From: Digestifier To: Linux-Misc@senator-bedfellow.mit.edu Reply-To: Linux-Misc@senator-bedfellow.mit.edu Date: Sat, 3 Sep 94 18:13:11 EDT Subject: Linux-Misc Digest #705 Linux-Misc Digest #705, Volume #2 Sat, 3 Sep 94 18:13:11 EDT Contents: Ctrl-Z Makes VC Unresponsive ! (Jim Nakamura) Re: **BRAND NEW**Texel CD-ROM Drives ***DIRT CHEAP***!!! (Goran Devic) Re: Unix programming question (Al Longyear) Re: VUE like system for Motif or Linux? (Bill Broadley) Re: How to use 14400bps with modem? (Jim Graham) Max size of SCSI HD? (Dale Elrod) Experiences with Adaptec 2842 bootdisk (Thomas Reutterer) Re: LJ#4? (Joseph W. Vigneau) Re: Promising Linux boxes in September 'Computer Shopper' (Bill Broadley) Re: WABI vs. SoftWindows? (Kevin Brannen) Re: vi for dos (Hugh Johnson) Re: Greater than 1024 Cyls and Linux support? (Bao Chau Ha) Re: Xconfig for Diamond SS24X ... (Craig Bates) Re: CAP for Linux? (Dave Platt) Re: dram -- bogoboost without patching! (Thomas G. McWilliams) Re: 16550AFN serial card (Rob Janssen) ---------------------------------------------------------------------------- From: jzero@netcom.com (Jim Nakamura) Subject: Ctrl-Z Makes VC Unresponsive ! Date: Sat, 3 Sep 1994 15:06:53 GMT I've just realized that when I used Ctrl-Z at ANY time, I get an unresponsive VC as a result. I have to go to another VC, and kill the first VC. This problem is driving me nuts. Has anyone had any similar experiences? (Linux v. 1.1.8). -- No Tijuana insurance! jzero@netcom.com S.F., CA ------------------------------ From: goran@cs.utexas.edu (Goran Devic) Subject: Re: **BRAND NEW**Texel CD-ROM Drives ***DIRT CHEAP***!!! Date: 1 Sep 1994 19:13:37 -0500 In article , H. Peter Anvin wrote: >This is a rip-off. Don't buy one. You can get a 2x-speed drive for >only a few more dollars these days. > My friend just bought one double speed CD ROM for $129 from one retail outlet in Dallas... Not SCSI though... -- o"o +----oOO--=U=--OOo------+ | goran@cs.utexas.edu | +-----------------------+ Smash your forehead on any key to continue... ------------------------------ From: longyear@netcom.com (Al Longyear) Subject: Re: Unix programming question Date: Sat, 3 Sep 1994 15:51:04 GMT mai@wumpus.cc.uow.edu.au (Van Dao Mai) writes: >I have programmed UNIX for a long time and feel frustrated with the way >software is installed on the system. Under UNIX people often have to >hardwire the paths and settings into the executable at compile time. >This is in contrast with DOS that passes the full path name of the execuatble >as argv[0] so that you can search for library + data files. >This is why DOS programmers can install all needed files into one >directory making it easier to maintain and backup. In UNIX after a while >you have no idea where to find the files. You also have binary >distribution problem where you have to make so many links and forget about >them. Just because your program is running does not mean that the code file is still addressable under the name that you have in argv[0]. The argv vector is simply that -- arguments. There is absolutely no requirement that argv[0] even contain the name of the program. Several programs even use it for an additional argument. Let's face facts. UNIX is not MSDOS. You are not alone in the computer. An other program running in the system may change the configuration information which will make the data in argv[0] meaningless. [BTW, this facility did not exist in MSDOS until 2.0. The file name is not in the argument string. It is in a different area of memory -- tacked on to the end of the environment strings.] >The UNIX shell can be modidied to pass the full path name into a program >but this will crash many existing programs or causing some real nuisance. No, I doubt that. Passing the full name will not crash any program of which I am aware. The UNIX programs must be written to accept full names as well as just their own name. It is totally legal and valid to use "fred" as well as "./fred". One has a path, one does not. (They may not be the same program however.) >I just wonder if there is any one working on proposing a standard for Linux >software layout that solve this problem. What I have done is to make the >binary fully configurable after compilation. > - Provide fee space within the executable at compile time > - Provide a program 'register' that scan the executable and > modify information in it (safely) to produced a customised one > that knows where it is installed. It would seem to me that you would be better to place the configuration file in a specific location, such as the /etc directory. Then, in that file, put the file names to the data files. -- Al Longyear longyear@netcom.com ------------------------------ From: broadley@turing.ucdavis.edu (Bill Broadley) Subject: Re: VUE like system for Motif or Linux? Date: Fri, 2 Sep 1994 02:13:23 GMT Vue runs well if you have enough ram. But for me olvwm is a MUCH nicer environment. Takes less ram, more functionality for multiple workspaces. I.e. drag windows into different workspaces, push windows offscreen into another desktop etc. Olvwm gives you the multiple workspaces without getting in the way or taking up to much memory. Of course memory use because you have more workspaces is to be expected. -- Bill Broadley Broadley@math.ucdavis.edu UCD Math Sys-Admin Linux is great. Bike to live, live to bike. PGP-ok ------------------------------ Crossposted-To: comp.os.linux.help,comp.os.linux.admin From: jim@n5ial.mythical.com (Jim Graham) Subject: Re: How to use 14400bps with modem? Date: Fri, 2 Sep 1994 23:12:57 GMT NOTE: Followups re-directed to comp.dcom.modems. In article <778389210snz@vertex.demon.co.uk> Steve@vertex.demon.co.uk writes: >Compression comes in 2 flavours; MNP 5 and V42bis. Well, actually, there are others, but..... :-) >MNP 5 is capable, theoretically, of 2:1 compression [ .... ] >V42bis theoretically manages a max of 4:1 compression Actually, both of these so-called theoretical maximum compression ratios are nothing but simple numbers the marketing types can remember. Both MNP5 and V.42bis (which, btw, really is normally the better of the two) are capable of *MUCH* higher than 2:1 or even 4:1. I have some detailed specs somewhere, but of course, they're not in the directory they should be in (and my copy of Recommendation V.42bis is at work). Of course, the numbers don't really mean much unless the data is basically designed to be highly compressible, but.... :-) Later, --jim -- 73 DE N5IAL (/4) < Running Linux 1.0.9 > jim@n5ial.mythical.com ICBM: 30.23N 86.32W || j.graham@ieee.org Packet: N5IAL@W4ZBB (Ft. Walton Beach, FL) E-mail me for information about KAMterm (host mode for Kantronics TNCs). ------------------------------ From: dale@gate.dungeon.com (Dale Elrod) Subject: Max size of SCSI HD? Date: 3 Sep 1994 16:35:55 GMT What is the max size that a single partition on a SCSI based hard drive? I would like to place two 9 gig drives on my ftp server but want to be sure that Linux can handle this. -- +-----------------------------+------------------+---------------------+ | Dale Elrod | dale@dungeon.com | 3 Hazel Close | | Dungeon Network Systems | +44-638-711550 | Mildenhall, Suffolk | | Internet Access for 10 quid | info@dungeon.com | IP28 7HU | +-----------------------------+------------------+---------------------+ ------------------------------ From: reutterer@wu-wien.ac.at (Thomas Reutterer) Subject: Experiences with Adaptec 2842 bootdisk Date: 3 Sep 1994 18:03:30 GMT I have been looking around for a bootdisk to set up Linux on a System with an Adaptec AHA 2842 SCSI-Adapter. Recently I«ve been told that one (still at the "interim-stage"?) is available at: sunsite.unc.edu:/pub/linux/kernel/images/boot284x.tar.gz Did anyone succeed to set up Linux with this bootdisk yet? Any information is appreciated. Thanks in advance thomas ------------------------------ From: joev@garden.WPI.EDU (Joseph W. Vigneau) Subject: Re: LJ#4? Date: 3 Sep 1994 17:56:16 GMT In article <1994Sep1.104456.7388@durie.wanganui.gen.nz>, Liam Greenwood wrote: >Joseph W. Vigneau (joev@garden.WPI.EDU) wrote: >> Has the Linux Journal #4 been distributed yet? I just received LJ#3 >> yesterday... > > I received the Linux Journal #5 today. I got my LJ 5 yesterday... Now I'm just waiting for #4... -- joev@wpi.edu, joev@hotblack.gweep.net WPI Computer Science Linux! Click Here! ------------------------------ From: broadley@turing.ucdavis.edu (Bill Broadley) Subject: Re: Promising Linux boxes in September 'Computer Shopper' Date: Fri, 2 Sep 1994 02:31:15 GMT Heres the most interesting quote I've seen for awhile: $2895 Comtrade p5-90 256k L2 cache (20ns) Flash bios, 4 pci, 4 ISA 16 MB ram 540 MB scsi. 15" monitor 1024x768 NI .28mm 1.44" floppy. 2 MB vram stealth 64 with the BT485 Ramdac PCI fast scsi-II ncr controller 2 serial 1 parallel, 1 game 101 keyboard 3 button 400 dpi mouse dos 6.2 win 3.11 with manuals mini tower 5 bay 200 watts energy star complaint $ 70 16 Bit ethernet $ 100 upgrade to 512 k L2 cache $ 550 upgrade from 15" to 17" MAG MX 17F .26" dot pitch (sony tube???) Others available much cheaper I.e. CTX for $320, mag DX $450 or so) $ 350 upgrade from 540 to 1080 MB connor 10ms 256k cache Total $3975 p5-90, 512k l2, 16 Mb ram, 1Gb scsi, 17" monitor with sony tube, all fairly linux compatible as far as I can tell. I've heard they use cheap keyboards/mice but I'd be happy to replace both to get a machine like this for that price. Comments? -- Bill Broadley Broadley@math.ucdavis.edu UCD Math Sys-Admin Linux is great. Bike to live, live to bike. PGP-ok ------------------------------ Crossposted-To: comp.unix.unixware From: kbrannen@csfb1.fir.fbc.com (Kevin Brannen) Subject: Re: WABI vs. SoftWindows? Reply-To: uunet!csfb1!kbrannen Date: Fri, 2 Sep 1994 16:51:51 GMT In article <346jqr$r1r@mis.nu.edu>, dmarner@mis.nu.edu (Dan Marner) writes: |> In article , |> Kevin Brannen wrote: |> >[With UnixWare] I also get complete and up-to-date man pages (this is |> >a real sore spot with me and gives me many negative feelings towards the FSF |> >despite the good utilities they produce). |> > |> Complete? Well, by some definitions. Novell ships UnixWare without |> man source, which renders commands such as 'apropos' not terribly |> useful. Note that 'apropos' DOES ship with the system, and, when |> asked, Novell says "Yup, it's broken alrighty, nope we aren't gonna |> fix it, why aren't you using the Fingertip Librarian?" Au contraire, el capitan... (sp?) There is hope, read on... |> Or did you mean complete as in "all commands have man pages?" Try |> "man dfspace." One of many I have encountered without a man page. Unfortunately, I'm at work and can't test. |> In general, I like UnixWare, I support a dozen or so systems |> running it. It definitely has its good points, but the man pages |> certainly aren't one of them. IMHO, of course. |> -- Dan, I understand your feelings, I had them several months back. However, I didn't want to wait for Novell to fix it and did it myself. I've posted my "How to make `man' work better" series here (c.u.u), but I'm always willing to sent it to whomever needs it (the reason I'm following up instead of replying). The series helps you to set up your environment to find all the man system pages, the Motif man pages (assumes you have the SDK), and create a whatis database so the apropos and man -k commands work. So, UW does provide it them, they just need a little tweaking. My comment (about Linux man pages) was directed to the comments in the gcc man page (about not bothering to keep them up to date) and the fact that some of my commands *definitely* do not have a man page (or perhaps I retrieved an incomplete Linux, but since 95+% of my commands have man pages, I suspect they just weren't included in the release). [ Note: follow-ups changed as this is becoming UW only. ] Kevin Brannen -- work: uunet!csfb1!kbrannen or csfb1!kbrannen@uunet.uu.net home: kbrannen@metronet.com ------------------------------ From: hugh@snafu.seada.com (Hugh Johnson) Subject: Re: vi for dos Date: 2 Sep 94 15:38:03 Reply-To: hugh@seada.com In article <778525462snz@panache.demon.co.uk> raph@panache.demon.co.uk (Raphael Mankin) writes: There are some very good vi clones on garbo and simtel. Have a look in the appropriate directories. I'm sorry I don't have an exact reference to hand. Also, try the ftp site of oak.oakland.edu, if memory serves. They have a slew of vi clones for DOS and such. -- | Opinions? I doan' have no steenkin' hugh@seada.com or | opinions. They doan' let me have 76317.2234@compuserve.com | none. ------------------------------ From: habaoch@eng.auburn.edu (Bao Chau Ha) Subject: Re: Greater than 1024 Cyls and Linux support? Date: Sat, 3 Sep 1994 17:52:27 GMT In article <343eqj$ohq@rigel.infinet.com> zureal@infinet.com (Jeffrey Oxenreider) writes: > >: At the same time I am planning on partitioning my 330 meg ESDI hard drive into >: MSDOS and a LINUX partitions. The drive has 1224 cylinders and I know that >: MSDOS and Smartdrive have problems with anything over 1024 so..... > >: Obvious question? Will LINUX recognize the partition if it's at the end of >: the ESDI drive and also will it have a problem with CYL > 1024? > >Should work. I don't know much about ESDI drives, but I've got a MAXTOR >345 meg SCSI that has over 12k+ cylinders on it, and DOS will only deal >with the 1st 1024 cyl, so I put DOS at the 1st 900 or so, and then my >swap partition and then the linux box at the end. No problems here. > What SCSI host adapter do you have? The Adaptec will remap the drive to a drive with only 345 meg., 1 meg for every cylinder. Just curious, and I still don't believe that LILO can load the kernel image from a partition which spans beyond 1024 cylinders. Bao ------------------------------ From: cgb102@psu.edu (Craig Bates) Crossposted-To: comp.os.linux.help,comp.os.linux.admin Subject: Re: Xconfig for Diamond SS24X ... Date: Sat, 3 Sep 1994 14:24:19 EST In article <345pio$fr6@charnel.ecst.CSUChico.EDU> hwang@ecst.csuchico.edu (Tony Wang) writes: >Path: news.cac.psu.edu!howland.reston.ans.net!europa.eng.gtefsd.com!MathWorks.Com!yeshua.marcam.com!charnel.ecst.csuchico.edu!hwang >From: hwang@ecst.csuchico.edu (Tony Wang) >Newsgroups: comp.os.linux.misc,comp.os.linux.help,comp.os.linux.admin >Subject: Xconfig for Diamond SS24X ... >Date: 1 Sep 1994 23:54:00 GMT >Organization: California State University, Chico >Lines: 22 >Message-ID: <345pio$fr6@charnel.ecst.CSUChico.EDU> >NNTP-Posting-Host: guzzler.ecst.csuchico.edu >Xref: news.cac.psu.edu comp.os.linux.misc:25709 comp.os.linux.help:55774 comp.os.linux.admin:14377 >Hello there: > I remembered someone posts an Xconfig file is specific for the >"Diamond SS24X", and it can display in 1024x768 mode by adjusting the >freq of the vga card. > I can not remembered where I got that file because it has been a while. >I tried the sunsite, but it seems not there. I can only find the generic one >for Diamond, and I can not make it to display in 1024x768 mode. > Can anyone who has this file send me a copy or tell me where I can find >this file? > Any help will be very much appreciated. >have a nice day, :) >Tony Wang >e-mail: hwang@ecst.csuchico.edu I would like a copy also please. I just can't get it to go in 1024x768 mode. Craig Bates cgb102@psu.edu Go Nittany Lions! ------------------------------ From: dplatt@3do.com (Dave Platt) Subject: Re: CAP for Linux? Date: Fri, 2 Sep 1994 22:15:22 GMT >I just downloaded the most recent CAP package and was disappointed >to see that Linux isn't one of the systems supported. Has such a >port been done, and if not is anyone planning to do one? It's available under a number of guises. All of them require hacks or extra hardware, or both. You can get a patch (at munnari.oz.au, I believe) which adds Linux to the list of supported CAP systems. It configures CAP to allow IPTalk operation on a Linux system - you'd need a FastPath-4, -3, GatorBox, or some similar IPtalk-to-EtherTalk router to handle protocol conversion. I was able to install CAP on my Linux system using a hacked-up version of UAR as the network interface. This UAR allows the CAP interfaces to be bound to the "tnnl" DDP-in-IP tunneling interface. I also set up UAR on my Sun SparcStation, acting as a router between the IP tunnel and EtherTalk Phase 2. It's an ugly hack but it does work. At least one group of people is working on adding native EtherTalk support to Linux. They've got Phase 1 partially working. Phase 2 (which is what most organizations are using these days) will be a more difficult capability to implement. It requires a consistent method of persuading the Ethernet interface to receive multicasts. A fair number of the current Linux ethercard drivers do not yet support multicasting, and [as far as I can tell] there is no consistent kernel interface for registering with the network in order to receive arbitrary packets (e.g. there's no equivalent to the Berkeley packet filter). -- Dave Platt dplatt@3do.com USNAIL: The 3DO Company, Systems Software group 600 Galveston Drive Redwood City, CA 94063 ------------------------------ From: tgm@netcom.com (Thomas G. McWilliams) Subject: Re: dram -- bogoboost without patching! Date: Sat, 3 Sep 1994 18:48:05 GMT Jerry Gaffke (jerryg@teleport.com) wrote: : You're almost certainly running these DRAMs far out of spec, on very : thin ice. Well, over 12 years of empirical evidence show that the bogoboost technique is safe and works fine. The 500 uS refresh timer is close to optimal on most machines. The margin of safety is considerable. You have achieved almost all of the boost possible at that point. You need to go to much, much longer delays before you experience parity errors: about 10000 or 20000 uS. But there is no point in pushing it to the threshold because you gain almost no performance increase over 500 uS. This technique is particularly well suited to older 386 motherboards. Owners of this hardware may well want to take a look at Paul Gortmaker's "dram" program and squeeze a bit more performance out of their older hardware. tgm@netcom.com ------------------------------ Crossposted-To: comp.os.linux.help From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: 16550AFN serial card Reply-To: pe1chl@rabo.nl Date: Sat, 3 Sep 1994 17:13:18 GMT In <347bmd$nrs@nermal.cs.uoguelph.ca> hpulley@uoguelph.ca (Harry C Pulley) writes: >Andre Fachat (fs1@aixterm1.urz.uni-heidelberg.de) wrote: >: Naji M. Khudairi (naji@myhost.subdomain.domain) wrote: >: : : >Does LINUX support any 16550AFN serial card? >: : Great, but how do you enable the 16 byte buffer in order to take advantage >: : of overrun protection? >: Another silly question >: Can I just remove the 8250 from an (very) old Serial Card and replace >: it with a 16550 to improve serial throughput? >You must change both the UART (8250) and the 2 support chips, unless the >support chips are newer than the UART. Wherever you buy a 16550 you usually >get the two little support chips too (1489 and 1488). The 3 chips are pin >compatible with your old setup. Just drop them in and all should be well. It >worked for me. There is no need to swap the 1488 and 1489, as they have no relation to the FIFO buffering. They are the drivers/receivers for the RS232 levels. Rob -- ========================================================================= | Rob Janssen | AMPRnet: rob@pe1chl.ampr.org | | e-mail: pe1chl@rabo.nl | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU | ========================================================================= ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Misc-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.misc) via: Internet: Linux-Misc@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Misc Digest ******************************