Re: can't find inetd.conf
At 03:21 PM 8/14/00 -0700, Nigel Osman wrote: Why i can't find inetd.conf in /etc after I installed RH6.2 as KDE workstation? How can I solve this problem? You should be able to. How are you looking? That is, may we seen the actual, exact command you enter and the actual, exact response you receive? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ethernet not working (part 3)
At 07:25 PM 8/14/00 -0400, Charles E. Gelm wrote: ... ln -s linux-2.2.16 linux tar -xzvvf /dosd/download/linux/linux_2_2_16_tar.gz cd linux make mrproper config ;no modules Actually your config file says you are supposed to be creating a lot of modules. But not the NIC drivers; that's true. However, you do need to pass an io= paramter to the NE2000 driver -- how are you doing that? (I don't actually recall how, when you don't use the ne.o module.) make dep clean zImage install reboot During boot-up I see something like: Configuring eth0 as 192.168.1.2 SIOCSIFADDR: No such device I think I've suggested before that you run "ifconfig -a" and tell us what output it produces. I wonder if all those boot-up messages are stored somewhere?!? Three places to look: 1. /var/log/messages . 2. dmesg 3. Backscroll on the console (SHIFT-PgUp), if Slackware 7 supports that. http://gelm.net/config.txt http://gelm.net/dmesg.txt "Document not found "The requested document /dmesg.txt was not found on this server. This could be for a variety of reasons, including ..." I didn't bother to try the rest of them, given this response. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: no inetd.conf
At 02:39 PM 8/14/00 -0700, Nigel Osman wrote: I installed REDHAT 6.2 recently and found I can't telnet to it. I have no problem to go out. I found I don't have inetd.conf in /etc.. or xinetd.conf.. What is the reason? How can I solve? Well ... you can't telnet because you don't have inetd running (and you don't have a standalone telnetd daemon running, but no one really does it that way any more). So ... why no inetd.conf? I recall you mentioning that you sis a "workstation" install of RH. I'm not up to date on RH instalation defaults, but there is no terribly good reason to run inetd on workstations, so maybe the default doesn't install it. It's probably in the "netbase" package (that's the relevent Debian package, and these names don''t cary all that much from one distribution to another). Perhaps someone with specific RH experience can help more, but checking on that is all I can suggest. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ethernet not working (part 2aa)
Charles -- I don't have ready access to your earlier postings about your problem, so please forgive me if some of this repeats things that I (or others) have already said. You are setting your kernel config file to do the NE2000 driver as a module: CONFIG_NE2000=m According to the config file you posted for us to lok at, it is NOT "compiled into the kernel". Similarly, you are setting it to compile NFS support as a module: CONFIG_NFS_FS=m nfsd starts because some rc script tells it to. Th fact that nfs.o is not available may mean it doesn't work properly ... I always get a bit mixed up about what NFS components are needed for the client, what ones for the server, though, so I may be wrong there. Can't help you specifically with the NLS codepages, since I don't know what they are. But the question you really need answered is: Why aren't the modules compiling? To explore it ... What process are you following when actually compiling the kernel? Are you doing the compile on the same machine that you use the kernel on? If not, how are you transferring the kernel and the associated modules from the compiling host to the target host? Might you be missing the step of transferring the modules? When compiling, are you going through the steps "make modules" and "make modules_install" as described in the README file in the root directory for the kernel (probably /usr/src/linux/README)? Do you have the kernel source tree (probably /usr/src/kernel-source-2.2.13) or something close to that) symlinked the way the Makefile expects (probably as /usr/src/linux)? This is what I can think of offhand. If none if this identifies the problem, I'd suggest you review the procedures you use when compiling, and any error (not warning - you'll get a lot of those even with a erfecrt compile) messages you are getting from the compiler. At 11:59 AM 8/13/00 -0400, Charles E. Gelm wrote: CMOS setup IRQ 10 is set to 'Legacy/ISA'. Status: Apparently the computer system is not an issue. Slackware 7.0 install, kernel 2.2.13. NIC=Macromate MN220PTC in NE2000 mode, io=0x2a0, irq=10, ISA Kernel make config: http://gelm.net/config.txt I don't know where to start. My ne2000 clone NIC is, now, 'compiled into the kernel' I think I need to tell the kernel what port it is using (0x2a0). What HOWTO do I read to figure out how to get my kernel to recognize my NIC ? OBTW, find / -name *.o returns matches to only dummy.o in 4 places. There are no other .o files anywhere in this computer. Why, during bootup, do I see several modprobe ... can't locate module nls_*whatever ? I answered 'y' to only one codepage (437). Are these responses saved to a file? I notice that, upon boot, that nfsd is started. Why? Why does not nfs.o exist (or whatever.o) ? -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
RE: Accessing CR-RW
What your options are depends on how securely the system was set up in the first place. I'll skip several intermediate methods, since competent security will make them not work, to go right to the reliable one -- you need a rescue disk. Many distributions have boot disks that also serve as rescue disks. You don't say which Linux distribution you are using, so I can't be more specific than to suggest you track it down and see what it has to offer. Also, some general-purpose rescue disks -- tomsrtbt is pretty much king of this hill, but there are others -- also exist. Take a look at metalab's archive, in one of these spots: ftp://metalab.unc.edu/pub/Linux/distributions/ ftp://metalab.unc.edu/pub/Linux/system/recovery/ At 04:36 PM 8/13/00 -0700, Dan Bentson-Royal wrote: I have been enjoying the use of my Linux box that a student setup. That student is now gone and the root password is no longer being accepted. What's my next step? I can't seem to locate any kind of emergency boot disk either! -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Accessing CR-RW
At 07:14 PM 8/13/00 -0500, Jim Reimer wrote: assuming lilo is in use. At the "Lilo boot:" prompt, enter "linux single" instead of just "linux" (all without the quotes, of course). That will put you straight in as root, no questions asked, after which you can use "passwd root" to change the root password. Not always "no questions asked". This choice in lilo CAN be password protected. All depends on how lilo is configured. Same thing goes for the other standard trick, entering "linux init=/bin/bash". -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: partition problem
I find your question a bit hard to follow, so I apologize if this answer is not responsive. To use "logical" partitions (the ones with numbers above 4), you need to make one of the first 4 "real" partitions an "extended" partition. It in turn holds the logical partitions. Assuming you are using fdisk, choose "e" for one of the partitions 1-4. At 03:02 PM 8/8/00 +0530, gaurav rajput wrote: hi, well i am new to linux ,and having problem right at the very first step of partition. HDD 10.2 Gb with following partition : hda1,hda5,hda6 - Dos,extended Dos,extended Dos when i go in for partition (by using suse-linux 6.4 CD) /,/swap,/tmp,/var,/usr : it takes at the maximum hda2,hda3,hda4 only. on giving further partition it says no primary partition available. while the only Primary partition ,that i saw using fdisk is Dos(hda1) only. while hda5,hda6 is extended Dos. [html deleted] -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: telnet Shell Script
Plain shell scripts aren't good at controlling dialogs over socket connections. I hesitate to say they cannot do it, since there are people here more skilled in shell scripting than I, but doing it directly won't be trivial. I'd do it one of these ways: 1. Write it in perl, which has a Telnet module 2. Use the "Expect" package from a shell script; this manages telnet-like dialogs. At 11:02 AM 8/11/00 -0700, Gregory D. Burns wrote: Hi, Does anyone know how I could make a script that telnet into a router did a some commands, log the results, and then moved to the next router. I'm trying to write a little shell script that will add a ip address to a access list in 5 different routers. If you have any suggestion please let me know. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Plug-and-Play trouble
At 11:57 AM 8/10/00 EDT, Aaron Brown wrote: Extreme newbie here. I'm using Definute Linux 7.0, with kernel version 2.2.12-10. I have two plug-and-play cards, a soundcard (which we'll ignore for now, as I don't urgently need to use it), and a Motorola VoiceSURFR 56K modem, which I do want to use. So far I haven't gotten the modem to do a single thing. (I'm writing this on someone else's computer, of course.) I can't find a "Motorola VoiceSURFR" listed in the database at http://www.o2.net/~gromitkc/2809a.html so let me ask if it is an isa-bus or a pci-bus modem. If the second, it is *probably* a Winmodem and will not work with Linux. If the first, then describe in more detail what you are trying with isapnp. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Can Linux be installed over a LAN?
John -- How specific is your question here? 1. Can Linux be installed over a LAN? Yes. Easily. I've done both Slackware and Debian installs that way for years. Don't actually know about Caldera or Red Hat. 2. Can Linux be installed over a WAN? Yes. Easily. Over the Internet, via ftp, is one of the standard ways to install Debian, and I did it many times before we set up our Debian mirror here. Again, I don't know abotu Caldera or RH specifcially. 3. Can Linux be installed "on top of" Open Server? What does this mean exactly? Linux would replace Open Server, requiring new filesystems and replacement of all apps. The usual way to do a "network" install Linux is to use one, or a few, boot disks (or a CD) to set up the drives and install a small "base" system. This base system then handles the rest of the install. I'd be surprised if any distribution had a system lying around to install itself on a system already running SCO, but it might be possible to adapt some existing system if needed -- certainly "several thousand" cutovers justifies a bit of development effort. 4. Can Linux be installed remotely? I suspect this may be what you want to know. I hesitate to say NO flat out ... but if it is possible, it would be difficult. Somewhere in the process, you have to reboot the system to switch from the SCO kernel to the Linux kernel ... and a reboot interrupts any remote connection. If there is any problem with configuration of the new system that reboots (wrong NIC driver, a small error in interface or routing setup, etc.), there would be no way to get back on the system remotely to fix it. Probably your best bet would be to create a custom bootdisk (or CD-ROM) that could be run at each site to boot the machine with a temporary filesystem, repartion the hard disk, and do a network install. Again, this is definitely doable in Slackware or Debian (albeit with some work), and probably could be done with Caldera or Red Hat too. This all assumes the hardware involved isn't too weird or limited. With sufficiently limited systems, doing any install at all requires inventiveness, and network installs wouldn't be possible. I'm assuming at least 16 megs of RAM, mainly. At 04:19 PM 8/10/00 -0700, Baskette, John wrote: Hello, The place where I am currently working is contemplating converting several thousand stores that have back office systems running on SCO Open Server 5.0.5 over to some version of Linux. Probably Red Hat or Caldera. This is only preliminary discussion at this point. But someone asked me if Linux could be installed over a LAN (WAN?) on top of Open Server. Later this year, this company will be putting DSL into all their stores. -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: windows download
Yes. Just make sure that whatever download method you use -- ftp or http, most likely -- downloads the .rpms as binary files, not as text. (This can be a problem with some misconfigured Web servers.) But how do you plan to "take it" to the Linux machine? Many .rpm packages are too large to fit on a floppy. At 09:34 PM 8/10/00 -0500, lewis kimble wrote: a friend if mine has a cable modem. can i download linux software under the windows platform and then take it to my linux machine (red hat 6.1) and load the software into it thank you -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: WU-FTPD RedHat 6.2
At 10:51 AM 8/8/00 -0500, Jim Roland wrote: RedHat, just like any other distribution is a collection of tools and software, along with some customizations to scripts, etc. You can always download the latest version in rpm format and install it, just make sure to read the README or INSTALL files included with any of the rpms or tar files, to make sure you don't need extra stuff. RedHat is fairly stable, so it's doubtful you will need anything extra, but always read the READMEs. Jim -- This isn't *quite* correct. Like any (competently done) Linux distribution, RH regularly releases updates to specific packages that reflect, mostly, security fixes. Just downloading the "latest version" won't get these; you have do download them from the "updates" directory. For example, RH 6.2 provides the 2.6.0-3 version of wu-ftpd in the distribution directory, such as ftp://metalab.unc.edu/pub/Linux/distributions/redhat/redhat-6.2/i386/RedHat/ RPMS/ OTOH, it provides the 2.6.0-14.6 update of wu-ftpd (dated June 23) in the "updates" directory, such as ftp://metalab.unc.edu/pub/Linux/distributions/redhat/updates/6.2/i386/ Going back to Jesse's original question, it is clear that he has been doing his homework here, since he has the current RH update version. And I too find that a 2.6.1 update isn't yet available from RH. Nor do I know the answer to his actual question, whether the 2.6.0-14.6 version is vulnerable to the latest wu-ftpd bug. If the wu-ftld site doesn't tell you, I'd try to find the bugtraq archives and search them. -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: 2 very simple questions ...
At 07:46 AM 8/9/00 +0530, Webmaster - prayagonline.com wrote: hi, i have 2 ery simple questions : 1.) I need to issure the commands halt / reboot from a telnet connection logged on as a different user ... so what to do ?? If you are accessing the machine remotely, you have to be root to do this. (Do you *really* want anyone who can log into your system able to reboot or halt it? I don't.) So the only option is to su to root then execute the commands. 2.) I want to change the message that appears befor the login prompt ... i changes the /etc/issue and /etc/issue.net but ti guess that got overwritten when i rebooted .. Yes. Or at least probably yes. grep through the init scripts for the ones that write to issue and issue.net (varies by distribution) and comment out the troublesome lines. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: /sbin/lilo - segmentation fault
At 12:06 AM 8/8/00 EDT, [EMAIL PROTECTED] wrote: I don't know how to make lilo segment fault, so I don't know how to fix it. Could you be dead flat out of space in /boot? I would expect that to get error 28, but maybe not. df to see. I wouldn't expect a segfault from that problem either. More likely prospects are: 1. A libraries problem (lilo uses libc6, so there's always potential for a glibc2.0 vs 2.1 problem). 2. Bad memory. Subtle weaknesses in memory produce a lot of weird symptoms, including segfaults. Never seen it with lilo specifically, but why not? That said, I've never seen lilo segfault either. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ethernet not working
At 11:35 AM 8/7/00 -0400, Chuck Gelm wrote: .. A thought to the module programmers: I wonder why, if ne.o needs 8390.o, it doesn't just go do it its own self! Modules don't know anything about installing other modules (if that's what you mean by "just go do it"). Many different NICs use the 8390 chipset, so pulling this functionality out into a separate module does make sense. ne.o is hardly unique in this approach. The "modprobe" application takes care of catching this sort of dependency. IT works with a list generated by the application "depmod". You should use it instead of "insmod". -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail/minicom
At 06:26 AM 8/7/00 +, Richard Spencer wrote: Is it possible I have the same problem? I've just assumed that the ISP was busy, or something? It's hard to tell from these logs, but I'd read them as saying your end tries to use PAP to authenticate itself and the ISP end doesn't like what you send. Either the ISP doesn't use PAP, or your userid or password is wrong (is your password really "secret"?! or did you edit the password field here [as you should] without indicating clearly what you did [which you also should do]?). Does the connection always fail or just sometimes? If sometimes, how do the logs of a successful connection contrast with the failure samples you sent? ... Aug 7 05:14:12 localhost pppd[770]: sent [PAP AuthReq id=0xa user="[EMAIL PROTECTED]" password="secret"] Aug 7 05:14:15 localhost pppd[770]: No response to PAP authenticate-requests Aug 7 05:14:18 localhost pppd[770]: rcvd [PAP AuthNak id=0xa "Invalid Login"] Aug 7 05:14:29 localhost pppd[770]: Terminating on signal 15. Aug 7 05:14:29 localhost pppd[770]: sent [LCP TermReq id=0x2 "User request"] ... -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ethernet not working
At 12:21 PM 8/7/00 -0400, Chuck Gelm wrote: Hi, Ray: Thanks for the hints. :-) Yes, that is what I meant. Then following your suggestion: Do you mean that if I; /sbin/depmod -a /sbin/modprobe ne that 8390.o ne.o will be installed automagically? Yes. Well ... it isn't *really* magic ... just a "sufficiently advanced technology." -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: How to change resolution
About 2/3 of the way through the xf86config interaction, there is a spot where it asks you if you want to change the order of the resolutions. If you do so for the 8-bit ones, you should be able to get the system to default to 800*600. I think you do it by specifcying an order like 324, but that's from memory so don't trust it as exact. Once X is running, you can change resolutions with CTRL-ALT-(numeric)+ and CTRL-ALT-(numeric)- . But your problem *may* that your video card, or our screen hsync and vsync. do not support 800*600. If that is the case, then no improved configuration tool will help. Without details, I naturally can't say if this is your problem, only that it might be. Now, GUI tools. There are two. XF86Setup and Xconfigurator (think I have the capitalization right on them). If you get "command not found", it probably means the tool wasn't installed. (It could mean that it is on the system but not in your PATH; in this case, use the "find" command to track it down and run it by specifying its full path.) Some distributions will install one or the other in a "normal" installation, but others won't. Since you don't mention what you are running, I can't even guess. Check the CD for whatever you are running and install the apropriate packages (for example, in Debian there is xf86setup; the Debian package manager will take care of its many dependencies; I can't find a .deb for Xconfigurator). Personally, I don't like these tools, but preferences vary. At 04:25 AM 8/6/00 -0700, Steven Ackerman wrote: I've been having problems changing my resolution from 640x480 to 800x600. I tried running xf86config and when I do it sets my resolution to 640x480. Someone told me there's a gui deal to do it called xconfigurator. When I try running that from a command line as root I get "command not found". ... -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Module problem: can't load tulip.o device busy.
Please post again with a more exact description of what you are doing. I use the Linksys LNE100TX NIC here all the time and have no trouble with the tulip driver provided with either Debian Potato or LRP distributions. I doubt RH is shipping an outdated tulip.o. So please clarify the following bits: I've tried building it into the kernel, no luck. What does "no luck" mean? What error do you get during compilation? Or during boot/init? After boot/init, what does "ifconfig -a" say? When I do an insmod, it says the device or resource is in use. Please quote the exact commend you enter and the exact response. What you wrote is similar to some messages I get from insmod'ing but I can't manage an exact match. Is it possible that you are insmod'ing with a kernel that has the driver compiled in? The card is using IRQ 11, IO=0xF400, ( cat /proc/devices). Are you sure of this? On my systems, /proc/devices doesn't contain any information about IRQ or IO. I need to consult /proc/interrupts, /proc/ioports, or /proc/pci (depending on what I'm after). Are RH systems that different from Debian systems? At 10:05 AM 8/6/00 -0400, Dances with Turtles wrote: I'm no guru at this module thing. I usually compile everything into the kernel, but I am trying to install a Linksys fastether 10/100 which is suppose to use the tulip driver (RedHat 6.2). My previous network card (linksys eth16) wasn't a problem to get working. I've tried building it into the kernel, no luck. So, I'm trying it as a module. When I do an insmod, it says the device or resource is in use. The card is using IRQ 11, IO=0xF400, ( cat /proc/devices). Nothing shows up on /proc/interrupts. Any help in how to debug this would be appreciated. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Repost of: module problem: can't load tulip.o with linksys 10/100lnetx V4.1 card
At 12:03 PM 8/6/00 -0400, Dances with Turtles wrote: I can not get my new linksys 10/100LNETX card to be recognized by Linux. It works fine under windows. Let me also state that my prior Linksys ether16 ISA card was working perfectly prior to my 'upgrade.' [details deleted] Well ... your more detailed post makes it clear why you're finding this process frustrating. You've pretty much tried almost everything I'd say to check. I'm left with 3 thoughts ... 1. You say the card works under Windows but not under Linux. Does this mean you have a dual-boot host? If so, the BIOS may be set wrong for Linux -- it should say NO for the Plug-and-Play OS question. 2. Have you tried the card in more than 1 slot? In my experience (not just with Linksys), some card/slot combinations work, others don't. 3. You probably did this right, but just in case ... you are using a a tulip module from the same compile as the kernel, right? Not just the same kernel humber -- the same actual compile. These suggestions are ... to use your own word ... "pathetic". Afraid it's all I can think of, though. Getting the wrong manufacturer name in the lspci output is a commonplace problem and indicates nothing of consequence. My Lynksys shows up as a "Lite-On", for example. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ethernet not working
Charles -- Some of what you post here is too vague to troubleshoot. I'll answer what I can. You write: If I: insmod ne /lib/modules/2.2.13/net/ne.o: unresolved symbol ei_open " ethdev_init " and about 4 more lines ne.o requires that 8390.o be installed first. So either do that with insmod or use modprobe to install ne.o (it takescare of dependencies for you). Now, you also write: I've compiled a kernel with lots of modules except my boot media. I have 'make dep clean zImage zdisk modules modules_install' seemingly withour error. I see lots of 'unresolved sysbols' when booting. It would help if *I* could see them too. But it sounds like you never actually install the zImage you create; at least you don't mention doing so. But the modules_install step may be causing the new modules to overwrite the old ones (depends on what kernel version you are using). Either install the new kernel hy hand or (if your kernel Makefile supports it) do a "make install" step. Be sure to run lilo after you do this. You write: If I: ifconfig eth0 eth0: error fetching interface information: Device not found. Probably because the module fails, so eth0 doesn't exist. You write: I did something somewhere while trying to get this NIC to work with isapnp, cause when the kernel boots I see this message: "don't know what to do with IRQ 10 on or around line 22". I wonder what file I edited to cause this. H. Me too. You aren't asking us to tell you what file you edited, are you? Without context (where in the boot/init dsequence is this? what precedes and follows it?) it's just a guessing game. Anyway, if you've assigned the card a fixed io and irq with its setup probram, you shouldn't be messing about with isapnp. You write: alias ne io=0x2a0 irq=10 #-- added line. ? that's a decimal ten right? Yes. You can enter the IRQ in hex, but you need to use the 0x prefix. so irq=0xA would also be acceptable. (As would nothing; ne.o requires the io= paramtere but not the irq= one.) At 03:37 PM 8/6/00 -0400, Charles E. Gelm wrote: Howdy, Y'all: I've been trying to get networking running on my Slackware v7.0.0 install on my k6-2-266, 96 Megabyte RAM; dual booting with windows98. ... -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Simple Serial port question
/proc/interrupts does NOT list all supported interrupts. It just lists the ones that have actually been used. If the lack of a listing here is the ONLY reason why you think your Linux kernel is not supporting the second serial port, then you are probably mistaken. If you have some other reason for believing that ... like you actually tried to access the unidentified "piece of hardware" on the second serial port and failed ... please describe in more detail what the hardware is, how you tried to access it, and what went wrong. Maybe we can help, but not if we don't know what the actual problem is. In any case, a linx kernel may or may not have serial-device support installed (either compiled in or insmod'ed). But it isn't selective. At 11:30 PM 8/7/00 -0400, Peter Howell wrote: My question is simply how do I get it to work. Here what info I have. I recently reinstalled Caldera Openlinux 2.3. I have a mouse on the first comm port (irq4) and that works just fine. Unfortunately, support for the second comm port doesn't appear to have been installed. It's not a conflict, since /proc/interrupts does not have a live for irq3 and when I check the IO-ports, 02f8-02ff and 03f8-03ff are both allocated to serial. I have a piece of hardware that is run through the second comm port. Unfortunately, I don't know which configuration file(s) I should edit to make it available. Any suggestions? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail/minicom
At 01:25 PM 8/6/00 EDT, [EMAIL PROTECTED] wrote: ... Aug 6 11:25:29 debian pppd[274]: rcvd [LCP ConfRej id=0x1 auth pap] Looks like the remote pppd is satisfied with the shell login and doesn't want to be bothered with PAP. Maybe you can try noauth in /etc/ppp/options. Good catch, Lawson. I'd missed that he had "auth" in that file AND didn't override it in the command-line options that invoke pppd (the usual workaround). But I wanted to follow up because the way you wrote your response, David (and others) might be misled about the meaning of "auth". From the pppd man page: auth Require the peer to authenticate itself before allowing network packets to be sent or received. This option is the default if the system has a default route. If neither this option nor the noauth option is specified, pppd will only allow the peer to use IP addresses to which the system does not already have a route. In other (perhaps clearer?) words, the "auth" option requires the *other* end to authenticate itself to *you*. I've never found an ISP implementation of ppp that does this; it's normally used with the so-called "server" end of a ppp connection (the end that answers the phone), not the so-called "client" end (the end that makes the phone call). From the frequency with which this comes up, I'd guess that "auth" is the most misunderstood option in pppd. -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail/minicom
Good. You've found the answer. Your chatscript reported using "davidturetsky" as userid, but the ISP expects "MSN/davidturetsky". Change the chstscript and see if pppd (however you invoke it - I don't have the backfile of your messages handy) now works correctly. At 12:18 AM 8/6/00 -0700, davidturetsky wrote: I ran minicom: ATDT410-727-0315 cr in response to CONNECT responded to login with MSN/davidturetsky (plain davidturetsky was rejected) responded to password with password Got: Entering PPP Session. IP address is 63.24.126.61 MTU is 1524 [rest deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: EDO vs non-EDO memory
No. And yes, this is a problem -- FPM memory is hard to find these days, and distressingly expensive when you do find it. You might try non-local shopping, at places like www.computergeeks.com (there are others, but not that are at the tip of my tongue), where used memory is sometimes available. At 09:41 AM 8/5/00 -0500, Jim Reimer wrote: Does, or should, EDO memory work on a non-EDO motherboard? (Desperate for more memory, and EDO is all I can get locally.) -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail
minicom wouldn't have a problem with name resolution, since it never does any. As to the problem you are experiencing with fetchmail ... how does your workstation make a ppp connection to the Internet? It sounds like it is not set up to autodial, resulting in the nameserver addresses in /etc/resolv.conf being unreachable (unless you started ppp menually before doing this). If you want real advice, though, you're going to have to provide a more detailed description of your setup. At 03:19 AM 8/5/00 -0700, davidturetsky wrote: I just downloaded fetchmail from the Debian site and installed it when I run: fetchmail -v -k -p POP3 -u MSN/davidturetsky email.msn.com I get: fetchmail: 4.6.4 querying email.msn.com (protocol POP3) fetchmail: POP3 connection to email.msn.com failed: temporary name server error fetchmail: Query status=2 fetchmail: normal termination, status 2 /etc/resolv.conf has the correct nameserver entries (minicom had no problem) No dial tone was heard [html duplicate deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail
David -- First of all, let's do one thing at a time. fetchmail won't work (nor will DNS, or much of anything else Internet related)) unless you can establish a PPP coonection. In a prior message you said ... With the change to ttyS2 and irq 10 I has successful in getting dialtone and connecting to msn ... so I assumed (as did others, no doubt) that you had PPP working. SO ... 1. IF you want help with PPP, tell us your actual setup. Telling us you "ran" pppconfig doesn't give us a clue if you did it right or not. Be less terse. Let us see the script that pppconfig creates, or tell us how you answered each question it asks. 2. IF you want help with fetchmail, start your PPP connection by hand (which you said you could do), try fetchmail then, and if it doesn't work, tell us how. Include the output of "ifconfig -a" and "route -n" (all of it, not just a summary of twhat you see as important) with any such posting. But don't muddle the two problems together. At 01:52 PM 8/5/00 -0700, davidturetsky wrote: How do I set up my Linux system to autodial? I'd be happy to provide any description that would be helpful. What information would be useful As for ppp, I ran pppconfig. As I previously posted, the related scripts to test it did not reach dialtone. I have been trying to explore all this with posts to the list, trying different workarounds and buying a slew of O'Reilly texts ... -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: installing sendmail
At 01:58 PM 8/5/00 -0700, davidturetsky wrote: I too have been trying to grope my way through sendmail. Yes it was installed with my system at /usr/sbin, but there is no file .fetchmailrc or sendmail.cf I wonder if I shouldn't download a fresh copy and (re)install it Bet you're wrong. You run Debian, yes? The default mailer for Debian is exim, not sendmail. It sets up a symlink "sendmail" that points to exim (any MTA does that, because way too many applications assume that name for the system's MTA to let them do anything else). Since fetchmail is a separate application, neither exim nor genuine sendmail would create any .fetchmailrc files. Try installing fetchmail.*.deb and fetchmailconf.*.deb for this. (I assume you know how to use apt-get by now.) Instead of sendmail.cf, you want to look for /etc/exim.conf (or, easier, run /usr/sbin/eximconfig). David [old stuff deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail
OK. All of your questions come back to your being unable to get a PPP connection established. So let me ask you what instructions you have from your ISP about how to start PPP. Since we're talking about MSN, the natural way to phrase it is: how to you do it under Windows? Or if you can't tell, try connecting with minicom and seeing what gets exchanged before the actual ppp connection (which, in a terminal window, looks like line noise) begins. In one place, you describe setting pppd up to make a pap-authenticated connection. There are at least 2 other common methods: chap and userid/password. Or at least I think you are; to be sure, I'd need to see the chat messages as well as the pppd ones. Especially since the chatscript seems to be doing userid/password authentication. You want to make sure that the exchange here is accurate. I notice you expect assword, not assword: -- that might be enough of an error to cause your problems (or it may just be a typo in the email you sent here). And to emphasise for other readers: this is an attempt to connecto to msn.com . There may be special problems using something other than Windows with MSN, and possibly someone else knows the specifics. Other odds and ends: apt-get (and its cousin apt-cache) assume that you can make an http or ftp connection to the server(s) listed in /etc/apt/sources.list . Until you can, the apt-get method won't work for you. Similarly, the smarthost setting of exim won't work until you can make a ppp connection to the service. Even then, I doubt "msn" is a sufficient identifier for the smarthost ... it needs to be a resolvable FQDN, probably something like smtp.msn.com (but this too they should have told you). Except for this, your setting slook fine. Don't waste your own or my time with "assurances" from an old ORA book. Practices change faster than books do, and I wouldn't be surprised if either pppd or chat doesn't have the setuid bit set. If not, you will need to be root to run pppd. Also don't waste your time running fetchmail yet. You have to solve the ppp connection problem first. Everything else depends on it. At 07:34 PM 8/5/00 -0700, davidturetsky wrote: Thanks, Ray, for all of the feedback. I've been trying to explore all of the questions you raise. Let me respond to each (I hope) and list the scripts involved ... With apologies for the length of this post. None needed. Much better to provide the information needed to describe your problem then to ask us to play guessing games. This is a model of how to do it right (except for the tiny slip in leaving out the messages from chat). -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: broken C++ compiler
The C++ compiler you want is probably g++, and it's probably not installed (check with the "find" or "which" command). I don't recall if the gcc/g++ mess was fixed by the time RH 6.0 came out ... it is pretty out of date, after all ... older versions of g++ simply were not decent C++ compilers, and the suggestion that gcc would act like a C++ compiler if the file's extension was .cpp, or something like that, was pure fantasy. Check thr RH site, or a mirror like metalab, for updates. The current version (in Debian) is 2.95.2-13, and it works pretty well. At 10:03 PM 8/5/00 +, Richard Spencer wrote: Hello gurus: I just wondered why I cannot install tarballs with ./configure make and make install; I get error messages as follows: snip checking for gcc... gcc checking whether the C compiler (gcc ) works... yes checking whether the C compiler (gcc ) is a cross-compiler... no checking whether we are using GNU C... yes checking whether gcc accepts -g... yes checking for c++... no checking for g++... no checking for gcc... gcc checking whether the C++ compiler (gcc ) works... no configure: error: installation or configuration problem: C++ compiler cannot create executables. I'm running rh6, mostly gui-free, and totally windoze-free rh let me leave something vital out, didn't it? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: fetchmail PROGRESS
OK. Good report. Nice and clear. Here's what is significant: the other end is not responding to your ppp initiation *after* authentication. You see that here in the chat messages: Aug 5 21:24:33 debian pppd[442]: Connect: ppp0 -- /dev/modem Aug 5 21:24:50 debian pppd[442]: peer refused to authenticate Aug 5 21:24:50 debian pppd[442]: Connection terminated. Everything else that goes wrong is the result of this failure, so focus on this part and fix it. So ... what's wrong? I can only guess that the particular userid and password you are entering do not start a PPP session. You need to investigate this. Here's how. Start minicom. Use it to dial the number and enter the userid and password. After the password gets accepted, see what happens. Possibly you'll be at a shell prompt, from which you'll need to run ppp manually. Possibly it's the wrong account and you need to do a 3-step login -- MSN, then userid, then password. Possibly the password won't be accepted (maybe your userid really is MSN/davidturetsky, like the docs you quote say). Possibly it's weirder than that. But you need to test possibilities. You do not have a Linux problem here; you have a problem of not knowing what your ISP wants you to do to start a PPP session. It might help if you tell us *exactly* how the Windows host is set up to initiate PPP. Assuming you're using the standard TCP stack, what is in the userid field of its "properties"? At 10:11 PM 8/5/00 -0700, davidturetsky wrote: Progress to report! In looking around further, I noticed in /etc/ppp/connect-errors: sh: my-chat-script: command not found ... -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: accessinf 2nd modem
Before we spend any time on the details here, you should make sure that your 56K modem works with Linux. Many do not (so-called "winmodems"). Look your make and model up at the canonical modems-and-Linux site: http://www.o2.net/~gromitkc/winmodem.html If it is listed there as working with Linux, please post again, including the modem brand and, under Windows, what IRQ and IO base port it uses. At 11:41 PM 8/3/00 -0700, mike mcmanus wrote: Hi, I have just been able to connect to my isp thanks to your helpful suggestions. I am on a compaq 486 with an original modem 14400. I added a 56k modem some time ago and have been using it for internet connections with win95. With Linux I can only get a connection with the 14400. The 14400 is on com2 .the 56k on com4. i've used control panel in linux xwindows to set the /dev/modem to cua3 for the 56k and used minicom -s to set speed; after saving configuration as df1 and exiting, the screen shows 'initializing modem' and then hangs like it can't find the modem. ican change back to 14400 modem and reset /dev/modem to cua1, change speed and everything works fine. Obviously I would like to use the faster modem. Any ideas on how to get the fast modem as the default. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: new drive - partitions sizes?
Try "du -sh /someting/other" At 05:31 PM 8/4/00 -0400, Renaud OLGIATI wrote: As an aside to this thread, is there a simple way to find the size of a given branch of the file-tree ? Or in other words, to find how big is /someting/other including all the subdirectories ? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: same old problem - part 2
This is obviously a hardware problem, but the question is where. The likely suspects are the IDE controller on the motherboard, the cable connecting the drive to the mobo, and the power supply. The "sometimes" nature of the problem makes the power supply an attractive guess ... I've had similar problems in the past, due to a weak connector. With the case open, this is easy to spot, as you can hear that the drive is not spinning. BTW, the POST test for memory is not an adequate test. I've had many DIMMs fail in actual use when they pass the POST test (and memtest). The best ways i've found to test memory are by running: a large ftp transfer, of a file that's 100 megs or so compiling something big; a kernel usually will do. I doubt that your problem is memory, though. At 09:44 AM 8/4/00 -0500, Jim Reimer wrote: After 30 minutes or so of "drive not ready" messages, the thing finally did a successful reset (without me touching anything) and finished booting into the install program. H (original message follows) The computer has a new drive, and I've swapped the memory (was 16meg, now 32meg). Condition of memory is unknown - came out of an old machine - but it passes the POST ok. RH6.2 installed ok, and I left it sitting at the login prompt. A while later, I had the "drive not ready for command" messages scrolling up the screen again. Computer would not respond to CTRL-ALT-DEL. Pressed reset, and the computer wouldn't boot (just like with the old drive). Cycled power, it started booting, but now I have a corrupted file system again. It's been turned off all night, just turned it on, and it won't boot at all from the hard drive. Boots ok from floppy. Booting from the floppy (with the install disk) goes ok until it starts checking for drives, then I get "hda: lost interrupt" over and over and over... P90, Award BIOS v4.50G, on-board IDE floppy controllers. Everything related to APM in the BIOS is either 'off' or 'disabled'. Where do I go now? -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: accessinf 2nd modem
I added the list back in to this exchange. At 09:47 AM 8/4/00 -0700, mike mcmanus wrote: Ray, The 56k is a 'creative modem blaster'(DI5660 1) . I did find it in the list you pointed to. It is isa and in the list it is listed as a pnp with no jumpers. In windows it is at com 4, irq 11, I/O 02E8-02EF and 0908-090F. OK. This makes it straightforward. Your problem is that the default setting for Linux's equivalent of COM4, /dev/ttyS3, is IRQ 3. So, to get things working, try using these three commands: setserial /dev/ttyS3 irq 11 rm /dev/modem ln -s /dev/ttyS3 /dev/modem Now see if minicom connects to the modem. Make sure ppp is using ttyS3 and it too should work (at least to the point of dialing, but if your 14.4 connection worked, you presumably have the rest figured out). The last loose end is with PnP configuration. If you're lucky, the modem will still be sitting at IRQ 11 when you boot Linux. If you aren't, you'll need to use the "isapnp" package to set it up. I don't have experience with this app, but others on the list do, and I'm sure someone else will respond to a problem report if you post one. (Be sure to read the appropriate HowTo first, of course.) -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Caldera Openlinux/GRUB
Well .. I've used the Debian grub package and found it pretty easy to install. Don't have it running at the moment, so I can only write from memory ... after installing the package, basically you copy a few files to /boot/grub, set up a simple config file, and run an installer (like you run lilo). Then at boottime you get a menu of available kernels for a 30-second countdown. Worked fine for me. But I don't know the details of how other distributions package grub (or even which ones do). But it might be worth your while to take a moment to tell us about your 1024-cylinder problem. It might be fixable, permitting you to stay with lilo. The usual solution is to create a small (20-50 megs) partition as the HD's *first* partition, then mount it as /boot . This forces the stuff in /boot, all the kernels and related stuff, to be within the first 1024 cylinders. At 01:03 AM 8/5/00 -0400, David Aikema wrote: I have been having problems getting Turbolinux and Corel Linux to work due the 1024 cylinder limit of lilo. On the Caldera site it talks about grub being included with Openlinux 2.4 to get around this limitation. However the user guide available on the site only makes mention of lilo. I am debating whether or not to purchase this distro but there seems to be a strange lack of documentation regarding it. How do I get grub running instead of the lilo the user guide talks about? I'm just hoping to finally get linux booting on the machine. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Caldera Openlinux/GRUB
At 01:24 AM 8/5/00 -0400, David Aikema wrote: ... I don't want to have to wipe the partition table (too much data on the drive and I haven't got a good place to backup to). I can pick up a copy of Openlinux for a whole lot cheaper the full partition magic so I figured I might as well give the alternate bootloader a shot. Well, you can give it "a shot" without having to buy OpenLinux. Just download the .tgz file from (for example) metalab. It's in ftp://metalab.unc.edu/pub/Linux/system/boot/loaders/ Good luck. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Linux Distro's
This question gets asked regularly enough that I think we're all too bored with it to start a "firestorm of debate". Generally, you'll do okay with any well-known brand of full-size Linux. When distribution-specific questions come up, the regulars on this list seem to be, as a group, most familiar with Red Hat, Slackware, and Debian. My best guess is that those three, plus SuSE and Mandrake, are the ones with the largest installed bases overall. As a group, we seem to know less of the specifics of the more commercial distributions, like Caldera and Corel, and the obscure ones, like Stampede. For some situations, one of the various "small linux" variants (ranging from 1-floppy ones like Linux Router Project up to Zipslack, a 100 mB file) have advantages. But here you need to know more of what you're doing and have a specific purpose in mind. Personally, I use and like Debian. I've used Slackware in the past and found it okay. I've used and disliked Red Hat, which I find hard to customize and (for the free version) too much work to keep updated with the security patches. But thses aren't strong opinions, just what I actually do. Were I trying to evaluate the relative merits rigorously, I'd probably try first to assess the timeliness of the security updates for the various discributions. But I haven't tried to do this ... but at least it would quickly clear the field of some old distributions that are still around but essentially abandoned. At 08:29 AM 8/5/00 +0300, Artie Ball wrote: Hi all.. At the risk of starting a firestorm of debate.. I am still rather new to Linux, and over the past couple of months I have begun to wonder. which is the best distribution of Linux... which one should I concentrate on Which is the most used.. the best supported . the most robust? Is this just a matter of personal taste, or is it something that can be quantified?? Your thoughts please. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Chat/ppp
At 11:59 AM 8/3/00 -0700, davidturetsky wrote: With the change to ttyS2 and irq 10 I has successful in getting dialtone and connecting to msn This worked with: pon but not with: pon connection When I changed /etc/ppp/pap-secrets so that the default 'provider' matched what was in 'connection' (actually msn), pon completed the connection to msn until local hung up. I will work on sending actual text messages to myself Nothing to suggest yet, since so far you haven't merntioned any unresolved problems. Following Lawson's recommendation, I added setserial /dev/ttyS2 irq 10 to /etc/rcS.d/S40network. There seem to be many rc? files and this one seemed appropriate It will do, I suppose. I don't think you've said which Linux distribution you are using, and they do follow slightly different practices about init scripts. I also tried /etc/ppp/ppp-on, but this generated no dialtone. Here I am following the text of 'Running Linux' Better to tell us what you did than to reference a book. I don't have a copy handy to check, and in any case the problem may be that you didn't "follow" what is in it very closely. I need to be able to build a capability to send a single message to a potentially large list of subscribers. In windows, Outlook Express apparently enforces a limit of 64. I am thinking along the lines of using sendmail for this purpose, directing a piece of mail to a file of subscribers, which is apparently a trivial undertaking for sendmail. I have been reading O'Reilly's 'sendmail' by Costales and Allma Yes. I wouldn't know about Outlook Express. Sendmail or a workalike (exim, smail, a couple of others) should be able to handle this as an alias file. Or writing a short program to send a message to a long list of people is a simple undertaking; I did it (in perl) years ago, to support an employee-surveying Web site I helped develop. Another option is to use a mailing-list package like mailman or majordomo. (But be careful if you use majordomo - a flaw that many people, but not majordomo's author, see as a serious security hole was discovered a month or so ago. Since the author doesn't see it as a problem, it won't be fixed, and majordomo's slightly restrictive license prevents anyone else from fixing it and distributing the fix. This seems to be speeding up what was a slow migration from majordomo to mailman.) Which way is best depends on the details of your need. Unfortunately, this capability is most often associated with the sending of SPAM ... but I don't actuall know what you have in mind. Suggestions on any of the above would be welcome [olds stuff deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Utilizes 24MB instead of 32MB / eliminate hda2 1MB
First, the 24 vs 32 meg issue: This is only a wild guess, but does your mobo perhaps have onboard video? Some mobos (not all) with onboard video use "shared" memory for video, which means that the video card uses a portion of the memory that would normally be available to the system. On the one such mobo I have, video grabs 8 mb -- eithe an interesting coincidence or, possibly, the reason for your missing memory. Only a guess, though. In any case, since (you say) it happens with Windows as well as with Linux, it surely is some sort of BIOS-level phenomenon, not a Linux problem as such. On the partition issue: first. hda2 is "eating" 1K, not 1M, of space by your report. (If this report were in megabytes, your hdd would be a 10-terabyte drive, a size even I would find daunting.) Second, it is surely an extended partition, sort of a "wrapper" that in turn holds the partition table for hda5 and hda6. IDE drives can have only 4 "real" partitions. They can be 4 primary partitions or 3 primary and 1 extended partition. In the second case, the extended can hold any number (or at least some large number) of additional "logical" partitions. Look at the drive a bit more carefully with fdisk or cfdisk and you will probably see this relationship. At 07:06 PM 8/3/00 +0530, Madan A S wrote: [root@Madan]$cat /proc/meminfo total:used:free: shared: buffers: cached: Mem: 22892544 22360064 532480 13049856 618496 13893632 Swap: 73986048 2600960 71385088 MemTotal: 22356 kB MemFree:520 kB MemShared:12744 kB Buffers:604 kB Cached: 13568 kB BigTotal: 0 kB BigFree: 0 kB SwapTotal:72252 kB SwapFree: 69712 kB This the output of cat /proc/memento in my Linux system The imp point over here is that Mem:22892544 but when I boot the initial memcheck counts untill 32768K which is a difference of 12MB , The same happens in my windows also which says my ram is 24MB Thus my system utilizes only 24 of 32MB available Physical check reveals it is indeed 4chips of 8MB in SIM Can anyone suggest what is wrong ? and what is the remedy ? --- [root@Madan]$cat /proc/partitions major minor #blocks name 3 06298425 hda 3 14024251 hda1 3 2 1 hda2 3 3 128520 hda3 3 41646662 hda4 3 5 425691 hda5 3 6 72261 hda6 2264 1073741823 hdd I think you have recognised my problem , i can't eliminate hda2 which is just eating 1Mb of space partition magic and fdisk doesn't display its existance Do you have any idea !!! -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Chat/ppp
First, let's keep this conversation on the list. I've added it back. I use mailman for the lists I manage and like it. I use the Debian Linux distribution here, and mailman is available as a .deb package. If you use something different, use a search engine to find what you need. Mailman is not public domain, but it is Open Source (more exactly, released under a DFSG-complaint license). I can't readily do an MTA comaprison for you, but if you look around, I'm sure you'll find more opinions about the relative merits of various ones than you can stand to read. I actually use exim and I find it easier to set up than sendmail (which I've used in the past). Both have actually workd OK for me. I haven't used smail or the others so don't have informed opinions about them. At 02:42 PM 8/3/00 -0700, davidturetsky wrote: Where is mailman available? Public domain? Do you recommend it? Can you compare sendmail vs exim/smail, and others? David -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: crossover cable??
At 11:39 AM 8/3/00 PDT, Chris S wrote: i am trying to connect my windoze machine to my linux box, and in the HOWTO it says i will need a 'crossover' cable. is this a special type of cable or can i just use any old RJ45? i set up samba and wasn't able to see anything with win98 machine, i thought the cable may be the problem. thanks in advance. Of course it is a a special cable; that's why they take the time to tell you you need one. It interchanges the Ethernet send and receive pairs, so you can connect two Ethernet cards directly, without using a hub in between. They are cheap to buy, or if you are adventurous, you can make one yourself. Here's how: Normal cableCrossover Cable 1 -- 1 1 -- 3 2 -- 2 2 -- 6 3 -- 3 3 -- 1 6 -- 6 6 -- 2 -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: crossover cable??
Richard -- this is incorrect. UTP-based Ethernet has the same kind of "two-endedness" as serial connections. Just as DTE and DCE pinouts are different for RS232, "hub" and "device" ends are different in UTP Ethernet. A UTP crossover cable is an exact analog to a modem-eliminator cable. See my separate response to the original question for the actual "pinout" differences. At 07:21 PM 8/3/00 +, Richard Adams wrote: On Thu, 03 Aug 2000, Chris S wrote about, crossover cable??: i am trying to connect my windoze machine to my linux box, and in the HOWTO it says i will need a 'crossover' cable. is this a special type of cable or can i just use any old RJ45? i set up samba and wasn't able to see anything with win98 machine, i thought the cable may be the problem. thanks in advance. Basicly you use whatever cable/connectors the card(s) "BOTH" support, be it coax cable rj-45 or whatever. Normally speaking one refers to a crossover cable when one talks of serial communication from one com port to another on a remote machine. So to answer your question in simple terms IF you mean an ethernet card, as long as both ethernet cards have the same type of connector then it does not matter what you use, the only exception to the rule is speed. -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: crossover cable??
At 07:58 PM 8/3/00 +, Richard Adams wrote: ... See my separate response to the original question for the actual "pinout" differences. ... I bet you also know that the pincher needed to make the cable costs much more than 100 meters of cable. Not here in California. Just looking at 1 catalog I have lying around, I find CAT 5 cable:$11/100ft =~ $36/100m RJ45 crimping tool: $20 Crossover cable, 10' $ 7 I could find better prices for all of these things with a bit of looking. But the reason not to make one's own cables isn't, in my view, the cost of the tool (I already own the tool, but I still make my own cables only in emergencies). It is that pre-made cables are cheap and reliable; the time they save outweighs their high materials cost relative to buying unrerminated cable and connectors. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Red Hat and NAT
At 07:36 PM 8/3/00 -0500, Eric Peeters wrote: Hello all, I am exploring the possibility of building a Linux-based firewall for a small network and I am wondering whether I can implement NAT (Network Address Translation) on a Red Hat Linux 6.2 and if yes, how I go about doing that. Yes, you can; people do it all the time. I think the RH stock kernel even has the stuff you need compiled in (other distributions don't and you need to compile you orn kernel). Look at the Firewall HowTo (at www.linuxdoc.org) for general guidence on the process, and don't hesitate to report with specific questions it leaves unresolved. The box in question would be connected at one NIC to the router and at the other NIC to the LAN's switch. This is a standard router/NAT/firewall setup for Linux. Remember to check if the connection between the Linux router and the other router requires a crossover cable (you don't say what the other device is, and both "device" and "hub" wired kinds exist). -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Red Hat and NAT
At 12:04 AM 8/4/00 -0400, Tammy Fox wrote: In Linux, NAT is called ipchains. Visit http://www.linuxheadquarters.com/howto/networking/ipchains.shtml for instructions on how to set it up for your network. It is much easier to follow than the LDP HOWTO. Actually, in Linux the most familiar form of NAT is called IP Masquerading, not ipchains. The 2.2.x kernel's "policy routing" features provide a capacity for doing 1-to-1 NAT as well, though relatively few people are familiar with it -- you use the facilities of the "ip" program (usually in the "iproute" package) to control it. Charles Steinkeuhler, a Linux Router Project (LRP) virtuoso, has built a nice example of this latter capability in his "EigerStein 1.1" script (check for it at lrp.steinkuehler.com ... though I'm not sure if 1.1 is posted yet). ipchains itself is a userspace application, used to configure firewall rules (packet filtering, mostly) in 2.2.x kernels. In that capacity, it sets some Masq'ing rules as well. You also need to know about the app ipmasqadm and about the use of the various ip_masq_*.o modules. The ipfwadm, ipportfw, and ipautofw userspace apps did the same thing for the 2.0.x kernels, BTW. Tammy's URL is certainly easy to follow, but that is because it leaves out most of the important details. If you limit yourself to following the instructions there, you'll find your LAN lacks a lot of the capabilities that the Masquerading code can provide (port forwarding, a working ftp service, irc service, and so on). There are also better ways to handle DNS than it provides, mainly by running a BIND forwarder on the Linux router. The Firewall HowTo, while certainly more complicated, actually covers these additional features. There are separate Ipchains and IP-Masquerading HowTos as well It's also a good idea to set up an actual firewall when doing a Masq'd router of this sort. While ipchains can do this, it takes more than one ipchains command to set up a proper firewall. Look at (for example) the Seattle Firewall project (seawall.sourceforge.net) for some good instruction here. Or look at lrp.c0wz.com for some of the LRP-based material on good firewall design. Or look at the "ipmasq" Debian package, which autogenerates a passable firewall. -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Chat/ppp
At 02:08 AM 8/2/00 -0700, davidturetsky wrote: I changed the ln entry per Mike's suggestion: My chat log now reads: pppd[155]: Connect script failed The connect script you posted looks okay, but there is no way to tell for sure without knowing WHY it failed. Could be a hidden typo, incompatibility with what the ISP wants, modem problems, ... or something else I'm not thinking of offhand. Surely the above line is not the ONLY log entry; chatscripts will log their own results is for run chat with the appropriate flag (I think -v). Tey again, and this time be sure to get chatscript log output. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: new drive - partitions sizes?
At 11:04 AM 8/2/00 -0500, Jim Reimer wrote: Well, the dead drive's in the trash, the new one is in the machine. This one is 10 Gig, the smallest available at the local store. In the past, I've always just dumped everything into one partition, but this drive is big enough that I think it should be broken down a bit. Requesting guidance on which root directories should have their own partitions ( I'm thinking /, /usr, and /home - any others? ) Conventional answer: the ones you listed, plus /var and /tmp . Possibly /boot (a small partition created as the first one) if needed to satisfy LILO constraints. and how big they should be - either as an absolute size or as a percentage of the whole. (This is a stand-alone system, one user, no mail, news, etc.) Much harder to answer. Almost any choices that are not obviously ridiculous will work in the setting you describe. I'd probably do this ... / - 1 gig /usr- 2 gigs /home - 4 gigs /var- 0.5 gigs /tmp- 0.5 gigs /boot - 50 megs /swap - as needed /spare - whatever is left (or add it to /usr or /home) -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Chat/ppp
OK. The chatlog you show indicates that your modem is not responding to the initial command sent by chat. Since you say ... As I recalled, when I first installed debian last year I used pppconfig and then pon (but with the wrong nameservers) and got dialtone and some dialog before the server disconnected, I'll skip asking about winmodems and other questions I'd ask if the modem *never* worked under Linux. But you still need to check that you are trying to use the correct serial port (i.e., the right /dev/ttyS* entry) and a speed that the modem can run at. So ... what serial port is the modem on? did pppconfig successfully find it there, or did you tell it the port identifier? what speed did you tell pppconfig to use for the port? Is it a speed that your modem understands? have you made *any* hardware changes since the time when you "first installed debian last year"? (I'm wondering about IRQ conflicts.) are you still running the same Debian install as you had then? At 03:06 PM 8/2/00 -0700, davidturetsky wrote: Well, I checked everything... including making sure I did not have the winmodem inadvertently installed. I also corrected the nameserver entries, although this should not prevent me from getting a dialtone The chatlog given below is accurate and only omits "pppd[484]: pppd 2.3.5 started by root uid0" and ends with "pppd[484]: Exit." I then went in and ran pppconfig (again) followed by pon This is what the chatlog showed: Aug 2 14:07:58 debian pppd[258]: pppdd 2.3.5 started by root, uid0 chat[259]: abort on (BUSY) chat[259]: abort on (NO CARRIER) chat[259]: abort on (VOICE) chat[259]: abort on (NO DIALTONE) chat[259]: abort on (NO ANSWER) chat[259]: send (ATZ^M) chat[259]: expect (OK) chat[259]: alarm pppd[258]: Connect script failed chat[259]: failed pppd[258]: Exit As I recalled, when I first installed debian last year I used pppconfig and then pon (but with the wrong nameservers) and got dialtone and some dialog before the server disconnected, but now... no dialtone, no chance for MSN to even kick me off! [old stuff deleted] -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Chat/ppp
At 05:14 PM 8/2/00 -0700, davidturetsky wrote: I just checked through windows, interrogating the modem: US Robotics 56K FAX INT PnP Identifier *USR3031,ISAPNP\USR3031 Port: COM3 Interrupt: 10 Addr 3E8 IRQ 10 is not the normal IRQ for ttyS2, so a call to ttyS2 won't connect to a modem at that IRQ. If the modem is in this location under Linux -- a big if, unless you are using isapnp to set it; if you aren't, I can't begin to guess where it might decide to put itself -- you need to use setserial to point a ttyS* device at it. For example setserial /dev/ttyS2 irq 10 will do (ttyS2 already has the right IO address). But don't count on this working, unless you've taken care of initializing the modem with isapnp. My impression is that under Linux, the modem is not responding at all (of course I am using it to communicate here), so I take it that I have some mismatch in the setup, etc I also had a second winmodem installed and physically removed it to avoid any potential of a conflict/confusion As per below, I changed the reference to S2 for COM3, changed the nameserver entries in /etc/resolv.conf (although that should not affect dialtone) I assume the settings in Windows are the same for Linux, but perhaps I should run a little experiment, trying S0 and S1 [old stuff deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: keepalive/wakeup script
A few observations.
First, the ping line in Jack's script surely should read:
system ("ping -c 1 $host");
so it sends only one ping packet instead of an unending stream of them.
Second, perl has "sleep" built in, so that line *can* read
sleep $sleep" ; }
(though the way Jack wrote it does work).
Third, if the program is run in the background (./alive ), won't the
"print" line block and stop the program from continuing to execute until it
is re-attached to a console?
At 11:08 AM 8/1/00 -0500, Jack Barnett wrote:
...
Also a quick perl script works, I have one on my system that looks something
like this (IIRC):
#!/usr/bin/perl -w
$sleep = "180"; $host = "myisp.net";
while (1 == 1)
{ system ("date");
system ("ping $host");
print ("Staying Alive, Staying Alive, o o o yea, Stayng
Allivee\n");
system ("sleep $sleep"); }
Then run it with something like
./alive
--
"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: crash!!
I bet you have "green" power management features enabled, so the drive spins down when idle, but for some reason isn't spinning back up (or at least not promptly enough to satisfy the kernel. Check your BIOS settings. At 02:07 PM 8/1/00 -0500, Jim Reimer wrote: No network card, no modem. 90MHz Pentium, 16meg ram, on-board ide Maxtor 90432D2 drive Red Hat 6.0 It's looking more and more like it's the drive. As long as I keep using it, it's ok. But if I go off for a while it's dead when I come back - like the drive's going to sleep and not waking back up. Rebooting won't wake it up, but cycling the power switch will. Could be power supply problem (?). ... -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: connecting to isp
mike -- While we welcome succinctness, you've taken it to a point that makes things a bit hard to follow. Still, I suspect your problem is in the chatscript portion of the ppp command. You say you send: Use this for real connection: pppd /dev/modem 57600 connect '/usr/sbin/chat " " ATDTxxx CONNECT " " Username: Password: ' noipdefault defaultroute Since you've obviously modified this from what you really use (all the x's), I'm not certain how exactly to read it. But I believe you are waiting for a space at the start, where you don't want to be waiting for anything. That is, where your script reads chat " " ATDT it should read chat "" ATDT without the embedded space. Not sure of this ... it's been some mnths since I last ysed pppd/chat ... but I think this is your problem. Leaving ip_forwarding off is fine in this application. Is this connection command all At 05:16 PM 7/29/00 -0700, mike mcmanus wrote: Hi, I've just installed linux 5.1 on a second hard drive. i know I need to do a lot more reading but I am impatient to get connected to my isp. i am using the MASTERING LINUX Arman Danesh book. My problem and details as succinctly as I know how : redhat 5.1ppp2.3.3-4 kernel 2.0.34. Am trying to connect to isp. Used mincom and can connect and reciever ip addr. Use this for real connection: pppd /dev/modem 57600 connect '/usr/sbin/chat " " ATDTxxx CONNECT " " Username: Password: ' noipdefault defaultroute There is no modem dialing. I checked /var/log/messages and, typical for each attempt, I get this: kernel ppp ver 2.2.0 kernel ppp dynamic channel kernel pppline disc. registered pppd pppd 2.3.3 started by root chatfailed pppd connect script failed pppd exit pppd line disc. unregistered Chat consistently fails so is that the prob and if so how to fix. I have gone to /usr/sbin/chat and it is there in hex so it not missing. only other thing I notice ip forwarding is off. I'm assuming that's ok. -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: DNS - a FAQ
You want to set up your BIND to do two things (on the LAN) --
1. be authoritative for the zone "fakedomain.lan"
2. resolve all external addresses, using either a forwarder
(the ISP's nameserver) or the rootnameservers.
To do the first, add to named.conf entries similar to this:
// add entries for other zones below here
zone "fakedomain.lan"
{
type master;
file "/etc/bind/fakedomain.lan";
};
zone "1.168.192.in-addr.arpa"
{
type master;
file "/etc/bind/reverse.192.168.1";
}
You then need to create the zone files themselves, but that's
standard BIND stuff, same as for any domain.
To do the second ... it depends on which way you want to do it. To use
forwarders, either add a forwarders section or uncomment the one in the
sample named.conf, in either case using the values the ISP provided, as follows
forwarders
{
204.156.128.1;
204.156.128.10;
};
To use the root servers, just make sure this part of the sample file (and
the file it points to) is present:
// prime the server with knowledge of the root servers
zone "." {
type hint;
file "/etc/bind/db.root";
};
I hope this is enough to get you started. If you have other, specific
questions, or if I misunderstood what you were asking, please don't hesitate
to post a follow up. I don't understand what your comment about dynamic
address sasignment on the external interface has to do with this problem,
for example.
PS - I almost missed your request for a CC ro [EMAIL PROTECTED]. A better way
to ask for this is to CC that address on your own message; then a
"reply-to-all" will pick the address up.
At 10:44 AM 7/28/00 +0300, Razvan Sandu wrote:
Hello!
Could someone please give me an answer to the following problem, which - I
think - it's almost a FAQ ? Or, at least, to point to the appropriate Web
page ?
I have a simple LAN (the simplest: one Ethernet segment). Workstations are
Win95. One machine can dual boot also Linux and, when in Linux, acts as a
gateway (it has a modem and it connects to the ISP through dial-up on a
normal, switched line). It provides WWW browsing and e-mail to all Win95
clients.
For various reasons, I want to set up a small DNS for INTERNAL use. Since I
DON'T HAVE AN OFFICIAL DOMAIN, I use 192.168.1.0 addreses and all my
machines are in the "fakedomain.lan" domain.
Reading Nicolai Langfeldt's DNS-HOWTO, I didn't understand how should I set
up my DNS for this type of configuration, in order not to disturbe anyone.
Cause my LAN is neither completely isolated, nor a part of my official ISP
domain.
As usual, my ISP allocates a dynamic, non-fixed IP address to the gateway,
every time when it connects through dial-up.
As a note, I HAVE to use the "notify yes" clause, since I also plan to have
a secondary nameserver on my LAN.
A Cc: to [EMAIL PROTECTED] in your response will be appreciated.
--
"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: A simple perl question.
Output to STDOUT is buffered. In perl, a LF forces the buffer to clear, so
it prints immediately. That's why the third version works the way it does.
The buffer also outputs when it gets full. That's why the first one does
what it does; it is printing fast enough to fill the buffer quickly.
Your second program is just enormously slow ... if you let it sit for
(maybe) 15 minutes, you should see some output, whenever it gets around to
filling the buffer.
There is a switch in perl for unbuffered output (well, sort of unbuffered).
Try putting the line "$| = 1 ; " before you start ptinting and see if that
changes the way number two operates.
This same behavior is often a problem in CGI programs written in perl.
At 03:03 PM 7/27/00 +, Dan wrote:
Here's one that's driving me mad ... I'm trying to write a script and I want
the output to continue on the same line, but when I add in anything that
slows down the loop I can only get an output if I add a line feed.
(RH 6.1 and perl-5.00503-6.)
Any ideas?
Dan.
Three example are below to try and explain what I mean.
-
#!/usr/bin/perl
while () {
print "hello world";
}
gives
o worldhello worldhello worldhello worldhello worldhello worldhello
worldhello worldhello worldhello worldhello worldhello worldhello worldhello
worldhello worldhello worldhello worldhello worldhello worldhello worldhello
worldhello worldhello worldhello worldhello worldhello wor
--
When I add something in that slows the loop down, a subroutine, or even a
simple sleep command ...
#!/usr/bin/perl
while () {
print "hello world";
sleep 1;
}
I get nothing at all, the cursor just sits there.
--
Then I add a line feed ...
#!/usr/bin/perl
while () {
print "hello world\n";
sleep 1;
}
And get this ...
hello world
hello world
hello world
hello world
hello world
hello world
hello world
hello world
--
--------"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: recursive find cmd
And one more basic one: find ./ -name whatever finds starting at the current directory location. This plus Richard's two will cover most situations you are likely to face. At 08:51 PM 7/26/00 +, Richard Adams wrote: On Wed, 26 Jul 2000, Chris S wrote about, recursive find cmd: i have already looked at the man pages on find, but could someone just tell me how i go about finding files/folders recursively throught the directory tree? find / -name pipo* Serach the whole disk for any file starting with 'pipo' find /usr/local/src/ -name pipo* Same again but now only look in '/usr/local/src/*' -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: dns issues
It is difficult to be certain of the cause of your problem, because your
explanation is incomplete in several ways. It sounds like you still have a
name resolution problem of some sort ... slow telnet responses are a typical
symptom of that ... but you've told us too little to let me pin it down.
I expect that if you run through the following checklist, somewhere in it
you will find the source of your problem. Please check each of the following:
1. That 192.168.5.3 is the IP address of your Red Hat host.
2. That you do not have a name zone that corresponds to the zone you created
for reverse lookup.
3. That you do not have any entries in your /etc/hosts file.
4. That you do have a file /etc/nsswitch.conf that includes this line:
hosts: files dns
5. That you do have in /etc/resolv.conf the entry
nameserver 127.0.0.1
6. That named.192.168.5 does include a reverse-address for the IP address of
your Windows workstation
7. That you have a line in named.conf that tells named the path to
named.192.168.5 (on my system, for example, a zone entry similar to yours
would contain the full path, or 'file "/etc/bind/named.192.168.5" ').
8. That if you test BIND from the Red Hat machine with "host 192.168.5.X"
(replace X with the right value for the Windows machine), you get a response?
If you have enerything identified in this list correct, please post again,
this time telling us
the hostname and IP address of the Red Hat host
the hostname and IP address of the Windows host
the domain name the LAN uses internally
the complete contents of named.conf
the lines from any zone files listed in named.conf for the
Red Hat host and the Windows host
the telnet service line from /etc/inetd.conf (to verify that
you are using tcpd, which I've assumed)
the contents of /etc/hosts.allow and /etc/hosts.deny
the content of the "hosts:" line in /etc/nsswitch.conf
the contents of /etc/resolv.conf
At 10:25 AM 7/26/00 +0530, Gaurav Agarwal wrote:
hi all,
i just installed red hat linux 6.0 on my system ...
now i was trying to configure the caching-only dns.
( as it takes quite long to get the prompt when i telnet from other machine
on the network )
for that i edited the /etc/named.conf file and added
zone "5.168.192.in-addr.arpa" {
type master;
file "named.192.168.5";
};
then i created the file /var/named/named.192.168.5
and put in the related hostname and domain name etc ..
finally i restarted the named : /etc/rc.d/init.d/named restart
still when i telnet 192.168.5.3 from my windows machine on the network, it
takes over 60 secounds to show the prompy
--
--------"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: Print: cannot create /var/spool/lpd/lp/.seq
The error message you are getting now is different from the one you quoted in your first message, which was: lpd: cannot create /var/spool/lpd/lp/.seq Now you say you are getting: lpr: cannot create /var/spool/lpd/lp/.seq child process exited abnormally lpd is the print daemon; lpr is the application users actually run (it in turn calls lpd). Please check the status of lpr now and see how it its ownership /or permissions might be set incorrectly. At 02:17 PM 7/23/00 -0400, [EMAIL PROTECTED] wrote: Probably the problem is in this line: drwxr-xr-x2 root lp 2048 Jul 22 11:03 lp The corresponding entry on my system is: drwxrwsr-x2 lp lp 1024 Nov 19 1999 lp To change this: chmod 2775 /var/spool/lpd/lp Well, I changed the permissions, and then the owner, but neither helped. Thanks for trying, anyway. Still get error: lpr: cannot create /var/spool/lpd/lp/.seq child process exited abnormally Here's the present status of lp: [root@localhost /root]# ls -l /var/spool/lpd | grep lp drwxrwsr-x2 lp lp 2048 Jul 22 11:03 lp -rw-r--r--1 root root6 Jul 22 10:33 lpd.lock -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Uninstalling slackware packages
At 05:11 PM 7/20/00 -0400, John Marr wrote: Hey all: I've been reading about linmux gateway securing, and I want to secure my slackware 7 box. The articles I have been reading talk about how to uninstall programs like telnet and ftp through the use of "rpm -e -nodeps ..." Is there some way to do the same sort of thing with slackware? Depends on the thing. Slackware uses a more simple-minded packaging system than RH (and every other major distro), so you probably don't want to use that approach. So ... identify what you want to do and do the needed customizing by hand. For example, I assume you want to close the ports that provide telnet and ftp *services* (not remove the *clients*, even though that is what your message actually says -- the services are provides by telnet and ftpd, while telnet and ftp are clients). The simplest way to do this is to edit the file /etc/inetd.conf . Look for the lines that start telnet and ftp service and comment them out (put a # character at the beginning of each line). Save the file and restart inetd (I think "killall -HUP inetd" will serve this purpose). Whatever else you want to do, there will be a way to do it in Slackware. To tell you the way, we need to know what the "it" is. -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Print: cannot create /var/spool/lpd/lp/.seq
Probably the problem is in this line: drwxr-xr-x2 root lp 2048 Jul 22 11:03 lp The corresponding entry on my system is: drwxrwsr-x2 lp lp 1024 Nov 19 1999 lp To change this: chmod 2775 /var/spool/lpd/lp At 09:45 AM 7/23/00 -0400, [EMAIL PROTECTED] wrote: Ray, You probably have a mismatch between those two sets of ownership/permissions. Running as root overrides the conflict (since root ... [details deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: redirecting the incoming mail to another mail server
Probably. Depends on what you mean by "redirect". You probably means forward it to an account somewhere else. The normal procedure for that is to set up a file in your home directory called .forward that contains the full e-mail address you want mail forwarded to. Pine has nothing to do with it. though; it is a feature of your MTA (sendmail, exim, or equivalent). At 11:52 AM 7/22/00 +0530, Sandeep Shetty wrote: Hi all I am using RH 6.1 and pine for mail. Is it possible to redirect all the incoming mail to another mail server say hotmail.com. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Print: cannot create /var/spool/lpd/lp/.seq
At 10:59 AM 7/22/00 -0400, [EMAIL PROTECTED] wrote: Apparently while fiddling with my domain name, I broke user printing. Root can print ok, but user gets stderr: lpd: cannot create /var/spool/lpd/lp/.seq lpd folder has rwxr-xr-x permission. Removing /var/spool/lpd/lpd.lock didn't help. /var/spool/lpd/lp/.seq exists and is rw-rw---x--x You skipped over the one we need to see. What is the output (all of it, not just the permissions) of ls -l /var/spool/lpd/lp ? Probably lpd cannot read from or write to that directory. Also tell us ls -l /wherever_it_is/lpd -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Print: cannot create /var/spool/lpd/lp/.seq
Sorry. I asked these questions imprecisely so got the wrong answers back. My fault, I'll try again. 1. I'd like to see the ownership of and permissions on the directory /var/spool/lpd/lp itself, not on its contents. To do that, I guess you need to run ls -l /var/spool/lpd |grep lp 2. The lpd I want to see ownership of and permissions on is the executable, not the spool directory. Probably " ls -l /usr/sbin/lpd" . You probably have a mismatch between those two sets of ownership/permissions. Running as root overrides the conflict (since root can, sort of, do "anything"), but is shows up when running as an ordinary user. At 01:39 PM 7/22/00 -0400, you wrote: ... I hope this is what you needed. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: XF86Setup?
At 02:42 PM 7/23/00 +1200, Clint Deckard wrote: Where can I find and use XF86Setup? I have used xf86config but am still having problems and can not find XF86Setup. I have RH 6.2 and S3 Virge 2mb video card. I don't have RH 6.2 here to check, but on my 6.0 disk it is its own .rpm, namely XFree86-XF86Setup-3.3.3.1-49.i386.rpm It relies on the VGA16 X server, so be sure to install that .rpm too. Tastes differ, but my experience with XF86Setup is consistently unhappy. If you are having X problems that xf86config can't handle, I doubt that using XF86Setup will help. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Confuguring @home with Linux
Hi, Mike. The usual problem with setting up @Home service with Linux is that many @Home installations use MAC-address authentication to identify the computers allowed to connect to the service. If this is the source of your problem, then the network is remembering the identity of the Ethernet card in your Win98 machine and will not lease "its" address to your Linux host. Four possibilities for getting around this: 1. Power down the cable-modem device and leave it off for about 5 minutes, to let it lose state (sort of equivalent to a PC BIOS losing its CMOS settings). Then plug everything in and see if the Linux host now gets a DHCP address. In some areas, the MAC-address stuff is in the cable-modem box and can be reset this way. 2. Tell @Home you have changed computers and they need to recognize the new NIC. Since this is NOT a Linux issue -- changing to a new Windows host would raise the same problem -- they shouldn't blow you off on this request. SOme areas control MAC-address settings from the head end. 3. Spoof the MAC address of the Win98 host's Ethernet card in the Linux host. If you can do this -- some NICs/modules will do it, but not all, and I don't know if 3c59x.o does -- you do it with a command about like the following: ifconfig eth0 hw Ethernet aa:bb:cc:dd:ee:ff replacing the last part with the MAC address you want to use (I'm not certain of the format for the address; the man page isn't clear on that, and I don't have an example here to check). 4. Connect the Windows computer, note the address and other information it gets in its lease, and hand-configure the Linux host with the same addeess, netmask, nameservers, etc. This might not work, and it is risky even if it does ... when your lease expires, you won't get a new address assignment ... but it will serve as a short-term test to verify that you don't have hardware problems we are missing. Another possibility is that you have to provide a specific hostname to the head end to get an address; if you do, they should have told you so (or put it into your Windows settings, if they did the setup for you). If so, that's a simple DHCP client option; there is probably a commented-out example in the dhcpcd config file. DHCP leases normally provide nameserver information, so your not having them is just another symptom of the failure to get a lease. Unfortunately, @Home service characteristics vary a lot from community to community, so while I (and others) can give you some general advice, you need to find someone who has installed Linux locally (in or near Mt View) to tell you what specific problems your instance of @Home imposes on Linux users. For more information, consult the Cable Modem HowTo, one of the HowTos at http://www.linuxdoc.org . Also, if you can search the archives of the [EMAIL PROTECTED] mailing list (it's archived at http://www.geocrawler.com/lists/3/Linux/303/0/), you will find a lot of discussion of problems connecting LRP router/firewalls to @Home. Sorry I can't give you more specific help than this. If you get partway through and have more problems, please don't hesitate to post a follow-up. At 08:04 PM 7/22/00 -0700, Mike Keithley wrote: Hello all I know this has been discussed and I've followed some of the suggestions but I'm deep in the woods as to getting My @home in Mountain View Ca to work at all. Help is really appreciated. ... I am using Slackware 7.1 under the 2.2.16 kernel. I know my eithernet card is working as ifconfig -a shows an entry for it. [other details deleted] -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Installing .deb packages ?
Well ... as a happy Debian users, if I had the opporite problem (needeing to install something I could get only as an .rpm), I would use "alien" to convert it from .rpm to .deb. I imagine alien works in both directions, though I've never tried it the other way. You might look for an alien.*.rpm package. At 11:12 AM 7/19/00 -0400, Renaud OLGIATI wrote: Am I in a nightmare ? I would like to install on my Mandrake system (which is a variant of RH from what I hear) the Wordperfect which resides on my Corel Linux CD. WP comes as a .deb file. So I need dpkg to install it But the only dpkg I have is also in a .deb file ;-( -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: setup cable modem howto?
At 04:17 PM 7/19/00 PDT, Chris S wrote: ... so to get on with it...i now want to get on the net via my cable modem that i use on my windows machine. so what do i do now? what's my next step? Probably to read the Cable Modem HowTo; there is one now, at www.linuxdoc.org . I can help with specific questions (as can others here, I'm sure), but you need something to walk you through the basics first. You need to know or decide things like -- what you want the machine to do: just connect itself, or serve as a router/firewall for other workstations? what services do you want to run? to access? how the cable service provides service -- e.g, does it use MAC-address authentication? Does it provide static or dynamic addressing? Does it use DHCP? These are only examples. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Linux and LS120 floppy drive
At 12:12 AM 7/18/00 -1000, David Jones wrote: Two small questions: 1. Will a Linux boot floppy work when booted from an LS-120 super floppy drive? Depends on the mobo. If the BIOS has support for booting from an LS-120, regular (1.44 mB) boot floppies will boot from it. I've seen reports that LS-120s choke on the "superformatted" sizes of floppies (e.g., 1.67 mB) but can't verify them personally. 2. Does Linux support the LS120 drive at all? Yes. You need to add in "ide-floppy" support, either compiled into the kernel or as a module. All 2.2.x kernels I've checked have it as a compile option. I have NOT been able to get Linux to boot from an LS-120 disk. This seems to be a geometry problem -- Linux and the BIOS see different geometries, and this causes LILO to choke. I've seen reports that syslinux will work from an LS-120 but have not been able to replicate them. I've also seen reports that one can do a two-step boot process, booting from a regular floppy, then inserting an LS-120 disk to use as the root filesystem, but I have not tried to replicate that procedure. If you do buy one and get it working, please send a report of what you did. Use of LS-120s appears to be murky, with a lot of misinformation circulating, and few reliable reports of what configurations work. Be cautious in relying on untested recipes for getting it working. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Win printer
There is a proprietary HP technology that was used to make what some people call "Winprinters". The problem was that HP didn't release the specs for the needed interface software, so no one could write an appropriate filter for Linux. But, if memory serves, someone managed to work around that. To find out, look up your actual printer by name (I can't do this since you didn't include a model number) in the database at http://www.linuxprinting.org/printer_list.cgi and see what it says to do about setting up a filter for lpr/lpd. Regular readers of the list might notice this is a new URL. It seems to be the same database previously at http://www.picante.com/~gtaylor/pht/printer_list.cgi Other than Samba, I don't know of a way for a Linux host to access printer resources on a Windows server. At 04:48 PM 7/18/00 +0800, Mary Christie Generalao wrote: I have an HP printer handed to me by another collegue and he said it's a Winprinter.. I assume this means that the printer require some Windows-specific components. But it works with a Windows Laserjet III driver... Is there a way I can use this with Linux directly? if it doesnt, is there a way I can share this printer to linux from a Windows machine if I we dont use samba? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: which is the default port
No question is "too dumb" for a beginners' list. 23 . You look up this sort of information in the file /etc/services . IF you want to connect to a different port, you specify that on the command line. For example, telnet somehost 25 will attempt to connect to port 25 (smtp) on somehost. At 05:28 PM 7/17/00 +0530, you wrote: sorry if the question seems too dumb... when i say telnet and a hostname which port does it normally connect to i.e which is the default port that i connect to. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: dial in line keeps dropping connection (addl info)
Sorry - I need to back up a step. What *getty are you running? More than that, what is the actual entry in /etc/inittab that watches the modem line for incoming connections? If you are (as I suspect) using mgetty, what does it log about these failed connections (probably in /var/log/mgetty/*, but check your man page to be certain if your version matches mine)? At 06:56 PM 7/17/00 -0400, Lewis Supply Co wrote [in part]: ... "I've tried a few different getty entries." May we see an example that fails (but that works on the SCO system, which is what I assume you mean by "Unix")? F9600 # B9600 CS8 CRTSCTS # B9600 SANE -ISTRIP HUPCL CRTSCTS # @SLOGIN # F9600 (that came in the gettydefs file as the default entry for fixed speed modems) nx # B9600 HUPCL # B9600 CS8 SANE HUPCL TAB3 ECHOE -IXON -IXOFF -IXANY -CTSFLOW -RTSFLOW ORTSFL # @SLOGIN #nx (that's what I use on SCO it works at 9600 and 19200) I've tried both of these at different speeds as well as the autobauding versions -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Kernel Patch failed?
I believe the process you've described patches kernel source. You don't mention the step of recompiling and reinstalling your kernel after doing the patch. Did you do these steps and not mention them as too obvious, or did you omit them? At 01:37 PM 7/16/00 -0400, 1stFlight ! wrote: ... I just tried updating my kernel from 2.2.15 to 2.2.16 using the patch. I used Netscape to download it, which changed it from a gz to the unziped form I think (when it arrived it the .gz extension had been removed). Then using "patch -p0 patch2.2.16" it appeared to update my kernel. But after rebooting and doing a "uname -a" it still shows my kernel as 2.2.15. Did I miss something? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Using pictures as bg in a XTerm
At 09:00 AM 7/15/00 -0300, OHPC wrote: How do I enable pictures as background in a XTerm? I have seen lots of pictures of X running terminals with this enabled, but didn't find anywhere the info about how to do it. You don't. You run the Eterm app to do this, not the xterm app. In Debian, the package name for installation is eterm. No doubt there is an .rpm with a similar package name (and the same for other packaging systems - I don't recall what distro you use). From the Debian description: Enlightened Terminal Emulator (ala xterm) with Imlib support A terminal emulator in the spirit of xterm or rxvt. However this one differs in that it uses an Enlightenment style config file, as well as themes. The Imlib graphics engine is used to render images. This version supports background images, pixmapped scrollbars, and the pseudo-transparency. Eterm also now uses libpthreads. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Using KDE and Wmaker from a command line
At 09:06 AM 7/15/00 -0300, OHPC wrote: How do I enable KDE, Gnome, or WindowMaker to be run from the command line? Every time I want to change my window manager I have to edit .xinitrc. I'm only replying because you say you didn't get any replies the first time -- this isn't really in my area of expertise. Since KDE, Gnome, and WindowMaker all require X in order to run, there is no way to run them *directly* from the command line. You'll need to set up some sort of scripts to start X with appropriate arguments (possibly telling it to use different init files, which I gether you already know how to write). I know this isn't much help, and I'm sorry for that. I do hope it is enough to steer you in the right direction. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Virtual desktop in Gnome (RH 6.0)
Well ... it depends on your X server and monitor settings. Without knowing them (and I don't have your earlier messages to refer back to for that information, if it was in them), I can only give you some general feedback. X will start up in the first listed size that it can achieve for the color depth that it starts in (you do the color-depth part by hand). But regardless of what one it starts in, it will use the resolution of the highest one specified. So even if your video card plus display will only support, say, 640*480, you will get that as a "window" into a larger, 1024*768 virtual screen. If the "Virtual" lines were set higher, you'd get a window into an even larger virtual screen. What you need to figure out is what in your video or monitor setting is holding you back to 640*480. It might be the available video memory, the hscan or vscan frequency, or simply a mismatch between the X server and your video card. Are you perhaps the poster who was having trouble getting X to work with an SiS chipset? did you get that part resolved? Might this be a continuing manifestation of the chipset problem? If you have 8 megs of video RAM, it certainly isn't a memory problem. Even if I've got you identified correctly, I don't think you ever mentioned the monitor scan frequencies. BTW, Gnome has nothing to do with this. It's an X server issue, not a WM/desktop environment issue. At 02:01 AM 7/14/00 -0300, ArlequÃn7W4= wrote: editing the Screen Section of the XF86Config file I've inserted this lines: Subsection "Display" Depth 16 Modes "1024x768" "800x600" "640x480" "320x200" Viewport0 0 Virtual 1024 768 EndSubsection Subsection "Display" Depth 8 Modes "1024x768" "800x600" "640x480" "320x200" Viewport0 0 Virtual 1024 768 EndSubsection and when I start the X server I have a giant desktop... I thought I was trying to increase my resolution (like in windows) but suddenly I have an enormous desktop!! he he How can I just keep my desktop fit into my screen but with a 800x600 or 1024x760 resollution ? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: o+o+o+o+o+o+o TO ALL THE PEOPLE SUBSCRIBED !!! +o+o+o+o+o+o+o
You're welcome. Oh, my mistake ... I thought you were thanking the people who took the time to read and respond to your question, not chastising us for making you have to press the DELETE key a few extra times. Let me address your current concern. Please consider that it is you, not the rest of us, who are new to this list. We know (most of us, anyway) how majordomo lists work. We also know (many of us, anyway) that sometimes this particular list cycles messages very S...L...O...W...L...Y. At times, I've seen over a 12 hour delay between the time I posted a message and the time I saw it on the list. In fact, the reply I sent to you about your video card hasn't yet come back to me on the list, and I sent it 5 hours ago. Also, some people post questions here without first joining the list. You might argue that such behavior is itself rude, and I tend to feel that way myself, in most situations. But this is a list for beginners, so it's worth cutting people some slack as regards "rookie errors". Various of us respond differently to this reality. Some of us send replies only to the list, pretty much sharing your view. Others of us (I'm in this group) send the replies to the poster and the list, assuming that the poster prefers a quick reply (and, to be honest, because it is easier than hand-editing the headers in EVERY DAMN REPLY I make). Perhaps others reply only privately (naturally, I wouldn't know, since thes messages would not appear on the list). Personally, when it is *you* asking *me* (and other like me) for help, I find being on the receiving end of a lecture about how majordomo works to be a bit insulting. Also, a technical note: if the list really wanted to encourage the behavior you describe, it would make it easy by inserting a Reply-to: header. I don't propose to open this argument -- for one thing, there is nobody actually on this list who *can* change its behavior as regards headers -- just to remind you that the list's default behavior is in part the result of a technical dscision made by whoever set up the list. I've obliged your procedural preferences this time. In the future, I'll try hard to remember not to reply to your messages at all. At 12:29 AM 7/13/00 -0300, ArlequÃn7W4= wrote: Hey, dudes !! Please, remember that we are subscribed to a Majordomo mail list!! We you post or reply a message, please send it ONLY to [EMAIL PROTECTED] I've received duplicated messages because some people put my electronic address into the `To:' line and put the list address into the `Cc:' or viceversa... Remeber that we all receive the mails posted to the list, so we don't have to carbon copy to the original sender ;-) We we reply a message, we must put into the `To;' line *only the list* electronic address, and then we all will receive the message !!! -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
MUA behavior (was: long, undescriptive Subject line)
Steve -- That we disagree about what is appropriate is no news to anyone, and it won't serve any purpose to extend the debate, since I think neither of us has anything new to say. BUt I did want to ask you about one technical comment you made that was news to me. At 05:02 PM 7/13/00 +1000, Steve Youngs wrote [in part]: ... - "when I hit 'reply' my mailer puts the original author in 'TO:' so I have to add the list address to Cc" (I'm sorry folks, but your mailer is broken. Configure it so it doesn't behave like this or get a mailer that will) ... (4) Configure your MUA to *ONLY* send to the list address. Gnus does this well. I checked into Gnus, and was disappointed to find that it is Emacs based. Without trying to start a big argument, I personally find emacs difficult to use so avoid emacs-based solutions. That said ... does anybody know of any other MUA that can be configured the way Steve describes here? Or is this a feature unique to Gnus (if so, I find it hard to accept the characterization that all the MUAs in the world except for one are "broken")? Of the many MUAs I've used over the years, I have NEVER found one that can be set to reply to the To: address in preference to the From: address. Or am I missing some other way that an MUA could be configured so that a reply would go ONLY to this list without hand editing the headers? Steve, perhaps you could sketch HOW you configure Gnus to do this, to help me (and perhaps others) figure out how to do something equivalent in other MUAs? Unless, of course, it is some special feature unique to Gnus. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: login problems
At 09:55 AM 7/13/00 PDT, Chris S wrote:
i have just created a new user but when i log in i get these error messages:
No direcory /home/cspielma!
Logging in with home = "/". -- i am guessing that it does this by
default??
Yes. Though I can see one certain typo in what you transcribed ("direcory"
for "directory"), and I suspect two other departures from what you actually
saw (the ! and . at the ends of the lines). I can't be sure because my
version of login presents a message quite different from yours when I have a
missing home directory.
bash: /home/cspielma/.bash_profile: Permission denied -- is this because
it wasn't able to log into the correct home directory??
Yes.
but there is a directory /home/cspielma!! i don't know why it won't log
into it.
Me either, except the vague understanding that you did something wrong.
1. How did you add the new user? If you used adduser or useradd, it's
probably fine.
2. Might there be a typo in "/home/cspielma"? Might you have entered, for
example, "/home/cspielma " (with a trailing space)? Or might the directory
name have been misspelled in the /etc/passwd entry (and you missed it in
transcribing the message for inclusion here)?
3. Might the permissions on /home or /home/cspielma be incorrect?
4. Am I making the wrong assumption in thinking the !'s are your editorial
additions and not part of the directory names? If I am, please note that
"/home/cspielma!!" is not the same as "/home/cspielma!"
--
"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: login problems
I added the list back into this exchange. It depends on *how* you add the user. It has to be done -- the user account needs to own its home directory (or at least have rwx rights to it) -- but programs like adduser and useradd normally take care of this detail for you. At 10:54 AM 7/13/00 PDT, Chris S wrote: i got a reply saying that i should run this command: chown cspielma /home/cspielma then when i logged in as cspielma i was able to access that directory. do i have to do this(the chown command) for every user i add? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Mounting a win32 file system from another partition?
Sure, in principle. Try mount /dev/hda1 /mnt (replace hda1 with the right partition identifier for your Windows partition) and see if it works. If it doesn't, your kernel probably lacks support for the "vfat" filesystem. See if "modprobe vfat" corrects this. If not, you may need to compile a custom kernel. At 10:20 AM 7/12/00 -0400, Chuck Wilson wrote: Hello ! I have a 6 gig harddrive partioned into two operating systems. A 5 gig partition for win 95 and a 1 gig partition. My question is, how can I mount the win 95 partition from linux? I have tried a lot of the how tos, but there isn't anything covering mount. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: converting text files from dos
fromdos . OR dos2unix . Or some variant of that. There are several of them, and distributions vary in which one they include. At 08:47 AM 7/12/00 -0500, Jim Reimer wrote: Well, at school, we had a little program called dtox that would strip out the CR's from a DOS text file, but now I'm without it. What's the best way (or any way, I suppose) to accomplish this under Linux? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Startx Won't Star
Given all the oddities you've encountered since this tech did his work, I'd really wonder if you have a complete picture of what he or she did. Specifically, I'd be curious as to whether you have the same video card that you previously had; these are not the sorts of changes I'd expect to see from a simple addition of memory. But in any case, your immediate problem is getting X running on a system that (apparently) has a "Diamond Stealth Video 2500, Alliance AT24" video card installed. The place to go for infor about video cards is http://www.xfree86.org/cardlist.html . The list there doen't include a 2500, but it does list five other "Diamond Stealth64 Video" video cards, all of which use the XF86_S3 X server. There's probably an XF86_S3*.rpm on your installation CD, and before I did anything major to the system, I'd try installing that and seeing if Xconfigurator (or a less automated but more informative installer like xf86config) can install it succesfully. To find out more about the actual video card you have in your system, use SuperProbe. At 02:10 PM 7/12/00 +0800, Peter wrote: Hi, I upgraded my memories in which process the videocard was removed. When everything was put back together W98 could not find the driver for the videocard. It was then installed and W98 run again. Going to Linux RH 6.2 startx did not work anymore saying "Fatal error, no screens found" Doing Xconfigurator my usual way, hitting on the first selection '2the MAXColor S3 Trio64V+' and then 'Custom' a message shows up: "Error dedecting video ram." When I go back and select my videocard Diamond Stealth Video 2500, Alliance AT24 which I never did before since I did not know my video card the program jumps out with the message "Server doesn't exist, can't continue. tried to use .../usr/X11R6/bin/XF86_SVGA. This is the same video card I always had. What do I have to do? Will reinstall RH 6.2 with or w/t formating / and/or /usr do the trick? Do I need a driver for linux for the video card? -- --------"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: how to set up domain?
At 06:39 PM 7/12/00 +0300, Nakajama wrote: hi what should I do to conect to my web server via browser? when I write IP it finds my web server, and when I write domain it doesnt find it. I assume you mean that you can connect to a URL like http://1.2.3.4/ but not to a URL like http://www.something_or_other.com/ even though www.something_or_other.com should resolve to 1.2.3.4 . If so, your problem is *probably* with name resolution. See if the same machine (the one with the browser) is also able to ping the server by IP address byt not by name (FQDN). If so, check whether it has functioning nameserver addresses listed in /etc/resolv.conf (or equivalent, if the browser is not running on a Linux workstation). -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Problem I have not come across or can think of a solution for UUMMMMMMMM
Are you sure it is a problem with Disk Druid and not the BIOS or drive jumpering? Does the BIOS recognize the drive as (approximately) 10 gB? During kernel boot, does the kernel report the drive as 10 gB capacity? If you try to partition with fdisk or cfdisk, what does the "print the partition table" display tell you about partitions and capacity? These programs are not as convenient as DD, but they are somewhat more informative. What prompts me to ask these questions is that you say the CPU is a "233 mmx". THat suggests to me possibly an old mobo with an old BIOS, perhaps one that has trouble with drives over 8 gB. Some 8gB drives have jumper settings that cause them to look like 8 gB drives, just to accommodate these old BIOSes. At 09:52 PM 7/12/00 +0100, Jill wrote: Hi all. I have a problem with disk druid, that I cannot find a solution for. What i am trying to do is run a Linux print server on a network of 10 machines. Using a 233 mmx with 2.5 gigs of fixed drive ( linux) and windoze 7 gig dual booting using RH 6.2 and KDE. The drive is fully functional with no bad sectors, and had a write test performed on it. Problems that I have are even though I delete all partitions and create new swap space and native . I get the error, cannot create as not enough room on the drive. I have even tried just asking for 1 gig native and 100 swap and still get the same message. I am ensuring that I do this on the correct drive. I think I am missing something here as on my machine at home I could install the two with a 100 meg swap and the 2.9 gig as native -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: no linuxconf
At 09:00 AM 7/11/00 -0700, Chris S wrote: i am trying to launch linuxconf from a terminal window, but when i enter the command all i get is [1] 687, which i know is the job number, but how do i access it? How are you launching it? This is the sort of message you get when you start something as a background process. If you are running bash, you can reattach the job to the console with "bg". Are you running the regular version of linuxconf or linuxconf-X? When you say "a terminal window", do you mean a console or an Xterm (or Eterm) under X? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: again, modutils
Alas, while you have deleted the answers you got, I have deleted your original question. And since this problem description, despite its length, leaves out a lot of important information, I can't give you very complete advice based on it. One very odd thing is this line: Jul 12 10:34:40 localhost modprobe: modprobe: Can't locate module /dev/ttyS1 Since /dev/ttyS1 is not a module (it's a device), modprobe should never be trying to locate it. This suggests to me that you have some fundamental configuration error that is not very well identified by the log output you sent us. Perhaps we need to see the script that runs pppd (intact EXCEPT that you should replace any password with Xs). More generally, I'm not sure why linuxconf would generate a hundred or so modprobe messages ... or any modprobe messages, for that matter ... I don't use linuxconf myself ... but it would be more instructive to see the messages generated during boot/init by modprobe or insmod than this batch. Similarly, the output of "ifconfig -a" and "route -n" would tell us (and you) more about your connectivity than the log messages. Finally, tell us some basics: what Linux distribution and version? What steps did you follow in compiling your kernel? Do the modules on the system match the kernel (does /lib/modules/2.2.16/ exist? did you do a "make modules_install" as the last step in your kernel compile process, before you moved the new kernel into place and re-ran lilo?)? At 11:26 AM 7/12/00 +1000, [EMAIL PROTECTED] wrote: I wrote in a little while ago about this, and unfortunately have deleted the responses, which amounted to telling me to recompile my kernel (2.2.16) with module support. While I was sure that I had done this in the first place (after all, vfat/sound/... module support were behaving well), I did it again to be sure. My main module concern was with ppp support, which I decided to compile into the kernel, i.e., not as a module. really, nothing's changed. Kppp (1.6.14), the only ppp app that I seem to be able to run not-root, still complains about no kernel ppp support -then does its merry wee thing all fine after that. Ok, weird, but it works. However, while it all kinda 'works', things are not right: here is what I see when I tail /var/log/messages, run an rxvt terminal and then run and exit linuxconf _without_ changing anything: (when I run the terminal): Jul 12 10:19:46 localhost modprobe: modprobe: Can't locate module char-major-5 (when I exit linuxconf): Jul 12 10:24:41 localhost modprobe: modprobe: insmod lo:0 failed Jul 12 10:24:41 localhost modprobe: modprobe: insmod lo:1 failed Jul 12 10:24:41 localhost modprobe: modprobe: insmod lo:2 failed Jul 12 10:24:41 localhost modprobe: modprobe: insmod lo:3 failed [...you get the idea...] Jul 12 10:24:46 localhost modprobe: modprobe: insmod lo:49 failed Jul 12 10:24:46 localhost modprobe: modprobe: insmod ppp0:0 failed Jul 12 10:24:46 localhost modprobe: modprobe: insmod ppp0:1 failed Jul 12 10:24:46 localhost modprobe: modprobe: insmod ppp0:2 failed [...ditto... (@ about 30 seconds of computation time)...] Jul 12 10:24:50 localhost modprobe: modprobe: insmod ppp0:48 failed Jul 12 10:24:50 localhost modprobe: modprobe: insmod ppp0:49 failed I can see that both ppp0 and lo are activated at boot in my boot messages. here is a /var/log/messages output of when I log onto the internet: Jul 12 10:34:39 localhost pppd[2081]: pppd 2.3.11 started by dasEs, uid 500 Jul 12 10:34:39 localhost pppd[2081]: Using interface ppp0 Jul 12 10:34:39 localhost pppd[2081]: Connect: ppp0 -- /dev/ttyS1 Jul 12 10:34:40 localhost pppd[2081]: local IP address 209.79.140.139 Jul 12 10:34:40 localhost pppd[2081]: remote IP address 203.56.11.29 Jul 12 10:34:40 localhost modprobe: modprobe: Can't locate module /dev/ttyS1 I'd like to be able to modprobe's job a little easier, can anyone tell me what, where, and when I should tell modprobe what it would like to know? -- ----"Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Closing port 113
At 11:27 PM 7/11/00 -0400, John Marr wrote: Does anyone know how I can close port 113 on my Linux box. I believe that this is used for irc connections, among other things. I am running Slackware 7.0. From /etc/services auth113/tcp authentication tap ident The ident service is used to authenticate your host, most often by IRC but sometimes by other services (e.g., smtp mail). To turn it off, just figure out what has it turned on. The likely candidate is a n entry in /etc/inetd.conf . If you find a line there for ident, just comment it out, then restart or re-HUP inetd (probably with "killall -HUP inetd"). -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ssh
You should be able to start sshd by hand, as root, with "sshd" at the command line. Then see if you can log in. For a more permanent solution, you want to have it run automatically. Just add an entry to start it to an appropriate rc file (I think RH uses rc.local for things like this). I don't recall if inetd will start sshd for you, and the sshd man page doesn't really address this question. At 11:33 PM 7/11/00 -0400, David Hearne wrote: ... Now, when I try to log in using my SSH client for my Win98 box (ttssh) I simply get an error : Connection Refused I don't get to enter any information at all. I'm assuming that the SSH daemon is either not running at all (ps doesn't show it) or that it's running incorrectly. The docs that come w/ OpenSSH don't seem to be of very great help either. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: again, modutils
Forgive what must seem like a foolish question this late in the game, but ... exactly what is the *problem* with your system? The network config report at the end of your message (the ifconfig and route output) look fine -- both ppp0 and lo look to be set up correctly. Does the system communicate properly over the ppp link? If not, how does it fail? Or are you concerned only about the ugliness of the modprobe errors, even though they reflect no actual, observed failures in your system's actual operation? Or just the delays that are introduced by the modprobe attempts, then failures, during boot/init? If so, I suspect that to troubleshoot them, we'll need to see whatever configuration file is generating them. At this point, I'm wondering if it is something as simple as that linuxconf is trying to modprobe modules that are already running (or in the case of ppp, is trying to modprobe a module for something that is already compiled into the kernel). In the end, I fear the problem is in your linuxconf settings. The fact that it wants to rerun LILO to produce a non-working version encourages that hypothesis. Since I don't run that app here, I am an unmlikely source for help on problems specific to its configuration. But with this preliminary work, perhaps someone more familiar with linuxconf will be able to focus in on the actual problem. BTW, this is what your reported LILO error means (for the little this explanation is worth): "LIL- The descriptor table is corrupt. This can either be caused by a geometry mismatch or by moving /boot/map without running the map installer." At 01:47 PM 7/12/00 +1000, [EMAIL PROTECTED] wrote [in part]: ... I'm running RedHat 6.2, kernel 2.2.16. My original original, full (c.d.) install was RH6.0, 2.2.5, but I have since updated. I updated modprobe when I did this (2.3.10). My new-kernel compilation(s) have been with make xconfig, compiling _in_ ppp support, as well as support for the kernel module-loader. ... The output of ifconfig -a after a very quick session online loLink encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:3924 Metric:1 RX packets:704 errors:0 dropped:0 overruns:0 frame:0 TX packets:704 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 ppp0 Link encap:Point-to-Point Protocol inet addr:203.109.153.98 P-t-P:203.56.11.99 Mask:255.255.255.255 UP POINTOPOINT RUNNING NOARP MULTICAST MTU:552 Metric:1 RX packets:685 errors:0 dropped:0 overruns:0 frame:0 TX packets:584 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:10 output of route -n Kernel IP routing table Destination Gateway Genmask Flags Metric RefUse Iface 203.56.11.990.0.0.0 255.255.255.255 UH0 00 ppp0 127.0.0.0 0.0.0.0 255.0.0.0 U 0 00 lo 0.0.0.0 203.56.11.990.0.0.0 UG0 00 ppp0 -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: no dialtone - FIXED
At 02:37 PM 7/10/00 -0300, OHPC wrote:
Well, I managed to fix the "no dialtone" problem.
I included some strings I found in the windozzz log.
It was trial and error, but it worked.
The dial string is now: ATS7=255X3DT
It is working, but I don't know exactly why.
Well ... let's tease it apart.
AT standard command prefix (why they are called "AT" commands)
S7=255 puts value 255 in register 7
X3 this is the X3 command I and others suggested
you try to turn off dialtone detection
DT the prefix for a DTMF ("Touch-Tone") dial
So ... the open question is: what does register 7 do? I don't know how
standard this is, but in the modem manual I checked, it tells the modem how
long (in seconds; default is 50, range 1-255) to wait for carrier before
giving up and hanging up.
Why does it work? Beats me, if you found that "ATX3DT" by itself did not
work. If you didn't previously find that, then it works because X3 is the
right command, as we guessed it might be.
--
--------"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: Hopefully a simple question...
This is normal behavior. Look in your environment ("env") for a variable
called PATH. It sets the order of directories that the bash shell will
search ehen looking for an executable. Normally, the directory you are in
(designated ./) is not part of that search PATH (at least for root; user
accounts don't seem to follow a standard in this respect).
It's more or less a security feature. If you are root, go into someone's
home directory, and execute (say) "ls", you want to be sure that what will
execute is the real /bin/ls program, not the script ./ls that does (to
choose a particularly nasty example) "rm -r /".
BTW, it's bad practice to convert directories, program names, etc., to upper
case. For example, I doubt you "have an executable in /USR/LOCAL/SBIN called
TCPDUMP". You are much more likely to have an executable in /usr/local/bin
called tcpdump .
In this instance, it didn't actually confuse. But since Unix/Linux commands
(unlike MSDOS commands) are case sensitive, it's a good habit not to change
the case of commands, directory names, and so on, when you are asking
questions about them.
Had you, for example, merely reported that the command "TCPDUMP" didn't work
and asked why, without mentioning that "./TCPDUMP" did work, I would have
explained that the correct command is "tcpdump".
At 02:10 PM 7/10/00 -0500, Brian Burns wrote:
Thanks to all who contribute to this list for us "newbies" to Linux.
Hopefully I can contribute back one day.
Here is the scenario:
I have an executable in /USR/LOCAL/SBIN called TCPDUMP.
When I am in the dir in a shell logged in as ROOT, I can only execute the
program by typing ./TCPDUMP instead of just TCPDUMP. The same goes for perl
scripts.
Am I missing something here or is this normal behavior?
[html repeat deleted]
--
"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: ftp server?
OK. Put differently, you have the ftp service running on host A but not on
host B. You did something different when you set the two hosts up. I can't
even begin to guess what you *did* that was different, but here is where to
look for the effect.
First, check the file /etc/inetd.conf, and see if it has a line for ftp. See
if the entry on B is the same as the one on A.
Second, that entry will name a program (probably /usr/sbin/frpd, or
/usr/sbin/in-frpd, or /usr/sbin/wu-ftpd). Make sure the program named on
this line is actually on the system. (I don't use Red Hat and don't recall
its package-naming conventions. But you may need to look for ftpd .rpms, not
just ftp .rpms. In many cases, *d.rpm provides the daemons (server end),
while *.rpm just provides the client programs.
Third, the line probably also mentions tcpd. Assuming it does, that program
uses the files /etc/hosts.allow and /etc/hosts.deny to control access. Check
them for anything that blocks access ("man 5 hosts_access" will give you the
details).
Fourth, am I right in guessing that you are attempting *anonymous ftp
access? It looks from what you quoted like you are using a Web browser, and
they typically assume anon access. You may not have set up B to provide
anonymous ftp. Check you a user "anonymous" in /etc/passwd . Also check
/etc/ftpusers to see if you have restricted access in it.
One of these suggestions will probably turn up the problem. IF not, please
post a followup, bu next time with a somewhat more complete description of
your setup.
At 12:31 AM 1/1/94 +, manolo wrote:
Hi all!
I have two linux A(192.168.1.1) and B(192.168.1.2) connected in a small
home network.
I used Redhat 6.0 in both , and I,m able to do from B
ftp A
and can go inside the linux box after log-in password.
but can not do from A
ftp B
""Connected to luna.laeliana.es.
421 Service not available, remote server has closed connection""
Can somebody do a sumary what I need to set-up to be able to
ftp both .
I planned to transfers files in both directions and seems
that this service is not avalaible.
I installs all rpms related with ftp but does not work.
thanks in advance
Manuel
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
--
--------"Never tell me the odds!"---
Ray Olszewski-- Han Solo
Palo Alto, CA[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-newbie" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.linux-learn.org/faqs
Re: Debian Potato
At 09:11 PM 7/10/00 +, Richard Adams wrote: On Mon, 10 Jul 2000, Chris Wrobel wrote about, Debian Potato: Hello everone, I need some help... :) I'm trying to get myself the new Debian Potato distro. That is I want to download it off the internet and use my cd burner to make an istallable cd. You can go to sites like tucows.com ... ftp://ftp.linux.tucows.com/pub/ISO Unfortunately, this site only has the older iso images, the ones for the last burn (2.1-r) of Slink. They aren't clearly labeled as Slink, but you can deduce it from the December 1999 date stamp. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Debian Potato
I haven't done a CD download of Potato, but I do use it on almost all my hosts. So let me suggest that instead of burning a CD, you try a network install, which should go well over a cable hookup. Just get yourself a set of 5 disk images (rescue.bin, root.bin, and 3 driver-*.bin) from (for example) ftp://ftp.us.debian.org/debian/dists/potato/main/disks-i386/current/images-1.44/ and use them to do the network install you prefer. Although Potato is now "frozen", it still gets regular updates to correct urgent (mostly security) problems, new kernels as released, and so on. The last iso's that I know of are about a month old, making them already a bit out of date. If you need to install to a lot of hosts, you might still do better to mirror the Potato archive, using rsync to keep your mirror up to date. If you really want to burn a CD, the latest Potato iso's are at http://linuxiso.org/debian.html Debian discourages their use, though. If you want to follow the advice of the Debian team, you'll follow the procedure described beginning at: http://cdimage.debian.org/ The Debian CD FAQ has information on the actual burning. See http://cdimage.debian.org/faq.html At 01:32 PM 7/10/00 -0700, Chris Wrobel wrote: Hello everone, I need some help... :) I'm trying to get myself the new Debian Potato distro. That is I want to download it off the internet and use my cd burner to make an istallable cd. Why I need help?? I have never attempted to do this over the internet as I got my cable not that long ago. So can someone tell me what do I need to make such a cd and where I can get the stuff, -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: Debian Potato
Since I don't burn Debian CDs, I'm only guessing. With that disclaimer ... ... I assume you are referring to this part of the ftp directory: potato-i386-1.list 14-Jun-2000 20:5687k potato-i386-1.raw 11-Jun-2000 14:28 643M potato-i386-2.list 14-Jun-2000 20:56 119k potato-i386-2.raw 11-Jun-2000 14:31 643M potato-i386-3.list 14-Jun-2000 20:5755k potato-i386-3.raw 11-Jun-2000 14:33 485M I'd read this as implying that Debian-Potato's i386 binaries fill 3 CDs. The *.raw files are the actual iso images (the date is about right, from what I recall), and the *.list files are the package lists that correspond to the similarly-numbered CD images. The usual Debian practice is to put the most important packages (from main, at least) on CD #1, so I'd begin by making an iso using potato-i386-1.list . AS you do this, remember that the iso images only cover main and, maybe, contrib. They don't cover non-free and non-US, leaving out a few packages you might care about (ssh is the obvious example). At 04:20 PM 7/10/00 -0700, Chris Wrobel wrote: ... one thing I was confused about is why are there three different ".list" files at this location http://cdimage.debian.org/potato_test-cycle-2_nonUS/ for the i386 potato I'm downloading the biggest one right now I hope that's right ;) what do you think? ... -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: PPP and Eth
Yes it is possible, if you mean eth0 connected to a cable modem ppp0 connected to a modem on a phone line Having the two connections themselves isn't really any harder than the separate work of setting each one up. What is hard is the routing table. In the simple sorts of static-routing setups typical for Linux routers, one or the other will be the default route and get all traffic not explicitly routed to the other. To have both interfaces, you need either to run a fancier routing protocol or to "bridge" the interfaces. For the first, look at zebra. The second is ISP specific and probvably not a real option for you. A more common setup is a failover system, in which the router periodically checks that the cable-modem link is up and, if it isn't, starts the ppp interface and changes the default route to it. Then it keeps monitoring eth0 until it comes back, at which point it drops ppp0. It's a bit tricker than I've made it sound, but people do get it working (there's been discussion of this on the [EMAIL PROTECTED] mailing list). BTW, I don't know this area in much more detail than what I've written here. So general follow-up questions won't get helpful answers (at least not from me). At 04:36 PM 7/10/00 -0700, Chris Wrobel wrote: I was just wondering is it possible to have an internet access throught cable on eht0 and through a phone line at the same time ?? Maybe its a dumb question but is that possible?? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: routing to a different machine
It depends on several things you haven't told us. First, do you already have the "dialup" machine set up as a router? I'll assume you do. Second, is it doing firewalling /or Masq'ing now? Probably. Third, what kernel is it using? 2.0.x and 2.2.x kernels are VERY different in how they handle all routing/firewalling.NAT'ing stuff. Assming my guesses are right about 1 and 2, you use either ipportfw (2.0.x) or ipmasqadm (2.2.x) to set this up. An example ipfwadm entry would be: ipportfw -A -t 1.2.3.4 /80 -R 192.168.1.42/80 (replacing 1.2.3.4 with the external interface address and 192.168.1.42 with the internal address of your Web server). The man pages for ipfwadm or ipmasqadm will give you useful help, as will the various HowTos: Firewall, Ipchains, IP Masquerading, I think Ipfwadm. Check for them at www.linuxdoc.org . At 04:02 AM 7/10/00 +0930, Pete wrote: I suppose the answer to this is simple but ... I want to route external http requests sent my Linux dialup machine, to a web server on my second (internal)machine. How would I tell my Linux machine to send all requests on port 80 to my other machine ? I want to setup a web server on my local network but I don't want it running on my dialout machine. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
Re: ? configuring NIC KNE110TX ?
Tulip support is almost surely part of the kernel tree Slackware provides. Look (in make menuconfig) under Network device support --- Ethernet (10 or 100Mbit) --- DECchip Tulip (dc21x4x) PCI support I just checked this on an old Debian host that still runs 2.2.13. The Kensington card I have (don't recall the exact model number) does use tulip. I needed to switch to a newer version of tulip to make it work, but that was with a 2.0.x kernel -- I'd expect the version included with 2.2.13 to be new enough. There is no compelling reason to compile NIC support into a kernel, though there is probably no urgent reason not to either. The tulip driver autoprobes, so all you would need to do (using it as a module) is add an entry for it to (if I recall Slackware's init setup correctly) /etc/rc.d/rc.modules . Surely the stock Slackware kernels include a tulip.o module. At 11:38 AM 7/8/00 -0400, Charles E. Gelm wrote: Howdy, All: I've just installed Slackware-7.0 - kernel 2.2.13. But I'm lost at where to find information about getting the kernel to recognize my NIC. It is a Kingston KNE110TX. I think that it may be a 'tulip'. I've browsed at Kingston and get refered to another web page and after some more browsing I found 'tulip.c' v0.92. Is this what I need? If yes, where do I put 'tulip.c'? How do I generate 'tulip.o' and where should it reside? I want to use this box as an IP Masquerade and Samba file server. I'd rather compile the NIC driver into the kernel instead of installing it as a module. I didn't see this NIC mentioned while doing a 'make config'. How do I include this NIC driver into my kernel? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
re: More strange occurances
Every distribution (Linux or *BSD) has security risks, since they all rely on a lot of third-party code. As problems become known, the (responsible) distributions release updates. Though I'm no Red Hat specialist, I know RH does this. You can find them, for example, at Metalab, in ftp://metalab.unc.edu/pub/Linux/distributions/redhat/updates/6.2/ You can also find them, for 6.2 and older versions, at RH's own ftp site, though I find the performance of it so poor that I'm unwilling even to spend the time to track down the exact URL. In any case ... whether you stay with RH or move to another distribution, you need to join its security-announcements list (I imagine RH has one; if not, join a general-purpose list like bugtraq) and install patches as they become available. In addition, RH may have some default setting that make it vulnerable. Examples are probrams set to suid that don't need to be and unneeded services in /etc/inetd.conf . Someone more familiar than I with RH's defaults would have to give you spscific guidence here; for general advice, try the Security HowTo at http://www.linuxdoc.org/HOWTO/Security-HOWTO.html . At 02:40 AM 7/7/00 -0400, Karthik Vishwanath wrote: ... I have grown on RedHat and am very hesitant about installing another distribution. What are the ways to "make it safer"? Is there any place that I can read about this? -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
re: More strange occurances
At 11:24 PM 7/7/00 -0400, Karthik Vishwanath wrote: Well, I did do a reinstall today, and got the patches from Redhat for the wuftd and a few more things lying around on my machine. I have still decided to keep with Redhat and have not yet checked Bastille linux out. Good start. Remember, though, that to be safe, you have to stay with it. That is, you need, as the bare minimum, to check for new RH updates regularly, and install them promptly (at least ones that are security related). For example, yesterday, Mandrake announced a security problem with the "man" command. Recently, the FreeBSD folks identified a problem with OpenSSH, and Debian announed a fix for a hole in the DHCP client provided by OpenBSD (and used by Debian). Several places have reported a problem in BitchX. This is all in the last week. I don't know what of these problems might affect Red Hat generally or your system in specific. You should find out. I don't know Bastille, but it too may be a good thing to try. Also, you can look at Seattle Firewall (it's a project hosted at Sourceforge) for what's reputed to be a good Open Source firewall setup. You should do a few basic checks too, like seeing what ports are open in inetd.conf and closing any you don't want open. Switching from telnet to ssh is another smart security move. Good luck. -- "Never tell me the odds!"--- Ray Olszewski-- Han Solo Palo Alto, CA[EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-newbie" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.linux-learn.org/faqs
