--------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00007 Date: 01/18/98 From: DAVID DRUMMOND Time: 08:56am \/To: EDWIN HERMANS (Read 4 times) Subj: uhm, help. :) Edwin, at 09:19 on Jan 15 1998, you wrote to David Drummond ... HT>>> And.. how come Win95 doesn't recognize my 850MB Linux HD? DD>> Simple, Win95 cannot read ext2 partitions. EH> Well there are programs for w95 that can read ext2 (And ext. EH> too i believe :)) The one i use is called lread (along with a EH> proggie called ldir to do an ls from w95) I don't doubt that there are addons to allow it to view such partitions, but native Win95 does NOT. David --- Msgedsq/2 3.10 * Origin: The JabberWOCky +61 7 3868 1597 (3:640/305) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00008 Date: 01/18/98 From: PATRICK MORAN Time: 11:47am \/To: TOMER GABEL (Read 4 times) Subj: Finally installed Linux.. * In a message to All on 01-05-98 * TOMER GABEL said the following: TG> 1. How can I mount (or do anything similiar) a DOS partition > permanantly, something like a permanant: TG> mount -o defaults /dev/hda1 /dosc > mount -o defaults /dev/hdb1 /dosd mount -t msdos -o defaults /dev/hda1 /dosc If you want to be able to use it with Linux filenames as well them mount it as UMSDOS: mount -t umsdos -o defaults /dev/hda1 /dosc add any or all of these entries to your /etc/fstab file so it is done automatically during boot or with a noauto entry in the /etc/fstab file for a particular device, then you can simply use mount /dev/device or mount /mountdirectory for the device if it is the only device that will use that directory for mounting in the /etc/fstab. TG> 2. Is it possible, after I mount "mount -o defaults -o ro /dev/cdrom > /cdrom" to change a disc without unmounting and remounting? or am > I doing this all wrong? mount -t iso9660 -o ro /dev/cdrom /cdrom > 3. How the h*ll do I setup my drivers, namely: why can't I use half > my applications just because they can't "initialize display" or > something of this sort? How do I define what drivers (sound, > display, printer etc.) the kernel should or should not use? You can compile them into the kernel or use most of them as modules. Depending on what distribution you are using, this is done in many different ways and and can be set to do most of it suring the boot process. > 4. HOW DO I DEFINE A USER OF MY OWN? I have to loggin as "root" all > the time (using Slackware) and it isn't very professional, to say > the least. adduser Pat Moran Computer Services (505) 867-9146 Bernalillo, NM bugsy@nmol.com --- SPEED 2.00 #2650 It's more than a reader. It's a message base manager! * Origin: * Binary illusions BBS * Albuquerque, NM * 505.897.8282 * (1:301/45) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00009 Date: 01/18/98 From: PATRICK MORAN Time: 11:47am \/To: TOMER GABEL (Read 4 times) Subj: LINUX installation * In a message to Patrick Moran on 01-05-98 * TOMER GABEL said the following: PM> As long as the boot files and kernel are located in the first 1023 PM> cylinders, you will be okay. For the lilo boot manager, you can put it PM> in any partition within the first 1023 cylinders. Newer BIOS may not PM> have the restriction. You can always use loadlin from a FAT partition PM> and boot it that way, but you have to enter DOS first. I have not heard PM> if they have been able to get Linux to boot from WINDOZE yet. TG> I see. Annoyingly enough, Linux will not boot for some reason, so I > have to enter my bootdisk in order to boot into Linux... other than > that, normal Windows95/MSDOS booting process is run. Have you installed LILO yet? If not, I would install it in a boot partition and not the MBR (WINDOZE 95 likes to wipe out the MBR and put it's stupid crap in there!) Pat Moran Computer Services (505) 867-9146 Bernalillo, NM bugsy@nmol.com --- SPEED 2.00 #2650 if=Gates of=/dev/null * Origin: * Binary illusions BBS * Albuquerque, NM * 505.897.8282 * (1:301/45) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00010 Date: 01/18/98 From: PATRICK MORAN Time: 11:47am \/To: TOMER GABEL (Read 4 times) Subj: S3 chips * In a message to All on 01-09-98 * TOMER GABEL said the following: TG> I have some serious trouble configuring X and running various > applications, all because stupid SVGALIB won't recognize my S3 chip > although x-config has my video card (Diamond Stealth 64 VRAM) on its > list (my S3 chip is 968). Is there a problem with X-config, or just > outdated S3 drivers? if so, where can I get the newest? Good thing I read some more messages before answering your message to me. I had no idea what card you had. Diamond cards are not well supported for Linux becuase the idiots at Diamond want you to sign a non-disclosure agreement and will NOT allow source code for drivers to be distributed. Thus under the GNU license these drivers can't be distributed as GNU. Your card and chipset are supported. You will need to read the Diamond HOWTO or whatever that file is for that card. I have a Diamond Stealth VRAM and finally someone figured out what chips were being used and now it works like a champ. It will be the last Diamond product I will buy unless they change their stupid policies. You need to use the S3 server and not the SVGA server. These cards will not work with the SVGALIB (anyway mine won't.) The next card I buy will work with both an accelerated server and the SVGALIB and will be PCI. I haven't figured out which I will get yet. You probably cannot use probe with that card, so when asked to probe, answer no. With my card it says not to probe it and when that comes up on the install, it bypasses the probing. Read the Diamond and S3 docs and HOWTOs. This should get you going. Pat Moran Computer Services (505) 867-9146 Bernalillo, NM bugsy@nmol.com --- SPEED 2.00 #2650 ...WINDOZE a virus? No! Virii do something! * Origin: * Binary illusions BBS * Albuquerque, NM * 505.897.8282 * (1:301/45) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00011 Date: 01/18/98 From: PATRICK MORAN Time: 11:47am \/To: PER K HANSEN (Read 4 times) Subj: uhm, help. :) 1/2 * In a message to Hans Tegnerud on 01-16-98 * PER K HANSEN said the following: HT> This sounds like a software solution to bad HD controllers. HT> Kind of like.. the info of partitions 'inside' the extended HT> partition isn't contained at the 'ususal' place for such info HT> but HT> at the beginning (or end) of the extended partition itself. PKH> It isn't a sulution for bad controllers, but i don't know why it > has the 4 primary limit, it might have been a max usable on a 200 > MB disk, but now with 6.2 GB it is something else ;) PKH> I don't exactly know how the extendend partition information is PKH> stored. I'll try and explain this to everyone. I am sure many people are curious about this. First let me state that in this discussion that as far as I am concerned, the first sector on a track is sector 0, but for some uncomprehending reason, the idiots that wrote MSDOS chose sector 1 as the first sector on a track (drive, cylinder, or whatever.) So if I use sector 0 or sector 1 in this discussion, it means the very first sector. I will try and use sector 1, but I learned DOS before IBM/MS/PC DOS was even around (Apple DOS) and from habit, I think it terms of counting from 0 with computers. This goes way back to when I went to school for IBM 360 computers and is very hard to change, just because Gates wants to run the world his way. Also, for this discussion, DOS will also refer to WINDOZE as well. Let's say we have a hard drive with 1,000 cylinders, 10 heads and 50 sectors per track and 512 bytes per sector. Let's divide this drive up into 5 partitions, each will be 200 cylinders. The first partition will be on cylinders 0-199, the second 200-399, the third 400-599, the fourth 600-799 and the fifth 800-999. The MBR will be located in the very first sector of the drive, that is sector 1 of head 0 in cylinder 0. The last 2 bytes of this sector will have the IBM identifier code of 55 AA (hex.) The 64 bytes immediately before this identifier code will consist of 4 entries of 16 bytes each for the partition table. The remaining 446 bytes in this sector will not be of concern in this discussion. It is code needed by the BIOS and most of the space isn't even used for anything. NOTE: This is the same identifier code that is also used in the last two bytes of the boot record. (Non-DOS boot records do not have to have this identifier and as far as I know, the coding was never completed to allow additional identifiers. The way I heard it, many years ago, was that IBM DOS would be able to recognize other types of file systems, but the idiots that wrote stupid easily crackable so-called copy protection schemes, asked IBM not to do this. Thus the additional coding for identifiers was not even implemented AFAIK. Residue of this capability is still embedded in the boot sector, i.e. sector size and other info. Had this code been implemented, you would have been able to directly read data files from Apple DOS, CP/M, Mac, and other file systems.) The four entries of 16 bytes will contain information about where the partitions are located and the type of partition and a boot flag. The partition information will include the starting relative sector and length of the partition in relative sectors. Most disk editors and FDISK programs will indicate the location in terms of cylinder, head and sector from this table. For use with DOS, one partition must be a primary partition. DOS FDISK will not even allow you to make and extended partition until you have made a primary partition. (Other FDISK programs will allow an extended partition without a primary.) From the above information about this drive, the MBR will be located in sector 1, head 0, cylinder 0. This is also known as track 0, but only the first sector of this track is used by DOS FDISK. The next four partitions will be similar to the MBR and their partition information will be located in the very first sector of that partition. Thus: partition 2 will have it's partition information in: sector 1, head 0, cylinder 200. Partition 3's info will be located in sector 1, head 0, cylinder 400, 4's: sector 1, head 0, cylinder 600, and 5's: sector 1, head 0, cylinder 800. If you have a disk editor such as Norton's or PCTools, you can look at this on your own hard drive, just substitute the location for the ones I am using. Make sure the disk editor is in read-only mode unless you know how to use them. This is the default configuration for Norton and PCT. Let's make this drive with 1 primary partition and an extended partition of 4 logical drives. the extended partition will be located from sector 1, head 0, cylinder 200 to sector 50, head 9, cylinder 999 and the primary partition in this example will be located from sector 1, head 0, cylinder 0 to sector 50, head 9, cylinder 199. This will be designated as drive C: by DOS. For this discussion, all the logical partitions will be DOS partitions. (This is for ease of understanding, the principle is the same for all partition types in the extended partition.) The entries for the MBR will be two. One for the primary partition and one for the extended partition. The other two entries will consist of zeros (thus unused.) The first entry will show a primary partition starting at sector 1, head 1, cylinder 0. NOTE: head can also be refered to as side i.e. side = head. So in this case side 0 of cylinder 0 has [ Continued In Next Message... ] --- SPEED 2.00 #2650 Gee officer, I just wondered how fast SPEED READ was. * Origin: * Binary illusions BBS * Albuquerque, NM * 505.897.8282 * (1:301/45) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00012 Date: 01/18/98 From: PATRICK MORAN Time: 11:47am \/To: PER K HANSEN (Read 4 times) Subj: uhm, help. :) 2/2 [ ...Continued From Previous Message ] been assigned to the MBR. Side 1 is where the boot record for the first partition is located or in other words sector 1, head 1, of cylinder 0. The second entry in the MBR will point to the extended partition, starting at sector 1, head 0, cylinder 200 and ending at sector 50, head 9, cylinder 999. Drive D: This location will also have an IBM identifier code of 55 AA in the last two bytes and will have two entires in the 64 bytes immediately preceeding the identifier code. The rest of this sector will be zeros and contain no information. The first entry will point to the location of this partition which will begin at sector 1, head 1, of cylinder 200. This is where the boot record of this partition is located. (Note the rest of side 0 is wasted just as with the MBR.) The second entery will be an extended entry pointing to the next logical partition, i.e. partition 3 or drive E: I do not recall exactly what information is contained in this entry. It will either be starting with the next partition's first sector and end with that next partitions last sector or the last sector of the extended partition. i.e. it will start with sector 1, head 0 cylinder 400 in both cases and end with either sector 50, head 9, cylinder 599 or sector 50, head 9, cylinder 999. Each additional logical drive will have similar entries in the very first sector of that partition and will actually start the data portion of each partition on side 1 with a boot record. The last logical partition will only have one entry in it's first sector. If we had included another primary partition, the first sector would be the boot sector. There would be no entries for it's location or the location of any other partitions, just the boot record. NOTE: DOS must start on a cylinder boundry, i.e. sector 1 head 0, cylinder xxxx. DOS boot record must be located in sector 1, head 1, cylinder xxxx. This is not the case for all OS's, Linux can start a partition on any head, cylinder. Linux and some other OS's can use the space wasted by DOS requirement for wasting the remainder of side 0 for each partition. Linux FDISK uses a maximize option to use the rest of this space. Since DOS is so stupid and doesn't even know of these partitions and can't read them, there is no reason to follow stupid DOS rules. For example, on my hard drive the second partition (hda2) is a primary partition for my Linux root partition and LILO resides in the very first sector of this primary partition. After all this has been said, you can actually have as many partitions on a hard drive as you want. There is no limit. For the hell of it I made 3 primary partitions and an extended partition (you can only have one extended partition) and 25 logical partitions on a single hard drive. Of course when I made all of the logical partitions, DOS partitions, FDISK ran out of letters, but they don't have to be DOS logical partitions, for example I made Linux partitions above drive Z. I used Caldera's OpenDOS FDISK v2.00 for this test. Linux IDE can have 20 partitions per drive (this is limited by the device entries in /dev) and 16 for SCSI. However, it would appear from the Linux device table that the 16th SCSI partition will link it up to the next SCSI drive, if one is present. i.e. sda16 device designation is 8,16 and sdb device designation is also 8,16. This is probably how several SCSI drives can be linked together and appear as a single device for up to a 2 terabyte filesystem. I have not read the docs about this, in-depth, yet. Thus on a single physical hard drive I could have a DOS primary, an OS/2 Boot Manager primary, a linux swap primary partition, 15 more Linux partitions, 24 DOS/OS/2 logical drives and other logical partitions for some other OS. Of course the 24 logical DOS and/or OS/2 partitions would have to be reduced for each additional drive added to the system, such as CDROM, RAM disk drives, etc. Of course all partitions located beyond what an OS can use will not be seen by that OS. (DOS may have a limit of 16 partitions on a single drive. I didn't think to check if DOS could see the partitions all the way to Z: and there was no way to let it install the CDROM drive because C: through Z: had already been assigned. Novell Netware and possibly other systems have a work around for the 26 DOS limit.) An interesting note: When I made than many partitions and started FDISK and selected to display the logical partitions, I could hear the heads do a bunch of seeks to get the information. Pat Moran Computer Services (505) 867-9146 Bernalillo, NM bugsy@nmol.com --- SPEED 2.00 #2650 Gee officer, I just wondered how fast SPEED READ was. * Origin: * Binary illusions BBS * Albuquerque, NM * 505.897.8282 * (1:301/45) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00013 Date: 01/18/98 From: PER K HANSEN Time: 07:30pm \/To: GERRIT KLEIN (Read 4 times) Subj: Linux CD'S Hej Gerrit Den 18 Jan 98 skrev Gerrit til Per og selfoelgelig skulle jeg blande mig. PKH>> Den 12 Jan 98 skrev Gerrit til Per og selfoelgelig skulle PKH>> jeg blande mig. GK> That isn't very easy to understand, but i now what it means, i GK> beleef GK> Op 12 jan 98 schreef Gerrit aan Per, that is all i can figure GK> out. Well, I havn't configured my reader for international use ;) it says: On the 12 Jan 98 Gerrit wrote to Per and ofcourse I had to break into the conversation. ;) PKH>> Sorry, I was wrong. It is one my old dec. 1996 CD set, but PKH>> not on your new one ;) GK> OK accept, all i found was a mini HOWTO with the right name, i GK> have to wait for the next CD set, downloading takes to much time, GK> while the newest version is 47 MB. I hope redhat 5.0 is on the next Infomagic set. Med venlig Hilsen Per K Hansen (perkhan@usa.net) --- FEddi 0.9pl7 via ifcico * Origin: FEddi (C)1993-96 by Oliver Graf (2:235/203.61) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00014 Date: 01/18/98 From: PER K HANSEN Time: 07:45pm \/To: DANIEL PARREIRA (Read 4 times) Subj: Re: Linux and AmigaDOS Hej Daniel Den 18 Jan 98 skrev Daniel til Per og selfoelgelig skulle jeg blande mig. DP> Yeah, but IIRC the homepage now uses Universal. That is also more fitting now. DP> It's very usable for something many people ever think wasn't DP> possible. I know, i have tried it myself. (Funny to look at amigafreaks when they see that their 'god' runs together with C64 emulators ;) ) Med venlig Hilsen Per K Hansen (perkhan@usa.net) --- FEddi 0.9pl7 via ifcico * Origin: FEddi (C)1993-96 by Oliver Graf (2:235/203.61) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00015 Date: 01/18/98 From: PER K HANSEN Time: 08:00pm \/To: JAN WAGEMAKERS (Read 4 times) Subj: Linux and AmigaDOS Hej Jan Den 18 Jan 98 skrev Jan til Per og selfoelgelig skulle jeg blande mig. EF>> Well.. most probably as mount -t amifs /dev/fd0 /mnt JW> So, with the above information in mind, I think that this will JW> not work. It won't on a floppy, but a harddisk will work fine. Med venlig Hilsen Per K Hansen (perkhan@usa.net) --- FEddi 0.9pl7 via ifcico * Origin: FEddi (C)1993-96 by Oliver Graf (2:235/203.61) --------------- FIDO MESSAGE AREA==> TOPIC: 241 LINUX Ref: F1P00016 Date: 01/18/98 From: PER K HANSEN Time: 08:06pm \/To: DANNY CAUTAERT (Read 4 times) Subj: FEddi serious problem :( Hej Danny Den 18 Jan 98 skrev Danny til Per og selfoelgelig skulle jeg blande mig. DC> How did you do that? I have no idea :( It just happened one day, when i was tossing my mailpacket. [CUT] DC> 7) Run 'futility rescan NETMAIL' to correct the marker-counts. DC> Done. Yes, I figured that out by myself (I couldn't live without netmail :( ) Isn't there a way to change the flags of the msg base ? futility can list them, but not change them. Med venlig Hilsen Per K Hansen (perkhan@usa.net) --- FEddi 0.9pl7 via ifcico * Origin: FEddi (C)1993-96 by Oliver Graf (2:235/203.61)