Sun-Spots-Request@Rice.edu (William LeFebvre) (10/22/88)
SUN-SPOTS DIGEST Friday, 21 October 1988 Volume 6 : Issue 268 Today's Topics: Re: NFS mount mail RE: C/A/T troff to ditroff Re: Large screen display for SUN video output Summary: 3rd Party Sun 3/60 Memory Bug in Contool 2.0 (with patch) wierd NFS hangup Portmap, Gateway problems learn for Sun 3/260? color hardcopy of Sun screen - how? news reader for sunview, news or x? Experience with "NSE"? PostScript printers for Suns? locking down systems? 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: Mon, 17 Oct 88 14:31:19 EDT From: beck@svax.cs.cornell.edu (Micah Beck) Subject: Re: NFS mount mail In Sun-Spots v6n258, two notes (from white@cs.unc.edu and Wayne Folta) offered solutions to the problem of reading mail from a Sun client. Both solutions involved directing all mail to one server which delivers it to an exported directory. Clients then mount that directory and read mail from it. I am nervous about doing this because I know that the semantics of concurrent access to NFS files is different from that of uniprocessor Unix. I also know that mail programs tend to rely heavily on the semantics of concurrent access. Of particular concern are non-atomic actions like deleting a file and rewriting it. Can anyone out there provide an authoritative answer: is it safe to read mail using NFS in this fashion? Is there a potential problem when mail is read while new mail is delivered? In the meantime, I read my mail on an overloaded central Vax, while sitting at a Sun-3/60! Micah Beck beck@cs.cornell.edu Dept of Computer Science Cornell University ------------------------------ Date: Mon, 17 Oct 88 12:06:36 -0700 From: greg@aerospace.aero.org Subject: RE: C/A/T troff to ditroff >Is it possible to get ditroff output on a Sun with SunOS 4.0? If not (and >I think not) is there a free C/A/T troff-2-ditroff out there? All replies >welcome.... >[[ Am I missing something here?... --wnl ]] Perhaps I didn't state my question clearly enough. SunOS provides a nice text formatting program "troff". However, as far as I know, it is totally useless to most people because its output in NOT device independent. It is specifically designed to be used with a Computer Aided Typesetter or C/A/T. There is a rich assortment of software out there which uses ditroff style output (ascii), but it is useless with the output from SunOS "troff". My question, then, is "Is there a "ditroff" program provided with SunOS or a way to make troff output ditroff style (ascii) commands?" Once again, If there is no free way to directly get ditroff type output commands, is there a free program to convert the device dependent "troff" output to ditroff style output? As a final note, sense I have not gotten any positive responses about the free filter mentioned above, I have started writing one (it doesn't appear to be as hard as I first thought) and it should be available soon. [[ AAhhhh. I see. You cat get the "device independent troff" package straight from AT&T Bell Labs. I don't know how much it costs, tho. It is probably worth it, tho, because "ditroff" is much better than "troff". --wnl ]] ------------------------------ Date: Mon, 17 Oct 88 13:41:56 EDT From: dpk@morgan.com (Douglas P. Kingston) Subject: Re: Large screen display for SUN video output The Barcographics 400 will handle Sun Color systems and many other things. It is a high-end projection television system. It was the best of the units that we evaluated. -Doug- PS. See some of the Sun-Spots articles about 6 months ago for more info. ------------------------------ Date: Mon, 17 Oct 88 13:54 EDT From: "Eugene C. Libardi" <AIDEL@UMAECS.BITNET> Subject: Summary: 3rd Party Sun 3/60 Memory Sorry this took so long, but things got crazy around here and I just wasn't able to get to it. I didn't get as many replies as I thought I would. Of the 5 I got, none were bad comments on any of the 3 vendors mentioned. These vendors are Parity, Clearpoint, and Helios. Two were good recommendations for Parity, two were good recommendations for Helios, and one was a good recommendation for Clearpoint. Add my good comments on Clearpoint and it's a three way tie. The pricing and service seem to be pretty much equal, too. Price/Meg (US$): List Univ. Disc. ---------------- ---- ---------- Parity (actually Stout Assoc, 489 405 Parity's rep in my area) Helios 490 410 Clearpoint 495 410 All three have a lifetime warranty, 24 hour repair/replace policy (although Parity says only if it's critical), and deliver in a week. Clearpoint has a 24 hour, 7 days-a-week 800 support line. Parity was iffy on support, and Helios I don't know about. Both Clearpoint and Parity were willing to wheel and deal, but not Helios. Some interesting info: Helios is the oldest 3rd party Sun supplier. They supply 3rd party products for Sun's whole product line. Clearpoint is the biggest, selling memory for many hardware platforms (DEC, Appollo, Sun, ...), and have a large sales/support force. I've dealt with Clearpoint before. My salesperson (Rebecca Haberman) is very good and very helpful. Clearpoint has kept all their promises and has been there with the support and come through on the repair/replace policy. They also publish the remarkably objective and well-written "Designer's Guide to Add-In Memory", packed with a lot of good info. Though they are marginally more expensive, we decided to go with them because of the benefits a large company offers, the experience of their sales staff, and the desire to keep the number of vendors we deal with down to a minimum. How to contact them all: Clearpoint Inc. 99 South Street Hopkinton, MA 01748 (617) 435-5395 Parity Systems, Inc. 20 S. Santa Cruz Ave. #102 Los Gatos, CA 95030 (408) 354-1500 Helios (408) 432-0292 I don't have Helios's address, but here is a local vendor: RML Associates (303)799-6525 Larry McNiff Helios c/o RML Associates 6535 S. Dayton Suite 1950 Englewood, Co 80111 I hope someone is able to use this information and that it is accurate. These prices are as of 10/17/88. The way chips have been going lately, I'm sure these will get lower before they get higher. Gene Libardi Mechanical Design Automation Lab Internet: aidel@ecs.umass.edu Dept. of Mechanical Engineering Bitnet: aidel@umaecs University of Massachusetts/Amherst Tel.: (413) 545-3599 Amherst MA 01003 ------------------------------ Date: Mon, 17 Oct 88 14:19:36 EDT From: Chuck Musciano <chuck@trantor.harris-atd.com> Subject: Bug in Contool 2.0 (with patch) A bug in contool 2.0 caused contool to hang when presented with a filter with a bad regular expression in it. The following patch fixes the bug. Thanks to Doug Curtis of GE who led me to this bug. Chuck Musciano Advanced Technology Department Harris Corporation (407) 727-6131 ARPA: chuck@trantor.harris-atd.com [[ The patch has been placed in the archives under "sun-source" as "contool.patch.shar". It is 5355 bytes long. 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, send a mail message containing the word "help" to the address "archive-server@rice.edu". --wnl ]] ------------------------------ Date: Mon, 17 Oct 88 14:31:33 EST From: mckay@courageous.ecn.purdue.edu (Dwight D. McKay) Subject: wierd NFS hangup I've noticed on several occasions the following NFS hangup: Client sends NFS_WRITE of slightly less then 8K or 4K. Write is send as a fragmented UDP packet. Nothing come back from the server. The write is retried. Nothing comes back. Eventually client (and later server) crashes running out of MBUFS. Equipment involved: Server (gus) 4/280, 2 Hitachi DK815-10's, second ethernet board, 32 Meg. memory Clients seen to exhibit this behavior so far: (gus7) 4/110, color, FPU, 8Meg. (gus13) 3/60, color, FPU, 4Meg. All running SunOS 4.0, kernels built from source tape I've observed this behavior using etherfind running on another Sun on the same ethernet as gus and her clients. I've also looked at the MBUFs on gus using netstat -m. The 4/110 seems to have run out of swap space or crashed and be in the process of rebooting when this condition is noted. I do not have a crash "partition" setup for it or any other clients. I've only had this group of Suns up for two weeks and it's happened twice that I've been here to observe. Once after a 4/110 crash and once after the 4/110 has run out of swap space. Has anyone seen this before? Is there a workaround or better still a fix? --Dwight D. Mckay, ECN Workstation Software Support --Purdue University, Engineering Computer Network --Office: MSEE 104f, Phone: (317) 494-3561 --ARPAnet: mckay@harbor.ecn.purdue.edu, Usenet: ...rutgers!pur-ee!mckay ------------------------------ Date: Mon, 17 Oct 88 14:15:44 PDT From: jimc@math.ucla.edu Subject: Portmap, Gateway problems We have two bizarre problems that we cannot figure out. All occur on Sun 3/180 and 3/280 machines running SunOS v3.5. Does anyone recognize these, and does anyone have a solution or even an explanation? 1. Route-add hangs. We do not run routed. The network looks like this: 128.97.64.x ------ 128.97.64.16 (4.3BSD VAX gateway) (SunOS 3.5) 128.97.12.8 ------- 128.97.12.1 (PC-RT gateway) (Also 4.3 VAX) 128.97.28.1 ----- 128.97.28.x The VAX gateway is up and the PC-RT is down. When booting, the Suns (also the VAXen) do the command /usr/etc/route add uclanet-cs 128.97.64.16 2 where YP's networks.byname says uclanet-cs 128.97.28 All machines hang in this situation. They will hang for half an hour if the RT doesn't revive (which is all too frequent). But when the RT comes up, route-add completes and rebooting continues. I believe the same thing happens when a gateway on our net dies (like 128.97.64.16). We can't see anything in the source where /usr/etc/route, or the kernel end of the route-add ioctl, sends anything to the gateway or expects any reply. It mystifies us how the machine doing route-add can know the address of either the RT gateway or any machine on the 28-net, so that it would even be capable of sending such a request or recognizing such a reply. Yet our machines (SunOS v3.5 and 4.3BSD VAXen) hang. Help! 2. Portmapper croaks. On one day last week, portmap started crashing repeatedly. It would just silently go away, sometimes leaving a core file. We were unable to use dbx because /etc/portmap is stripped, and we could not generate from source an executable version sufficiently identical to the distributed one. It died of a segmentation violation; the routine that it died in is in the RPC library; its name is svc_sendreply() and the relevant code is: bool_t svc_sendreply(xprt, xdr_results, xdr_location) register SVCXPRT *xprt; xdrproc_t xdr_results; caddr_t xdr_location; { struct rpc_msg rply; rply.rm_direction = REPLY; rply.rm_reply.rp_stat = MSG_ACCEPTED; rply.acpted_rply.ar_verf = xprt->xp_verf; /* DIED HERE */ rply.acpted_rply.{various stuff} = various stuff; return (SVC_REPLY(xprt, &rply)); } The pointer xprt was bogus; interpreted as characters it was 'a.ma', reminiscent of that site's hostname of SONIa.maTH.UCLA.EDU (lower case used for suspicious fragment). It's not clear who called the routine; there are five or six possibilities. Portmapper crashed many times that afternoon. The problem went away after a reboot. There was possibly one instance the next day. There was one definite instance the next week, and one (cross fingers) instance this week. Has anyone seen anything like this? Thanks to Vahe Sarkissian of Mathnet for figuring out what little we know about portmapper and for going over the sources on the route problem. James F. Carter (213) 825-2897 UCLA-Mathnet; 6608B MSA; 405 Hilgard Ave.; Los Angeles, CA 90024-1555 UUCP:...!{rutgers,ucbvax,sdcrdcf,{hao!cepu}}!ucla-cs!math.ucla.edu!jimc ARPA: jimc@math.ucla.edu BITNET: jimc%math.ucla.edu@INTERBIT ------------------------------ Date: Mon, 17 Oct 88 13:03:45 CDT From: david@wubios.wustl.edu (David J. Camp) Subject: learn for Sun 3/260? I am looking for the learn program that we used to have on a uVax for a Sun 3/260 running SunOS 4.0 . I checked the comp.sources.unix index to no avail. I would also be interested in other unix tutorial software. Can anyone give me a lead? -David- (314) 362-3635 Mr. David J. Camp Room 1108D ^ Box 8067, Biostatistics 706 South Euclid < * > Washington University Medical School v 660 South Euclid Bitnet: david@wubios.wustl Saint Louis, MO 63110 Internet: david%wubios@wucs1.wustl.edu uucp: uunet!wucs1!wubios!david ------------------------------ Date: 17 Oct 88 19:19:03 GMT From: mkhaw@teknowledge-vaxc.arpa (Mike Khaw) Subject: color hardcopy of Sun screen - how? We have a need to produce some prints of a color Sun screen. Does anyone in the Bay Area rent either color printers or photographic apparatus designed to take screen shots? Thanks, Mike Khaw -- internet: mkhaw@teknowledge.arpa uucp: {uunet|sun|ucbvax|decwrl|uw-beaver}!mkhaw%teknowledge.arpa hardcopy: Teknowledge Inc, 1850 Embarcadero Rd, POB 10119, Palo Alto, CA 94303 ------------------------------ Date: Mon, 17 Oct 88 14:08:25 CDT From: Max Ziff <max@gide.uchicago.edu> Subject: news reader for sunview, news or x? I sure hope this is a naive question. Does anyone have a good window-oriented news reader? I am using lwall's rn (with nntp hacks), and I'm sick of it. I run sunview, but I have friends here who run news (and are considering switching to x) and I'd be interested to hear of options for those environments. Donald Ziff (Max) ARTFL Project University of Chicago (312) 702-0343 max@gide.uchicago.edu ------------------------------ Date: Mon, 17 Oct 88 10:48:45 MDT From: pvi!rickg@boulder.colorado.edu (Rick Gillespie) Subject: Experience with "NSE"? We are about to begin a large project porting an application to the Suns. My question is "does anyone have experience with NSE"? We are debating whether to use SCCS or RSC for source code control and we believe a consideration may be that SCCS will be "easier" to move under NSE. Does anyone have an (informed :-)) opinion? Rick Gillespie UUCP: (boulder,ncar)!pvi!rickg Internet: rickg%pvi@boulder.colorado.edu ------------------------------ Date: Mon, 17 Oct 88 11:38:39 edt From: daniel@dg-rtp.dg.com (Stephen Daniel) Subject: PostScript printers for Suns? I am interested in a laser-printer that meets the following requirements: 1) Postscript capable 2) Works with Sun's transcript postscript software 3) Has an RS232 or ethernet connection 4) Is fast and holds lots of paper 5) Costs less than $15,000. Less than $10,000 would be even better. I'm fairly ignorant of current laser printers. What should I buy? I'll summarize any suggestions I receive. Steve Daniel Data General Corp. daniel@dg-rtp.dg.com Research Triangle Park (Home of DG/UX), NC. mcnc!rti!dg-rtp!daniel 919 549 8421 [[ I saw a recent chart that claimed to list all PostScript laser printers. Not one of them claimed to have an Ethernet connection with IP/TCP support (the DEC printer does DECNet only), even as an option. The closest you can get if you really want it on an Ethernet is an Imagen with "UltraScript" support. Almost all PostScript laser printers have RS232 connections: the most popular one is the LaserWriter. What's "fast"? If you want anything above 8 ppm, you will probably have to pay well over $15000. --wnl ]] ------------------------------ Date: Mon, 17 Oct 1988 10:15:00.65 CDT From: <byers@UKANVAX.BITNET> (Ralph Byers) Subject: locking down systems? We have a number of sun workstations in faculty offices. Although they are kept locked, they are not very secure. How would one go about chaining a workstation down? Are there products on the market? R.B. [[ Someone just asked this question, but I have not seen any responses yet. --wnl ]] ------------------------------ End of SUN-Spots Digest ***********************