding which process is accessing a filesystem Reply-To: dupuy@columbia.edu After seeing Richard Tobin's find-fs program, I was inspired to add shared text reference checking to my ofiles program, just so that it will do everything that find-fs does, plus the whole lot more it already did. The README describes it fairly well - it's at the front of the shar file. Now who wants to port it to 4.0? @alex inet: dupuy@columbia.edu uucp: ...!rutgers!columbia!dupuy [[ The shar file has been placed in the archives under "sun-source" as "ofiles.shar". It is 33998 bytes long, and it can be retrieved via anonymous FTP from the host "titan.rice.edu" or via the archive server. For more information about the archive server with the request "send sun-source ofiles.shar", send a mail message containing the word "help" to the address "archive-server@rice.edu". --wnl ]] ------------------------------ Date: Tue, 16 Aug 88 13:03:40 PDT From: gandalf@csli.stanford.edu (Juergen Wagner) Subject: `bmx' - Bitmap Conversion This is to announce the new release of `bmx' (previously called `bm' which caused some naming confusion with the fast Boyer-Moore grep). I have posted the new version to comp.sources.unix on Saturday, so if Rich Salz approves the package, it will be available within the next couple of weeks. The package is also available via anonymous ftp from csli.stanford.edu (Internet#36.9.0.46) from pub/Gandalf/bmx* (do it at night). For those of you who can neither ftp the files nor get them from comp.sources.unix, I should point out that there are some mail-based unix-sources servers around. In exceptional cases I will mail the package. The program comes in three shar files, one README, and some man pages. I am not going to mail any sample bitmap files because you should be able to get some elsewhere. If not, I can point to pub/Gandalf/bits.tar.Z (again on csli.stanford.edu). This file contains some sample bitmaps in some of the formats I support. If you want to test the package, try two things: o a Sun rasterfile, o a X10/X11 bitmap in header file format. Both should work fine on any machine I've tried the program on. The only problems I've encountered is with some compilers which don't like the typedef void Void; in BM/include/gandalf/bitmap.h (line 19). You should just replace that by # define Void void and everything should compile fine. If you are compiling on a Sun4 with SunOS 4.0 (i.e. have shared libraries), link everything -Bstatic. Sometimes I get weird unexplainable errors if they are linked with -Bdynamic. I am still working on the package, extending it to support more formats and operators. If you happen to have a weird format, let me know. Please, include a format specification and some sample bitmaps, so I can try to implement them. (FAX format is not ready yet, I am working on GRF and GIF) Mail bug reports, suggestions, comments, etc. to gandalf@csli.stanford.edu. Please, note that my package is not supposed to be a converter between graphics languages (like PostScript, HPGL, ImPress, etc.) but just a mere conversion program for pure raster images. I am not sure whether it is a good idea to include fonts here. I may think about reading one character out of a font, but generating one font from another is outside the scope of `bmx'. Note, that the package is not in the public domain. It is copyrighted, and I don't want to have it included in any licensed software without my consent. I hope the `bmx' package with its bitmap i/o library is of use for you. Enjoy, Juergen "Gandalf" Wagner, gandalf@csli.stanford.edu Center for the Study of Language and Information (CSLI), Stanford CA PS: Here is an extract from the program's help message: *** Bitmap Conversion program: `bmx' (Version beta-2.2) *** [BitConv, Copyr 1988, Juergen Wagner (gandalf@csli.stanford.edu)] The command line syntax is: bmx [ operators... ] where 'operators' is a list of operation names followed by some optional arguments. Operation names are: -in Read a bitmap file -out Write a bitmap file -info Print information about a bitmap -list List all bitmaps currently loaded -log Write an action log -inv Invert all bits -rev Reverse bits (word-wise) -rev2 Reverse bits (byte-wise) -swap Swap bytes (word-wise) -frame Frame the bitmap -concat Concatenate bitmaps -help Print help information The available input/output formats are: gandalf (gandalf, g) Gandalf's favourite format ascii (asc, a) ASCII representation raster (ras, r) Sun Rasterfile iconedit (icon, i) Sun Iconedit file impress (imp, l) imPRESS format dlionbm (dlion, d) Xerox-1108 InterLISP bitmap laserjet (lj, h) HP LaserJet format macpaint (macp, m) MacPaint documents postscript (ps, p) (nano-)PostScript (images only) printronix (ptx, t) Printronix printer format sixel (six, s) Sixel format (LN03 printer) ucbvax (ucb, u) Weird format of ucbvax:pub/xbackground/* xbitmap (xicon, x) X bitmap (header file) xwindow (xwd, w) X Window Dump format ccittfax (fax, f) G3/G4 FAX Format (CCITT Rec. T.4/T.6, VII.3) cvl (cvl, c) CVL Format pppf (ppp, pp) Program Products Picture File Format ------------------------------ Date: Tue, 16 Aug 88 15:45:00 EDT From: "Ruth Milner, Systems Manager x2746" <SYSRUTH@UTORPHYS.BITNET> Subject: priority vs. niceness? Part of this may not be particularly a Sun question, but here goes anyway (send those responses directly to me if you don't think they belong on the list). 1. I have recently had occasion to bump our nfsd's (12) on our Sun 4/280 serving 10 diskless 3/50's (and exporting fs's to a few other Suns) up to (down to?) a niceness of -5. The result of this was that a diskless client running a very disk-intensive program to munge many files into a few files dropped from 30 minutes elapsed time to about 3.5 minutes. Does anyone know of any potential problems doing this? The Sun 4/280 is also used as a number-cruncher for low-priority CPU-intensive jobs and, to a much lesser extent, for interactive jobs (mostly editing, compiling, and reading mail). The Sun 4 is at 3.2FCS; the 3/50's are at 3.5 (if this matters). 2. The non-Sun-specific question has to do with "priority" and "niceness". The display from "top" (thank you, WNL, for a very nice utility) lists both of these numbers, but I can't find any clear description either in manuals (I admit I haven't gone through with a fine-toothed comb) or verbally from some of the local UNIX gurus on the real difference between priority and niceness. What is each used for? How exactly do they differ? How is priority calculated (something to do with the niceness, whether the process has a tendency to use up its CPU quantum - how long is that, anyway? - etc. etc., I assume)? On our Sun 4, most system processes (daemons, etc.) run with niceness 0 and priority 1-5. Interactive logins have niceness 0 and priority 15 or 25 (root). CPU-devouring background jobs run at niceness 19 and priority 102. How does the allocation of the CPU work? Are lower-priority jobs guaranteed a given portion of the CPU based on the priority level, or is it absolutely based on priority so that lower-priority jobs only get CPU if there is *nothing* at a higher priority wanting it? Does Sun do this any differently than the rest of the UNIX world, or is it something defined by "if UNIX then schedule- in-this-way"? If you don't feel like explaining this, please let me know if there is a *good* description of this in either the Sun manual set or someone's book. I would really like to understand how this works. Thanks a lot. Ruth Milner Systems Manager University of Toronto Physics sysruth@utorphys.bitnet sysruth@helios.physics.utoronto.ca [[ Well, there probably isn't a very detailed explanation in the Sun documentation. This is all deep Unix internals stuff. "Niceness" is exactly the figure you set with "nice" and "renice". The lower the number, the less "nice" you are being to other processes. In other words, the lower the number the more preference your job is given. But niceness is only one factor used in calculating the actual scheduling priority of processes (and the only one directly settable by the user). "Priority" is the final number that the scheduler uses to pick the next process to run. Several facts are used to calculate the priority, and I'm not even sure what they all are. If your process blocks waiting for input, its priority (which will only matter once the read completes) depends on the speed of the device it's waiting on (this includes waiting for text and data pages). Again, preference is given to processes with a lower priority. Further discussion and details about this is probably more appropriately discussed on one of the Unix interest lists. --wnl ]] ------------------------------ Date: Wed, 17 Aug 88 10:52:08 CDT From: drl@vuse.vanderbilt.edu (David R. Linn) Subject: /usr/lib/calendar: determining if a file is NFS mounted? In trying to figure out why calendar(1) quit working when we took our YP master to 4.0, I have discovered that /usr/lib/calendar now takes an argument (a filename) and somehow determines whether the file is NFS mounted or not. Using "strings -2 - /usr/lib/calendar", I fail to find a "/etc/fstab" or a "/etc/mtab" so I don't think the program is using getmntent(3). I am working on a program that needs to distiguish local files from remote files. Can anyone tell me how /usr/lib/calendar does it?? David David Linn drl@vuse.vandebilt.edu ...uunet!vuse!drl ------------------------------ Date: Wed,17 Aug 10:29:32 1988 From: Mario Wolczko <mario%mushroom.computer-science.manchester.ac.uk@nss.cs.ucl.ac.uk> Subject: automounting ... why bother? Under SunOS4.0, NFS filesystems can be either mounted in the normal way (using mount), or dynamically mounted "on demand" using automount. Apart from saving startup time when booting, why bother automounting? If you have fewer filesystems mounted, is your system faster? Is access to an automounted filesystem faster or slower than to a mounted fs? Mario Wolczko Dept. of Computer Science Internet: mario%ux.cs.man.ac.uk The University USENET: mcvax!ukc!man.cs.ux!mario Manchester M13 9PL JANET: mario@uk.ac.man.cs.ux U.K. Tel: +44-61-275 2000 extn 6146 [[ One of the advantages to automounting is that the file system is automatically unmounted when the last file on that system is closed. This avoids unnecessary hanging problems on the client if the server were to go down. There might also be a difference in overall system performance since fewer system resources are used when the file system is not mounted. --wnl ]] ------------------------------ Date: Wed, 17 Aug 88 09:18:46 EDT From: (Tony Cardazzi) <@williams.edu:tony@cc.williams.edu> Subject: Plot10 for Sun's? We currently are running FORTRAN programs that use Plot10 FORTRAN subroutine calls on Tektronix 4107 terminals on our VAX 11/785. The Tektronix terminals produce color plots by dumping what is on the screen to the Tektronix 4696 color printer. We are looking for a similar package for the Sun's, contacts, or directions to who may sell such a package. The reason we are looking for such a package is that we already have written many programs using the Plot10 library and don't have the time or manpower to convert these programs to the Suns. We have a 3/180 file server and 3/50 b/w clients all running under SunOS 4.0. Any information or suggestions would be greatly appreciated. Anthony Cardazzi Computer Center Williams College E-mail cardazzi@cc.williams.edu Bitnet cardazzi@williams Telephone (413) 597-3077 ------------------------------ Date: Wed, 17 Aug 88 15:39:04 PDT From: celeste@coherent.com (Celeste C. Stokely) Subject: how reliable is the Sony color monitor? I'm considering getting a Sony 16" or 19" color monitor for my 3/60. How reliable are Sony monitors? I understand the 19" isn't even out yet, but have you ever had the 16" monitor break? My Sony TV lived many many years before it died, but my Sony walkman didn't last a month. Testamonials and horror stories to me by email, please. ..Celeste Stokely Coherent Thought Inc. UUCP: ...!{ames,sun,uunet}!coherent!celeste Domain: celeste@coherent.com Internet: coherent!celeste@ames.arpa or ...@sun.com or ...@uunet.uu.net VOX: 415-493-8805 SNAIL:3350 W. Bayshore Rd. #205, Palo Alto CA 94303 ------------------------------ Date: Wed, 17 Aug 88 11:59:01 EDT From: reidar <reidar@cucard.med.columbia.edu> Subject: a WORM for the 3/160? We have a 3/160 and would very much like to hang a WORM on it. Has anyone done this? If so, did you treat it like a disk (i.e. file system) or a tape of some sort? Thanks in advance, Reidar Bornholdt Columbia University, College of Physicians & Surgeons reidar@cucard.med.columbia.edu cmcl2!cucard!reidar ------------------------------ Date: 17 Aug 88 06:01:46 GMT From: mkkam@wael.cs.uh.edu (Francis Kam) Subject: Transputer products on Sun? If anyone has news/hints/pointers on transputer products which can plug onto the VME bus of a Sun 3/280, or as an external box connected via SCSI port or anything to a 3/[50,60], please forward me a copy. The only company I know now is Niche, but I believe there should be more. Thank you. CSC-3475 Francis Kam Computer Science Department Internet: mkkam@sun1.cs.uh.edu University of Houston CSNET: mkkam@houston.csnet 4800 Calhoun Phone: (713)749-1748 Houston, TX 77004. ------------------------------ Date: Wed, 17 Aug 88 14:13:11 PDT From: chaos%gojira.Berkeley.EDU@jade.berkeley.edu Subject: Anti-aliased lines? Does anyone have an anti-aliased line drawing routine for pixrect level of SunView? That is fast? (Using put() pixel by pixel is slow). Jim Crutchfield Physics, UCB (415) 642-1287 ------------------------------ Date: Tue, 16 Aug 88 21:04:03 PDT From: lim@isl.stanford.edu (Kelvin O. Lim) Subject: Trying to locate ibmtape and ansitape There was some recent activity about reading and/or writing IBM tapes. Two public domain programs, ibmtape and ansitape, were mentioned. Could someone give me leads as to where I can get them? Ftp would be preferable but I'm sure other methods are possible. Thanks, Kelvin O. Lim lim@isl.stanford.edu ------------------------------ Date: Tue, 16 Aug 88 23:01:39 EDT From: Don Hopkins <don@brillig.umd.edu> Subject: Re: Format of a ".o" file > From: attcan!utzoo!henry@uunet.uu.net > Danger! Your Sun software license (if it's like ours) specifically forbids > decompiling or disassembling! > > (I suppose if you can sight-read 68020 binaries that's okay! :-)) Speaking of which, does anybody have a PostScript font for sight reading SPARC binaries? (Can you say "cat core"?) -Don ------------------------------ End of SUN-Spots Digest ***********************