From feedle at feedle.net Thu Dec 6 11:27:31 2001 From: feedle at feedle.net (C. Sullivan) Date: Thu, 6 Dec 2001 19:27:31 +0000 (GMT) Subject: [buug] Low end web server In-Reply-To: <20011130113735.U32527@linuxmafia.com> Message-ID: On Fri, 30 Nov 2001, Rick Moen wrote: > > Hardware-wise how low should we go, in your opinion? > > If you go under a 486 with 64MB RAM and a 2.1 GB hard drive, you'll > probably regret it. But maybe not. A well tuned 486 should have more than enough performance. Any Pentium-class machine should be more than enough. As a point of reference, here I run an MP3 streamer box on a Pentium 200 machine with 64 megs of RAM. It gets a couple of hundred web hits per day by housemates requesting music for the stream. It has no trouble handling the webhits AND dishing out a single mp3 stream. Similarly, I have a moderate volume (300+ inbound messages per day) mailing list machine that's a Pentium 133. It barely breaks a sweat. > I assume that puts your hardware picture in a bit better perspective? > People new to Linux -- especially servers -- usually make the mistake of > vastly overestimating the need for CPU power, while slightly > shortchanging I/O and in particular the disk subsystem. This cannot be overstated. People seem to forget how lousy performace really is on IDE, especially if you're talking an older box that might only support a 33MHz IDE bus. Running two disks as master/slave only increases the suckyness. Heavy disk usage on a low-end Pentium box can really bring things to a full and complete stop. This is why it's also important to get as much RAM as you can. Reduce disk paging, especially if you are using IDE. Apache can be a memory hog (especially if you load lots of modules). Smaller, thinner HTTP servers (like boa) can also help here, especially if you don't need fancy shiznat like PHP. A 486 with 64 megs of RAM and a lean Linux install using Boa as a webserver should perform nicely, with enough room to spare. > Better have a backup strategy. Using software RAID-1 (mirroring) will > make you feel clever right up until you realise that something's gone > from _both_ drives. Also keep in mind that RAID recovery on Linux can be a scary experience. It WILL work, but the warnings that raidtools gives you will (and should, really) make you want to change your pants. Better to not depend on the RAID to save your butt: there's no excuse for good backup strategy. Also, RAID will also cause performance issues. > And it'd be nice if there were a nice, safe choice in reliable, > well-supported PCI ethernet chipsets. It's difficult to say what's your > safe bet, these days. Intel is playing variation-du-jour with its > EthernetExpressPro 100 chipset, and the glory days of the DEC > 1040/21041/21140 "Tulip" chipset ended shortly after Intel bought and > discontinued the production line. Someone might be able to recommend a > reliable Tulip clone used by somebody. (NetGear? Samsung? LinkSys? > ADMtek? ASIX? LiteOn? MXIC? STmicro? Kingston? D-Link?) Although the driver is still a bit sketchy, I've noticed a trend to the RTL8139 chips. YMMV. The Macronix versions of this chip were poorly supported, but that may have improved recently (I had nothing but trouble with 2.2.19 and a Macronix chipset) Linksys is making a good Tulip clone card. There are still a small number of Netgear 310TX cards available, but the newer ones have a slightly buggy Tulip clone (LiteOn) chipset. Check Fry's (ick).. the Fry's here in Phoenix AZ has a good quantity of the "new" Netgear 310 cards. The common trend on the new good (that is to say, usable) Tulip clone cards is they are LiteOn chipsets. > You see, the other mistake newcomers to hardware for Linux keep making > is to assume that newer is better. Older, standard components are > likely to have mature, well-tested drivers. If you stick to quality > parts, you can do very well indeed. (This is why a lot of people > stocked up on $24 NetGear cards that had genuine DEC Tulip chips on > them, before supply ran out. They're still good.) Even the LiteOn clone Netgear 310TX cards didn't suck. I couldn't agree more that newer isn't always better in the Linux world. Don't get me started on VIA 686 motherboards. Argh. From steven-liang at 21cn.com Mon Dec 17 19:17:12 2001 From: steven-liang at 21cn.com (Steven Liang) Date: Tue, 18 Dec 2001 11:17:12 +0800 Subject: [buug] How to install the Redhat 7.2 in IBM Thinkpad iServer Message-ID: <003301c18772$7e355350$0f01010a@stevenlaptop> Hello every one: does anyone know how to install the Redhat 7.2 in IBM's Thinkpad iServer laptop? Thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL: From steven-liang at 21cn.com Mon Dec 17 19:21:10 2001 From: steven-liang at 21cn.com (Steven Liang) Date: Tue, 18 Dec 2001 11:21:10 +0800 Subject: [buug] How to install the Redhat 7.2 in IBM Thinkpad iServer Message-ID: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> Hello every one: does anyone know how to install the Redhat 7.2 in IBM's Thinkpad iServer laptop? Thanks. From john at jjdev.com Mon Dec 17 19:36:56 2001 From: john at jjdev.com (johnd) Date: Mon, 17 Dec 2001 19:36:56 -0800 Subject: [buug] How to install the Redhat 7.2 in IBM Thinkpad iServer In-Reply-To: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> References: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> Message-ID: <20011218033656.GA949@vette.jjdev.com> On Tue, Dec 18, 2001 at 11:21:10AM +0800, Steven Liang wrote: > Hello every one: > does anyone know how to install the Redhat 7.2 in IBM's Thinkpad iServer laptop? > Thanks. What specific problems are you having? When I installed linux on *my* laptop, the only problem I had was that the CD wouldn't boot. I booted off a floppy and it all worked. johnd ps - > get slack From p_b_h at lycos.com Tue Dec 18 09:28:54 2001 From: p_b_h at lycos.com (Johnny Ho) Date: Tue, 18 Dec 2001 09:28:54 -0800 Subject: [buug] Installing X on Dell with i810 chipset Message-ID: I'm new to *nix and FreeBSD, and I've tried following the handbook but still could not get X configure correctly. Has anyone gone through this headache or is it just me. :( Thanks, From evan at theunixman.com Mon Dec 17 19:35:55 2001 From: evan at theunixman.com (Evan Cofsky) Date: Mon, 17 Dec 2001 19:35:55 -0800 Subject: [buug] How to install the Redhat 7.2 in IBM Thinkpad iServer In-Reply-To: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> References: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> Message-ID: <20011218033554.GC815@hack.theunixman.com> On Tue, Dec 18, 2001 at 11:21:10AM +0800, Steven Liang wrote: > Hello every one: > does anyone know how to install the Redhat 7.2 in IBM's Thinkpad iServer laptop? > Thanks. > > What specific problems are you having installing it? -- "Considering the number of wheels Microsoft has found reason to invent, one never ceases to be baffled by the minuscule number whose shape even vaguely resembles a circle". -- unknown, but _very_ sharp Evan Cofsky evan at theUNIXman.com DSA key DA253F39 http://www.theunixman.com From bill at wiliweld.com Tue Dec 18 09:36:37 2001 From: bill at wiliweld.com (Bill Schoolcraft) Date: Tue, 18 Dec 2001 09:36:37 -0800 (PST) Subject: [buug] Installing X on Dell with i810 chipset In-Reply-To: Message-ID: At Tue, 18 Dec 2001 it looks like Johnny Ho composed: > I'm new to *nix and FreeBSD, and I've tried following the handbook but still could not get X configure correctly. Has anyone gone through this headache or is it just me. :( (word_wrap=70) What type of Dell machine is this ? A laptop ? -- Bill Schoolcraft PO Box 210076 -o) San Francisco CA 94121 /\ "UNIX, A Way Of Life." _\_v http://forwardslashunix.com From bill at wiliweld.com Tue Dec 18 09:38:05 2001 From: bill at wiliweld.com (Bill Schoolcraft) Date: Tue, 18 Dec 2001 09:38:05 -0800 (PST) Subject: [buug] How to install the Redhat 7.2 in IBM Thinkpad iServer In-Reply-To: <20011218033554.GC815@hack.theunixman.com> Message-ID: At Mon, 17 Dec 2001 it looks like Evan Cofsky composed: > On Tue, Dec 18, 2001 at 11:21:10AM +0800, Steven Liang wrote: > > Hello every one: > > does anyone know how to install the Redhat 7.2 in IBM's Thinkpad iServer laptop? > > Thanks. > > > > > > What specific problems are you having installing it? > > Do you see it here at http://linux-laptop.net/ibm.html ? -- Bill Schoolcraft PO Box 210076 -o) San Francisco CA 94121 /\ "UNIX, A Way Of Life." _\_v http://forwardslashunix.com From bferrell at baywinds.org Tue Dec 18 09:43:08 2001 From: bferrell at baywinds.org (Bruce Ferrell) Date: Tue, 18 Dec 2001 09:43:08 -0800 Subject: [buug] Installing X on Dell with i810 chipset References: Message-ID: <3C1F802C.E9B14B39@baywinds.org> You may find some help at yolinux.com. I don't own it, but is sure looks to have ANYTHING you'd want in terms of linux info or just simple X stuff Bill Schoolcraft wrote: > > At Tue, 18 Dec 2001 it looks like Johnny Ho composed: > > > I'm new to *nix and FreeBSD, and I've tried following the handbook > but still could not get X configure correctly. Has anyone gone > through this headache or is it just me. :( > > (word_wrap=70) > > What type of Dell machine is this ? A laptop ? > > -- > Bill Schoolcraft > PO Box 210076 -o) > San Francisco CA 94121 /\ > "UNIX, A Way Of Life." _\_v > http://forwardslashunix.com > > _______________________________________________ > Buug mailing list > Buug at weak.org > http://www.weak.org/mailman/listinfo/buug -- Bruce One day at a time... One second if that's what it takes From feedle at feedle.net Tue Dec 18 12:23:29 2001 From: feedle at feedle.net (C. Sullivan) Date: Tue, 18 Dec 2001 20:23:29 +0000 (GMT) Subject: [buug] Installing X on Dell with i810 chipset In-Reply-To: Message-ID: The Intel 810 chipset has a couple of unique issues. I cannot speak for FreeBSD, but on Linux there is special kernel-land stuff for handling this chipset. It is likely that XFree86 for FreeBSD will require a backport of this stuff. You might try a Google search on "Intel 810 FreeBSD"... I did that and got some promising looking information. -Fedl On Tue, 18 Dec 2001, Johnny Ho wrote: > I'm new to *nix and FreeBSD, and I've tried following the handbook but still could not get X configure correctly. Has anyone gone through this headache or is it just me. :( > > Thanks, > > > > > _______________________________________________ > Buug mailing list > Buug at weak.org > http://www.weak.org/mailman/listinfo/buug > From john at jjdev.com Wed Dec 19 11:16:37 2001 From: john at jjdev.com (johnd) Date: Wed, 19 Dec 2001 11:16:37 -0800 Subject: [buug] mapped io Message-ID: <20011219191637.GA265@vette.jjdev.com> Can someone refer me to some docs on how to read from a port address? something like the inp funtion from the dos lib conio.h I'm looking for something to do this: ----------------------------------- I/O Mapped I/O To place things in the I/O map the CPU must be forced to activate the I/O control signals. The machine code instruction for IN and OUT perform this action. There are no corresponding ANSI-C commands. But almost every i86 architecture C compiler provides some support. Two or more functions are declared in the conio.h header file. Then these functions are defined and linked into the C library. These two groups of functions are known as inp and outp. Inp Function The inp functions has one argument that is the port address. The function reads the port and returns the data read at that port address. Typically there is one function to read a single byte called inp and a second function to read a word called inpw. These functions are declared as follows: int inp(unsigned port); unsigned inpw(unsigned port); To read the byte at port address 0x45 and to save that byte in the integer ch would be coded as: #include int ch; . . . ch=inp(0x45); From itz at speakeasy.org Wed Dec 19 11:25:24 2001 From: itz at speakeasy.org (Ian Zimmerman) Date: 19 Dec 2001 11:25:24 -0800 Subject: [buug] mapped io In-Reply-To: <20011219191637.GA265@vette.jjdev.com> References: <20011219191637.GA265@vette.jjdev.com> Message-ID: <86n10fau6z.fsf@speakeasy.org> john> Can someone refer me to some docs on how to read from a port john> address? something like the inp funtion from the dos lib john> conio.h man 4 port -- Ian Zimmerman, Oakland, California, U.S.A. GPG: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087 In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles. From davids at idiom.com Wed Dec 19 11:48:08 2001 From: davids at idiom.com (David S.) Date: Wed, 19 Dec 2001 11:48:08 -0800 Subject: [buug] mapped io In-Reply-To: <20011219191637.GA265@vette.jjdev.com>; from john@jjdev.com on Wed, Dec 19, 2001 at 11:16:37AM -0800 References: <20011219191637.GA265@vette.jjdev.com> Message-ID: <20011219114808.A24214@malign.rad.washington.edu> On Wed, Dec 19, 2001 at 11:16:37AM -0800, johnd wrote: > Can someone refer me to some docs on how to read from a port address? > > something like the inp funtion from the dos lib conio.h > > I'm looking for something to do this: > ----------------------------------- I'm not sure what you mean by "port address", but I think that you just want read(2). David S. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 229 bytes Desc: not available URL: From john at jjdev.com Wed Dec 19 11:54:19 2001 From: john at jjdev.com (johnd) Date: Wed, 19 Dec 2001 11:54:19 -0800 Subject: [buug] mapped io In-Reply-To: <86n10fau6z.fsf@speakeasy.org> References: <20011219191637.GA265@vette.jjdev.com> <86n10fau6z.fsf@speakeasy.org> Message-ID: <20011219195419.GA513@vette.jjdev.com> > > man 4 port > > -- > Ian Zimmerman, Oakland, California, U.S.A. I've come up with this: -------------------- . #include #include int main(void) { ioperm(0x300, 5, 1); printf("inv: %d\n", inb(301)); } ------------------------------- This program will only run as root...(It must be compiled with the -O flag or it wont work) In linux, can a program access port memory directly with out being run as root? From itz at speakeasy.org Wed Dec 19 12:32:30 2001 From: itz at speakeasy.org (Ian Zimmerman) Date: 19 Dec 2001 12:32:30 -0800 Subject: [buug] mapped io In-Reply-To: <20011219195419.GA513@vette.jjdev.com> References: <20011219191637.GA265@vette.jjdev.com> <86n10fau6z.fsf@speakeasy.org> <20011219195419.GA513@vette.jjdev.com> Message-ID: <86d71bar35.fsf@speakeasy.org> johnd> This program will only run as root...(It must be compiled with johnd> the -O flag or it wont work) In linux, can a program access johnd> port memory directly with out being run as root? No, but you can always use 1/ the setuid bit (dangerous) 2/ sudo 3/ (on debian) super -- Ian Zimmerman, Oakland, California, U.S.A. GPG: 433BA087 9C0F 194F 203A 63F7 B1B8 6E5A 8CA3 27DB 433B A087 In his own soul a man bears the source from which he draws all his sorrows and his joys. Sophocles. From will_sargent at yahoo.com Thu Dec 20 01:56:13 2001 From: will_sargent at yahoo.com (Will Sargent) Date: Thu, 20 Dec 2001 01:56:13 -0800 Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP Message-ID: I throw myself on the mercy of the mailing list. I have RTFM'ed to the depth of my understanding, and I'm still hosed. I can't get my IBM Deskstar 75GXP to be recognized as more than 2 GiB under Linux. I'm using the 30 GiB model (IBM-DTLA-307030). I've tried Mandrake 8.1 and Redhat 7.1, and both distros will come up on boot recognizing the drive type, but then assigning it as a 2 GiB drive. The CHS is printed as 255/255/63, if CHS still has any meaning. I'm using an Asus A7V133 motherboard, revision 1004. This drive and motherboard worked fine for me under Windows 2000 (so it's not a bad drive) or outdated BIOS (which is unlikely). The Asus board includes an integrated Promise ATA/100 controller, and I've tried using that instead of the standard ide channels, but that doesn't seem to help (again, it recognizes the drive on /dev/hde, but has the wrong CHS and number of sectors). I've even tried specifying the CHS directly using bootparam. The command line I'm currently using is: linux ide2=0x9400,0x9002,autotune hde=3737,255,63 which at least forces hdparm -g to show the right CHS, but still doesn't change the number of sectors (i.e. it still thinks it's a 2 GiB drive). I don't have anything wierd in my system that I know of. There's a Macronix Tulip type ethernet card and a SB Live soundcard. I also have the 50x CD-ROM drive plugged into ide2 so it gets recognized by the BIOS and can boot up (although it's still a secondary slave). Anyone who knows whats happening here shall receive my eternal gratitude. Will. _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com From psoltani at ultradns.com Thu Dec 20 09:06:37 2001 From: psoltani at ultradns.com (Patrick Soltani) Date: Thu, 20 Dec 2001 09:06:37 -0800 Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP Message-ID: <3DBB075EEB95944492E127F2B9A96FAF0CE08A@ultra-exchange.UltraDNS.com> Is the drive scsi or IDE? If ide, then have you checked the LBA, Logical Block Addressing, in the bios? I have run into situations that simply instructing the bios to handle the ide disk with LBA enable resolves the size problem. If scsi disk, then I would do a low level format of the scsi device. Usually the adaptor allows you to do this via a key-stroke/menu or the manufacturer has a utility that you run off of a floppy, usually DOS based, and takes care of this. Good luck. Regards, Patrick Soltani. -----Original Message----- From: Will Sargent [mailto:will_sargent at yahoo.com] Sent: Thursday, December 20, 2001 1:56 AM To: buug at weak.org Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP I throw myself on the mercy of the mailing list. I have RTFM'ed to the depth of my understanding, and I'm still hosed. I can't get my IBM Deskstar 75GXP to be recognized as more than 2 GiB under Linux. I'm using the 30 GiB model (IBM-DTLA-307030). I've tried Mandrake 8.1 and Redhat 7.1, and both distros will come up on boot recognizing the drive type, but then assigning it as a 2 GiB drive. The CHS is printed as 255/255/63, if CHS still has any meaning. I'm using an Asus A7V133 motherboard, revision 1004. This drive and motherboard worked fine for me under Windows 2000 (so it's not a bad drive) or outdated BIOS (which is unlikely). The Asus board includes an integrated Promise ATA/100 controller, and I've tried using that instead of the standard ide channels, but that doesn't seem to help (again, it recognizes the drive on /dev/hde, but has the wrong CHS and number of sectors). I've even tried specifying the CHS directly using bootparam. The command line I'm currently using is: linux ide2=0x9400,0x9002,autotune hde=3737,255,63 which at least forces hdparm -g to show the right CHS, but still doesn't change the number of sectors (i.e. it still thinks it's a 2 GiB drive). I don't have anything wierd in my system that I know of. There's a Macronix Tulip type ethernet card and a SB Live soundcard. I also have the 50x CD-ROM drive plugged into ide2 so it gets recognized by the BIOS and can boot up (although it's still a secondary slave). Anyone who knows whats happening here shall receive my eternal gratitude. Will. _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com _______________________________________________ Buug mailing list Buug at weak.org http://www.weak.org/mailman/listinfo/buug From will_sargent at yahoo.com Thu Dec 20 20:20:25 2001 From: will_sargent at yahoo.com (Will Sargent) Date: Thu, 20 Dec 2001 20:20:25 -0800 Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP In-Reply-To: <3DBB075EEB95944492E127F2B9A96FAF0CE08A@ultra-exchange.UltraDNS.com> Message-ID: I think ATA is just the official term for IDE, so they're the same. It's not LBA addressed by the BIOS right now; it's handled by the ATA 100 controller. As far as the BIOS goes, it doesn't even HAVE a hard drive :-) Will. > -----Original Message----- > From: buug-admin at weak.org [mailto:buug-admin at weak.org]On Behalf Of > Patrick Soltani > Sent: Thursday, December 20, 2001 9:07 AM > To: Will Sargent > Cc: buug at weak.org > Subject: RE: [buug] Can't get Linux to see more than 2 GiB on 75GXP > > > Is the drive scsi or IDE? > If ide, then have you checked the LBA, Logical Block Addressing, in the > bios? > I have run into situations that simply instructing the bios to handle > the ide disk with LBA enable resolves the size problem. > If scsi disk, then I would do a low level format of the scsi device. > Usually the adaptor allows you to do this via a key-stroke/menu or the > manufacturer has a utility that you run off of a floppy, usually DOS > based, and takes care of this. > > Good luck. > > Regards, > Patrick Soltani. > > -----Original Message----- > From: Will Sargent [mailto:will_sargent at yahoo.com] > Sent: Thursday, December 20, 2001 1:56 AM > To: buug at weak.org > Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP > > > I throw myself on the mercy of the mailing list. I have RTFM'ed to the > depth of my understanding, and I'm still hosed. > > I can't get my IBM Deskstar 75GXP to be recognized as more than 2 GiB > under > Linux. I'm using the 30 GiB model (IBM-DTLA-307030). I've tried > Mandrake > 8.1 and Redhat 7.1, and both distros will come up on boot recognizing > the > drive type, but then assigning it as a 2 GiB drive. The CHS is printed > as > 255/255/63, if CHS still has any meaning. > > I'm using an Asus A7V133 motherboard, revision 1004. This drive and > motherboard worked fine for me under Windows 2000 (so it's not a bad > drive) > or outdated BIOS (which is unlikely). The Asus board includes an > integrated > Promise ATA/100 controller, and I've tried using that instead of the > standard ide channels, but that doesn't seem to help (again, it > recognizes > the drive on /dev/hde, but has the wrong CHS and number of sectors). > > I've even tried specifying the CHS directly using bootparam. The > command > line I'm currently using is: > > linux ide2=0x9400,0x9002,autotune hde=3737,255,63 > > which at least forces hdparm -g to show the right CHS, but still doesn't > change the number of sectors (i.e. it still thinks it's a 2 GiB drive). > > I don't have anything wierd in my system that I know of. There's a > Macronix > Tulip type ethernet card and a SB Live soundcard. I also have the 50x > CD-ROM drive plugged into ide2 so it gets recognized by the BIOS and can > boot up (although it's still a secondary slave). > > Anyone who knows whats happening here shall receive my eternal > gratitude. > > Will. > > > _________________________________________________________ > Do You Yahoo!? > Get your free @yahoo.com address at http://mail.yahoo.com > > > _______________________________________________ > Buug mailing list > Buug at weak.org > http://www.weak.org/mailman/listinfo/buug > > _______________________________________________ > Buug mailing list > Buug at weak.org > http://www.weak.org/mailman/listinfo/buug > _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com From bill at wiliweld.com Thu Dec 20 22:30:10 2001 From: bill at wiliweld.com (Bill Schoolcraft) Date: Thu, 20 Dec 2001 22:30:10 -0800 (PST) Subject: [buug] [Solaris-8] Multiple hostnames on aliased IPs? Message-ID: Hello Guru's I've setup a machine that has 3 aliases to the one NIC and of course three seperate hostname entries in /etc/hosts, which all work fine. I've been asked to make it so when you login to foo, foo1, foo2, or foo3 you could execute the command: echo $HOSTNAME and receive a unique hostname based on the particular interface you logged into, which would in essence make the entries in: /etc/hostname.iprb0 <------ foo loghost /etc/hostname.iprb0:1 <------ foo1 /etc/hostname.iprb0:2 <------ foo2 /etc/hostname.iprb0:3 <------ foo3 be the hostname when you "echo $HOSTNAME" /etc/hosts 10.1.0.100 foo loghost 10.1.0.101 foo1 10.1.0.102 foo2 10.1.0.103 foo3 I'm thinking that maybe in ~/.login I could maybe setup some script to querey the login's origin, compare to /etc/hosts and set export HOSTNAME=`$(information_acquiring_script)` I'm a little lost on the mechanics of how the login would reverse map the originating IP requested to enter the machine then look into /etc/hosts to grab the hostname. Any pointers would be appreciated. Thanks From rick at linuxmafia.com Fri Dec 21 10:58:34 2001 From: rick at linuxmafia.com (Rick Moen) Date: Fri, 21 Dec 2001 10:58:34 -0800 Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP In-Reply-To: <3DBB075EEB95944492E127F2B9A96FAF0CE08A@ultra-exchange.UltraDNS.com> References: <3DBB075EEB95944492E127F2B9A96FAF0CE08A@ultra-exchange.UltraDNS.com> Message-ID: <20011221105834.H27273@linuxmafia.com> begin Patrick Soltani quotation: > Is the drive scsi or IDE? He said "/dev/hde", so IDE. My status as an inveterate SCSI bigot gets more deeply entrenched all the time, I must say. With SCSI, these things just don't crop up. Will, you have to enable sector translation (LBA, or ECHS, or something like that) in your ROM BIOS Setup. Because IDE (or more properly, ATA) is brain-dead, and relies on BIOS drive-table information to a depressing extent. -- "Is it not the beauty of an asynchronous form of discussion that one can go and make cups of tea, floss the cat, fluff the geraniums, open the kitchen window and scream out it with operatic force, volume, and decorum, and then return to the vexed glowing letters calmer of mind and soul?" -- The Cube, forum3000.org From rick at linuxmafia.com Fri Dec 21 13:36:22 2001 From: rick at linuxmafia.com (Rick Moen) Date: Fri, 21 Dec 2001 13:36:22 -0800 Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP In-Reply-To: References: <3DBB075EEB95944492E127F2B9A96FAF0CE08A@ultra-exchange.UltraDNS.com> Message-ID: <20011221133622.M27273@linuxmafia.com> begin Will Sargent quotation: > I think ATA is just the official term for IDE, so they're the same. That's one way of putting it. I prefer to say that IDE is the marketroid's term for what is more _accurately_ called ATA. > It's not LBA addressed by the BIOS right now; it's handled by the ATA > 100 controller. Sorry, but ATA is inherently dependent on BIOS drive-type tables. Which ATA has had problems over the years handling consistently and reliably. Thus your current problem. For a snapshot of ATA as it was a number of years ago, please see: http://linuxmafia.com/pub/hardware/eide.txt Nothing fundamentally has changed since then: The standards have been updated, but the trends cited have continued. -- "Is it not the beauty of an asynchronous form of discussion that one can go and make cups of tea, floss the cat, fluff the geraniums, open the kitchen window and scream out it with operatic force, volume, and decorum, and then return to the vexed glowing letters calmer of mind and soul?" -- The Cube, forum3000.org From rick at linuxmafia.com Fri Dec 21 13:40:03 2001 From: rick at linuxmafia.com (Rick Moen) Date: Fri, 21 Dec 2001 13:40:03 -0800 Subject: [buug] Installing X on Dell with i810 chipset In-Reply-To: References: Message-ID: <20011221134002.N27273@linuxmafia.com> begin C. Sullivan quotation: > The Intel 810 chipset has a couple of unique issues You mean, like having no processing power whatsoever, stealing all video RAM from system RAM, and in general being just a few cents of glue circuitry? Configuring X11 can sometimes be a problem. But if it's on an i810, then you have _two_ problems. (Yes, I'm a good-hardware bigot. What's your point? ;-> ) -- Cheers, DISCLAIMER: I didn't do it. Rick Moen You can't prove it. Nobody rick at linuxmafia.com saw me. The sheep are lying. From rick at linuxmafia.com Fri Dec 21 13:43:00 2001 From: rick at linuxmafia.com (Rick Moen) Date: Fri, 21 Dec 2001 13:43:00 -0800 Subject: [buug] Installing X on Dell with i810 chipset In-Reply-To: References: Message-ID: <20011221134300.O27273@linuxmafia.com> begin Johnny Ho quotation: > I'm new to *nix and FreeBSD, and I've tried following the handbook but > still could not get X configure correctly. Has anyone gone through > this headache or is it just me. :( We feel your pain. My friend Google just whispered this in my ear: http://www.osfaq.com/article.php3?sid=67 But, honestly, I'd personally slip a nice Matrox G400 into one of those handy PCI slots, and forget the i810's even there. From rick at linuxmafia.com Fri Dec 21 13:44:21 2001 From: rick at linuxmafia.com (Rick Moen) Date: Fri, 21 Dec 2001 13:44:21 -0800 Subject: [buug] How to install the Redhat 7.2 in IBM Thinkpad iServer In-Reply-To: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> References: <004c01c18773$0d24aa20$0f01010a@stevenlaptop> Message-ID: <20011221134421.P27273@linuxmafia.com> begin Steven Liang quotation: > Hello every one: > does anyone know how to install the Redhat 7.2 in IBM's Thinkpad > iServer laptop? Almost certainly. That was a vague answer, for a vague question. Try again? From jammer at weak.org Fri Dec 21 13:41:48 2001 From: jammer at weak.org (Jon McClintock) Date: Fri, 21 Dec 2001 13:41:48 -0800 Subject: [buug] Can't get Linux to see more than 2 GiB on 75GXP In-Reply-To: ; from will_sargent@yahoo.com on Thu, Dec 20, 2001 at 01:56:13AM -0800 References: Message-ID: <20011221134148.C8147@weak.org> On Thu, Dec 20, 2001 at 01:56:13AM -0800, Will Sargent wrote: > I can't get my IBM Deskstar 75GXP to be recognized as more than 2 GiB under > Linux. I'm using the 30 GiB model (IBM-DTLA-307030). I've tried Mandrake > 8.1 and Redhat 7.1, and both distros will come up on boot recognizing the > drive type, but then assigning it as a 2 GiB drive. The CHS is printed as > 255/255/63, if CHS still has any meaning. > > I'm using an Asus A7V133 motherboard, revision 1004. This drive and > motherboard worked fine for me under Windows 2000 (so it's not a bad drive) > or outdated BIOS (which is unlikely). The Asus board includes an integrated > Promise ATA/100 controller, and I've tried using that instead of the > standard ide channels, but that doesn't seem to help (again, it recognizes > the drive on /dev/hde, but has the wrong CHS and number of sectors). The machine this system is running on is using pretty much the exact same configuration, except with a PCI Promise ATA/100 controller instead of the on-board version. I don't give it any extra boot parameters, and it works fine. You may need to upgrade to a 2.4.x kernel, which, AFAIK, is when they intregrated in the support for the Promise controller. Relevant sections of my dmesg: [It detects the IDE controller...] PDC20267: IDE controller on PCI bus 00 dev 50 PDC20267: chipset revision 2 PDC20267: not 100% native mode: will probe irqs later PDC20267: (U)DMA Burst Bit ENABLED Primary PCI Mode Secondary PCI Mode. ide2: BM-DMA at 0xfcc0-0xfcc7, BIOS settings: hde:pio, hdf:pio ide3: BM-DMA at 0xfcc8-0xfccf, BIOS settings: hdg:pio, hdh:pio ... [It detects the IBM drive...] hde: IBM-DTLA-307030, ATA DISK drive ... [Prints out the resources used by the IDE controller...] ide2 at 0xfc88-0xfc8f,0xfc9a on irq 11 ... [Prints out the geometry information about the drive...] hde: 60036480 sectors (30739 MB) w/1916KiB Cache, CHS=59560/16/63, UDMA(100) This is with a 2.4.0 kernel. -Jon From john at jjdev.com Fri Dec 21 14:38:14 2001 From: john at jjdev.com (johnd) Date: Fri, 21 Dec 2001 14:38:14 -0800 Subject: [buug] parsing a file Message-ID: <20011221223814.GA3499@vette.jjdev.com> Any suggestions? I have a transaction file...plain text. the first two lines I want excluded from the parsing (they are not transactions) I need to know about each line where the 6th field is '0' and the delimiter is ':' the first two lines don't have any delimiter characters, so if it would simplify things just to say parse lines that are delimitted, that would be ok. (cut has a flag for that) I came up with: cat trans | cut -d : -f 7 -s I'm thinking there's away to do it with some combo of standard commandline tools...else I'll just make a little script. the file looks like: 6685.23 1005 1000:john d:12/21/01:40.23:mem:cat:0 1004:bill d:12/21/01:45.00::bills:1 2000:sadie:12/21/01:23.00:dog food:pet:0 From jammer at weak.org Fri Dec 21 15:00:12 2001 From: jammer at weak.org (Jon McClintock) Date: Fri, 21 Dec 2001 15:00:12 -0800 Subject: [buug] parsing a file In-Reply-To: <20011221223814.GA3499@vette.jjdev.com>; from john@jjdev.com on Fri, Dec 21, 2001 at 02:38:14PM -0800 References: <20011221223814.GA3499@vette.jjdev.com> Message-ID: <20011221150012.E8147@weak.org> On Fri, Dec 21, 2001 at 02:38:14PM -0800, johnd wrote: > Any suggestions? > > I have a transaction file...plain text. > > the first two lines I want excluded from the parsing (they are not > transactions) > > I need to know about each line where the 6th field is '0' and the delimiter > is ':' > > > the first two lines don't have any delimiter characters, so if it would simplify > things just to say parse lines that are delimitted, that would be ok. (cut > has a flag for that) Well, if the 7th (I think you meant 7th) field is always the last field, then why not use grep? darkhawk:~/> grep ':0$' foo 1000:john d:12/21/01:40.23:mem:cat:0 2000:sadie:12/21/01:23.00:dog food:pet:0 That is, if you want the whole line; the grep output can always be piped into awk to pull out specific columns. You can also search on other columns: grep ':string:[^:]*$' # The second to last column grep '^string:' # The first column grep '^[^:]*:string:' # The second column man 1 grep for more information. -Jon From aris at pharoe.com Mon Dec 31 02:44:51 2001 From: aris at pharoe.com (Miki Shapiro) Date: Mon, 31 Dec 2001 12:44:51 +0200 (IST) Subject: [buug] Updating ports using cvsup Message-ID: <20011231123330.M19471-100000@pharoe.com> Hi all I have a FreeBSD 4.4 Release box installed from a mini distribution, with the ports collection. I went into /usr/ports/net/cvsup and did a "make install" Now I want to update my ports collection. I took the default cvs_supfile from the /usr/share/examples/cvsup directory, changed it to local settings and ran as in the handbook. It worked for a long while, then I had /usr/ncvs the relevant subtrees (src, ports, doc) full of files with ",v" appended to them. The whole procedure seems to have ended successfully. That's where the handbook just left off, and my /usr/ports is still not updated. I haven't the slightest clue what to do with all the new tree under ncvsup. (or how to convert those ,v files into what they're really supposed to be) 1. What do I do now? (some FM to RT would be wonderful) 2. Should I at some point remove the old /usr/ports and replace it with (a softlink to) /home/ncvs/ports ? 3. Where is updating the whole operating system (make world I think?) documented in a step-by-step guide? 4. When, regardless of the whole cvsup updating procedure, I go into /usr/ports/editors/pico and run "make install", where is the downloaded package saved on disk? Thanks a bunch! -- Miki Shapiro Unixophilic Software Developer --------------------------------------------- Tel: +972-(56)-322433 ICQ: 3EE853 --------------------------------------------- If at first you don't succeed... ... SkyDiving is probbably not for you. From lramos3 at satx.rr.com Mon Dec 31 20:40:50 2001 From: lramos3 at satx.rr.com (Luis) Date: Mon, 31 Dec 2001 22:40:50 -0600 Subject: [buug] Multiple problems Message-ID: <01123122405001.02817@MandrakeServer.NorthPole.com> I'm having several problems with my most recent attempt at running FreeBSD as my desktop software: 1-the os stoped booting and would stop when the last message was the discovery of the parallel port. This happened on an ASUS A7V board and neither openbsd nor Linux had this problem. Moving the scsi hd to an ASUS P2L97S board solved the problem. 2-kde won't with either 3.36 or 4.1, having installed using /stand/sysinstall, neither through startx or startkde. GNOME is running fine, as is xfce. In reference to GNOME, it was running the sawfish manager for a while and started to use twm after opening the configuration window for twm and clicking on try. After that, though, I could not get back to sawfish until, out of the blue, it booted by itself into sawfish and twm was nowhere to be seen. What can be the explanation? 3-cups would not install via the ports collection, had to download the tgz file and followed the instructions, including using gmake instead of make. However, the Canon 2100C bubblejet does not work and there is no Canon choice during printer setup. What can I do? Moving the printer to a Linux box was no problem, it worked fine. 4-can't use staroffice. had to do a package installation as the cdrom would not install the program. What more information do you need for me to solve this problems? What I'm looking for also, is systematic way of approaching these problems so I can solve them myself. For example, what steps would somebody have to take to be able to boot into kde or to diagnose the printer problem? Thanks. Luis R. Ramos, lramos3 at satx.rr.com From mjh at aciri.org Mon Dec 31 22:46:10 2001 From: mjh at aciri.org (Mark Handley) Date: Mon, 31 Dec 2001 22:46:10 -0800 Subject: [buug] Multiple problems In-Reply-To: Your message of "Mon, 31 Dec 2001 22:40:50 CST." <01123122405001.02817@MandrakeServer.NorthPole.com> Message-ID: <35215.1009867570@vulture.icir.org> >I'm having several problems with my most recent attempt at running FreeBSD as >my desktop software: 1-the os stoped booting and would stop when the last >message was the discovery of the parallel port. This happened on an ASUS A7V >board and neither openbsd nor Linux had this problem. Moving the scsi hd to >an ASUS P2L97S board solved the problem. 2-kde won't with either 3.36 or >4.1, having installed using /stand/sysinstall, neither through startx or >startkde. I'm running both FreeBSD 4.4R and 4.3R, and XFree86 3.3.6 and 4.1.0. I have a few Asus A7V boards running FreeBSD. And I'm running Kde 2.1 as I type this. You'd need to say what the error is with Kde for anyone to help. The A7V has a known problem with disks; both SCSI and IDE disks suffer corruption when writing large files. It's a BIOS problem, as Asus were too aggressive in their North Bridge tuning. There's a Bios fix available - when I came across this back in April, I flashed a7v10801b.zip, although that was a Beta release. It's still available from ftp://ftp.asus.com.tw/pub/asus/beta/mb/bios/a7v10801b.zip but there's probably a production release by now. I wouldn't use an A7V under any OS without updating the Bios. Cheers, Mark