Configuring Login.App
I have been using the Login.App program and like it better than the default login. But when it runs X, it runs it in 8 bit color mode. Is there a way to configure this differently? I always used the Xservers file in /etc/X11/xdm before, but Login.App doesn't seem to have a comparable file. [EMAIL PROTECTED]/GNU--2.1---Linux--2.2.1--- "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
GTK wants g_strcasecmp symbol
When I try to run "gimp" it balks and says that it can't link because the GTK library wants a symbol. Any idea what is wrong? gimp: error in loading shared libraries /usr/lib/libgtk.so.1: undefined symbol: g_strcasecmp I have the following from LDD: $ ldd `which gimp` libgtk.so.1 => /usr/lib/libgtk.so.1 (0x40011000) libgdk.so.1 => /usr/lib/libgdk.so.1 (0x400a5000) libglib.so.1 => /usr/local/lib/libglib.so.1 (0x400c5000) libXi.so.6 => /usr/X11R6/lib/libXi.so.6 (0x400cd000) libXext.so.6 => /usr/X11R6/lib/libXext.so.6 (0x400d5000) libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x400e3000) libm.so.6 => /lib/libm.so.6 (0x40186000) libc.so.6 => /lib/libc.so.6 (0x4019f000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x4000) All of the libraries are there. Is something out of sync? [EMAIL PROTECTED]/GNU--2.0 (frozen)---Linux--2.0.30--- THE LESSER-KNOWN PROGRAMMING LANGUAGES #12: LITHP This otherwise unremarkable language is distinguished by the absence of an "S" in its character set; users must substitute "TH". LITHP is said to be useful in protheththing lithtth. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
genclass, C++ and the new (GNU?) egcs stuff
I have been using the old G++ for some time and recently found that upgrading has caused the genclass script, String library, and other GNU stuff to disappear. Is there a package I am missing? Do I need to switch over to the STL? I am using the genclass types in a project for school -- The school Sparc labs have it but I don't know if they have STL yet. What packages do I need to downgrade to get genclass back, or is there some reason I should switch to STL? More importantly, is there a "standard" equivalent to the String class from the old libg++ -- my project depends heavily on the regex stuff. TIA, Kevin Bealer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
smail configuration?
I am trying to configure smail. I have read the documentation, searched the source code for an hour or so, and tried to use the generic configuration script at different settings. I have been at this for several days -- my mail is still bouncing. I am on a system which does not have a permanent connection, rather it has a dynamic IP through Penn State. When the mail bounces, it comes back with a message to the effect that the hostname in question does not exist. This is true of course, since it is the hostname of the local system, which is a dial up dynamic IP thing. When using the debian "configure smail" script, there are 5 or so options. The UUCP to smarthost seems to be the closest, except that instead of UUCP I want to use smtp. Unfortunately, there is not an option like this, and although I have tried to modify that setup to do the job, it seems to do about the same. Some hosts bounce, some do not -- I assume hosts that are trying to prevent spam etc are bouncing more than those that aren't. People have to be using it this way... what is the trick? It shouldn't make any difference, but I am using (mh). Thanks, Kevin Bealer -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
nmh and mail spool permissions
I installed the "nmh" program, and almost everything seems to work now, except that the program "inc" hangs, or at least hangs around for a while and then dies. After looking at the "strace" I see it is trying to create lockfiles in /var/spool/mail/ which is forbidden by the permission bits. It works fine if I make the permissions so that anyone can write to the directory. That is, anyone can delete/create anyone's mail file. It seems like this should not be -- what is the correct permissions for /var/spool/mail? Thanks, [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- The meta-Turing test counts a thing as intelligent if it seeks to devise and apply Turing tests to objects of its own creation. -- Lew Mammel, Jr. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Gigabyte 586S and Cyrix 200+ (was Dual Pentium Machines)
Jason Gunthorpe wrote: > > >On Thu, 17 Jul 1997, Kevin M. Bealer wrote: > >> After trying all sorts of measures, I have found one that seems so far >> to reliably fix the problem: run it with the case off. Since there is >> unlikely to be any "grounding problems" because of the way the case is >> laid out (its a full tower, the case is really just a "wrapper" for a >> solid frame, and touches no components), it is almost certainly >> thermal. > >Hehe, bet the gigabyte isn't using a switching power supply :> Find the >voltage regulators, very large heat sinks off the MB and touch them, they >should be excessively hot (Cant keep your finger on them). You might want >to put a fan near/on/over them to cool those puppies off. I know my MB has >3 large heatsinks on the voltage regulators, they get hotter than the CPU! > >I also have a P5 MB with a switching power supply built in, it has no heat >sinks and still carries the same load.. > >Jason > The system has been up for "32 minutes" according to uptime, and they are actually /not/ noticeably hot, just slightly above "cold metal" temperature. Large shiny pieces of copper-colored metal, large enough to /look/ like they should be hot. But right now, they're not. OTOH, maybe during a kernel compile they warm up. The set6x86 thing sets the chip to "suspend on halt" which is supposed to cool things down a lot. Total CPU usage since bootup is 5.8 %. So far it is "behaving itself" now that I have moved the ribbon cables from directly in front of the fan. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- *Don't* background it, just sit there looking at it. Tell everyone "I'll get back to you as soon as this here job is finished". Very relaxing. -- Lars P. Fischer -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Gigabyte 586S and Cyrix 200+ (was Dual Pentium Machines)
"Kevin M. Bealer" wrote: >"Kevin M. Bealer" wrote: >(clipping orig post and most of reply...) >> (again...) >(grinding teeth) > >Kernel compiles are failing (twice). No signal 11's, just the compile >complains about things as though random characters were inserted into >a compile pipe or something. Different errors, different files, >rebuild fixes the problem both times.. > >Anyway, I'll shut up for now. > >[EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- >"You are in a maze of twisty little passages, all alike." > Ok, I'm going to waffle one more time: When compiling before, I could reliably hang or oops the machine simply by running continuous "make"s of the kernel. By reliably I mean that the compile would fail, on the MAJORITY of the compiles, i.e. usually about 1/2 way through the process. After trying all sorts of measures, I have found one that seems so far to reliably fix the problem: run it with the case off. Since there is unlikely to be any "grounding problems" because of the way the case is laid out (its a full tower, the case is really just a "wrapper" for a solid frame, and touches no components), it is almost certainly thermal. So, for the moment I retract what I said earlier, but I may waffle again if I can't fix it by proper ventilation (I didn't realize how critical it was..) Now, if I could just get the serial ports to work so I can use my 3-b mouse... [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- Any sufficiently advanced operating system is indistinguishable from Un*x. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Gigabyte 586S and Cyrix 200+ (was Dual Pentium Machines)
Jason Gunthorpe wrote: > >On Tue, 15 Jul 1997, Kevin M. Bealer wrote: > >> "Kevin M. Bealer" wrote: > >> Following up on this: >> >> After: >> >> 1) installing set6x86 >> 2) Setting all BIOS settings to maximum delay timings (for RAM) >> 3) Setting the ISA bus frequency to 7.159 MHz (lowest setting) >>(only my CDROM, modem, and ethernet card are on the ISA anyway.) >> >> Everything seems to be stable. The serial ports don't work as >> far as I can tell, but the modem works fine, and the PS/2 mouse >> works fine. > >Just a suggestion, but if you bought a board it should be able to run at >'normal' clock rates and not require adjusting like that. IMHO you have >some kind of faulty hardware if this is required. You might have some >device/memory/whatever that doesn't like the 75MHz bus speed however... > >I had a board that would nuke Win95 in some strange cases, mostly due to >PCI bustmastering it seemed, turned out to be bad ram -- though the ram >didn't fail any of the tests the supplier put it through after I told him >about it. I'm sure if I put linux on it I would see symptoms like you do, >crashing during kernel compiles and such. > >Is this a 430TX based MB? > >Jason > Nah, it's a Gigabyte GA-586S, based on the SiS5571 chipset, Award BIOS. 32 MB EDO, 512K cache plus whatever the internal cache on a 6x86L 200+ is. It is on the Cyrix "recommended" list as well, and has specific documentation for setting up with this CPU (Cyrix 200+ has a 75 MHz bus speed; if your MB runs it at 60 MHz bus, you effectively have a 200 * (60/75) = 160.) I guess it's going back... I don't know whether I want to get another clone on another kind of board, or go to intel and get a Pent 120 for the same price. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- You have new mail in /dev/null -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Gigabyte 586S and Cyrix 200+ (was Dual Pentium Machines)
"Kevin M. Bealer" wrote: (clipping orig post and most of reply...) > >I would suggest subscribing to linux-smp, and look at the archives for >the list, too. I asked this question a week or two ago and then >decided that the Cyrix 200+ with a Gigabyte 586S board was the best >for the money(*). > (clipping info on SMP) > >(*): the board seems to have trouble with serial mice, and probably >serial anything. It has two PCI devices that are "unknown PCI >devices" to linux. The board has support for the Cyrix 200+ which >means it can generate a 75 MHz bus clock, without which the 200+ >becomes a 160+ to 170+ or thereabouts. > >Also, the board and CPU segfault without the Cyrix patch, and also >sometimes with it. I have been playing with the RAM settings to see >if I can get it more stable; so far it seems good, but I haven't >tested it much yet. You might want to email me before you buy one of >these, I am going to do some more testing tonight. Also see the other >sources above of course. > (now responding to self ...) Following up on this: After: 1) installing set6x86 2) Setting all BIOS settings to maximum delay timings (for RAM) 3) Setting the ISA bus frequency to 7.159 MHz (lowest setting) (only my CDROM, modem, and ethernet card are on the ISA anyway.) Everything seems to be stable. The serial ports don't work as far as I can tell, but the modem works fine, and the PS/2 mouse works fine. I did a "make -j bzImage" and made a 2.1.43 kernel. This went flawlessly, using all 32 MB of RAM and 70+ MB of swap at peak (I think). The generated kernel boots fine. Summary: if you want a Cyrix 200+ system you can get one on this motherboard for 220$ american, and it seems very quick. You may have to fiddle with the BIOS, all I did was set the ISA bus speed to slow as possible, and set the RAM timings that I could see to slow as possible (highest numbers.) (grinding teeth) Kernel compiles are failing (twice). No signal 11's, just the compile complains about things as though random characters were inserted into a compile pipe or something. Different errors, different files, rebuild fixes the problem both times.. Anyway, I'll shut up for now. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- "You are in a maze of twisty little passages, all alike." -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Dual Pentium Machines
Greg Vence wrote: >Hello, > >Its time to upgrade... What dual Pentium machines do people like and >what have known problems? I looked in LDP and didn't see anything >regarding this topic. > >Thanx -- Greg. > I would suggest subscribing to linux-smp, and look at the archives for the list, too. I asked this question a week or two ago and then decided that the Cyrix 200+ with a Gigabyte 586S board was the best for the money(*). Look at http://www.linux.org.uk/SMP/title.html for SMP Linux info, or look at http://www.linuxhq.com/lnxlists/linux-smp/ for mailing list archives. If you ask for general information on motherboards, you will probably be asked to look on the web for information. The place I was refered and re-refered to was http://sysdoc.pair.com/mainboard.html, but chop off the filename for lots of general hardware information. > >-- >TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to >[EMAIL PROTECTED] . >Trouble? e-mail to [EMAIL PROTECTED] . (*): the board seems to have trouble with serial mice, and probably serial anything. It has two PCI devices that are "unknown PCI devices" to linux. The board has support for the Cyrix 200+ which means it can generate a 75 MHz bus clock, without which the 200+ becomes a 160+ to 170+ or thereabouts. Also, the board and CPU segfault without the Cyrix patch, and also sometimes with it. I have been playing with the RAM settings to see if I can get it more stable; so far it seems good, but I haven't tested it much yet. You might want to email me before you buy one of these, I am going to do some more testing tonight. Also see the other sources above of course. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- $ sync sync: error, already sinking -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Enlightment packaging
Amos Shapira wrote: >Hi, > >I though I saw that someone was packaging the Enlightment window manager >(http://www.cse.unsw.edu.au/~s2154962/enlightenment, US mirror at >http://www.kickit.org/enlightenment/), but I can't find it in >hamm's Packages list. > >Does anyone know anything about this? > >Thanks, > >--Amos > >--Amos Shapira| "Of course Australia was marked for >133 Shlomo Ben-Yosef st. | glory, for its people had been chosen >Jerusalem 93 805 | by the finest judges in England." >ISRAEL [EMAIL PROTECTED] | -- Anonymous (as I heard it...) Enlightenment has a license that prevents it's migration to certain operating systems, which is a restriction on derived works, which makes it not "free software". It could be packaged into non-free however, as I understand it. This would discourage people from doing the packaging, however, there may still be someone doing it. >-- >TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to >[EMAIL PROTECTED] . >Trouble? e-mail to [EMAIL PROTECTED] . > > [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- $ sync sync: error, already sinking -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: XF86 3.3-3
Christopher Jason Morrone wrote: (clip) > >In any case, the XF86 3.3 packages are definitely broken. I'd fill out a >bug report, but unfortunately, I can't be any more specific than that >right now. > (clip) > >-- >TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to >[EMAIL PROTECTED] . >Trouble? e-mail to [EMAIL PROTECTED] . > > I found my keyboard/screen, and possibly the whole thing (don't remember) would lock up when I had an S3, but adding the "nolinear" option fixed the problem. This was with the older version, i.e. it happened with 3.1 or 3.2. However, other things changed with my config at 3.3, so maybe you might want to try turning options like this on and off. Video was an S3 on the mainboard, now I have upgraded both and "xf86config" got me a fairly good config, although my old compaq monitor can't keep up to the new video card frequency wise. (I know this post is a little light on the solution side, but noone seems to have "definitely hit the nail" yet, so.) [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- "You are in a maze of twisty little passages, all alike." -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Segmentation fault with mh
(about mh segfaulting) (sorry, but note these are mostly all general tips, some I have learned recently to help with diagnosing these kinds of problems.) What do you get for: dpkg -l | grep mh and ldd `which inc` I have been using mh for some time, and it seems to work fine here, granted, I have not "started an installation" with "inc" for some time, since I now have, well, enormous amounts of mail in my folders. The other thing to try is: strace -f `which inc` Which will show system calls that are being done to complete the "inc", and if there is poorly written code, it might not be checking the return value of system calls. If a call to make a directory fails right before the segfault, you might look into permission, for example. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- Programming is always harder than doing the same task manually. It's hard because you must completely understand the problem, take everything into consideration, and protect against every possible flaw. Then you never have to do it manually again. The great driving force behind programming is the fervent devotion to laziness: you wage a war to save typing later. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Afterstep problem solved
"Christian Hudon" wrote: > >On Jul 7, Jason Westervelt wrote > >> nah.. that only gets a few other icons to pop up... don't ask why, but >> the TOP button (the one that has the lock buttons, shutdown, etc) will >> ** NOT ** have an icon until you install procps and xproc.. very >> weird.. > >Even if you don't "swallow" an xload into your afterstep dock? Now, *that* >would be odd. > > Christian > Speaking of xload, where did it go? I'm sure I saw it is still available somewhere, but why doesn't it come standard with the rest of X anymore? [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- Programming is always harder than doing the same task manually. It's hard because you must completely understand the problem, take everything into consideration, and protect against every possible flaw. Then you never have to do it manually again. The great driving force behind programming is the fervent devotion to laziness: you wage a war to save typing later. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Well tested mode lines fail with 1.3.1
Brandon Mitchell wrote: >Can anyone verify that well tested mode lines, used with older X servers, >fail with the latest X servers in 1.3.1? Sorry for making this so long, >but I didn't want to leave anything out. > >Here's the info: >--- dpkg --list xserver-s3: >||/ NameVersionDescription >+++-===-==- >ii xserver-s3 3.3-3 X server for S3-based graphics cards > >--- Error message: >(--) S3: Diamond Stealth BIOS found >(**) S3: videoram: 1024k >(**) S3: Ramdac type: s3_trio64 >(--) S3: Ramdac speed: 135 MHz >(**) S3: Using Trio32/64 programmable clock (MCLK 59.957 MHz) (clip) This happened here as well, and I also have an S3. I had my old modes configured, but I had to readjust the monitor every time I switched text -> X -> text -> etc. Now X rejected those modes, so I edited a "close" modeline into a working one as per "Videomodes.doc" and now it works, and the screens text->X match up. I had given up on getting the old mode to line up, but when I redid the config, it seemed fairly easy. Did I ... a) know more, b) get the better luck of the draw, or c) a bug was fixed, and the new server guided me by rejecting my "broken" config? By the way, my card is not a Diamond Stealth, but rather an on-the-motherboard S3 on an IBM motherboard. Quite fast for a 486, but this weekend the new MB and the Matrox go in (hopefully). I think mine is the 805 chipset, if anyone is trying to figure out what happened. Maybe the changelogs say something. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- Programming is always harder than doing the same task manually. It's hard because you must completely understand the problem, take everything into consideration, and protect against every possible flaw. Then you never have to do it manually again. The great driving force behind programming is the fervent devotion to laziness: you wage a war to save typing later. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: ncurses3.4
Carey Evans wrote:
>"Kevin M. Bealer" <[EMAIL PROTECTED]> writes:
(clip)
>
>So you _do_ need the /lib/libreadline.so.2 link. bc in unstable is a
>libc6 program, so it shouldn't be concerned with what is in the
>libc5-compat directories.
>
>% dpkg -S libreadline.so.2
>libreadline2: /lib/libc5-compat/libreadline.so.2
>libreadlineg2: /lib/libreadline.so.2.1
>libreadline2: /lib/libc5-compat/libreadline.so.2.0
>libreadline2: /lib/libc5-compat/libreadline.so.2.1
>libreadlineg2: /lib/libreadline.so.2
>
>Is libreadlineg2 still installed properly after your problems with it?
>
>--
>Carey Evans <*> [EMAIL PROTECTED]
>
> "Our mail program accidentally deleted our remove list."
>- Real quote from UCE
Thanks for the help, all. "bc" works now. I don't understand how it
determined what libs it needed, i.e. why it refuses to link with
same-named libs in the same place, or where any of that is configured.
However, the appropriate libs now seem to be installed and work ok.
It would be nice if packages didn't leave Incoming until the packages
they depend on leave incoming, but I imagine the people who do that
stuff have enough complexity without worrying about that.
[ offtopic: ]
My new computer parts are in the mail "today". Cyrix 200+ ("6x86L"),
Gigabyte MB, full tower case, 32 MB EDO, and a Matrox Mystique w/ 4 MB
SDRAM for $535.
[EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30---
Programming is always harder than doing the same task manually.
It's hard because you must completely understand the problem,
take everything into consideration, and protect against every
possible flaw. Then you never have to do it manually again.
The great driving force behind programming is the fervent
devotion to laziness: you wage a war to save typing later.
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: ncurses3.4
Carey Evans wrote: >"Kevin M. Bealer" <[EMAIL PROTECTED]> writes: > >[snip] > >> Also, installing this packages causes dselect to get really excited >> about dependencies, since everything wants the old libreadline, and >> the new one doesn't want to coexist. What is the solution to this >> sort of thing? For now, I have left the new one installed, >> deselected it, and stopped upgrading for a bit, but "bc" doesn't >> run yet. > >Get the new, old libreadline2 from the same place as libreadlineg2: > (clip) Thanks, I don't know why I didn't see that before. However there is still a problem : $ bc bc: error in loading shared libraries libreadline.so.2: cannot open shared object file: No such file or directory $ ldd -v `which bc` ldd: version 1.9.2 libreadline.so.2 => not found libncurses.so.3.4 => /lib/libncurses.so.3.4 (0x4000f000) libc.so.6 => /lib/libc.so.6 (0x40054000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x4000) $ ls -la /lib/libc5-compat total 132 drwxr-xr-x 2 root root 1024 Jul 6 01:36 . drwxr-xr-x 4 root root 3072 Jul 6 01:36 .. lrwxrwxrwx 1 root root 18 Jul 6 01:36 libreadline.so.2 -> libreadline.so.2.1 lrwxrwxrwx 1 root root 18 Jul 6 01:36 libreadline.so.2.0 -> libreadline.so.2.1 -rwxr-xr-x 1 root root 129716 Jun 23 17:44 libreadline.so.2.1 I have, several times, run "ldconfig -v" as root. The directories /usr/lib/libc5-compat /lib/libc5-compat are both in /etc/ld.so.conf So why is it doing this? If I do an $ strace -f ldd -v `which bc` It doesn't seem to try to open /usr/lib/libc5-compat Also note that there were spurious links to libhistory.so and libreadline.so in the "regular" directories, i.e. /usr/lib and /lib, presumably from the last version. Removing these did not help. The /etc/ld.so.cache had a modification date from the last time I run ldconfig -v. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- ACHTUNG! Das machine is nicht fur gefingerpoken und mittengrabben. Ist easyschnappen der springenwerk, blowenfusen und corkenpoppen mitspitzensparken. Ist nicht fur gewerken by das dummkopfen. Dasrubbernecken sightseeren keepen hands in das pockets. Relaxen undvatch das blinkenlights!!! -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: ncurses3.4
Peter Mutsaers wrote: >I see a lot of packages appearing in unstable that depend on >ncurses3.4. But no packages providing it is available (only >ncurses3.0). > >Can anyone tell me where I can get ncurses3.4? > >-- > /\_/\ >( o.o ) Peter Mutsaers | Abcoude (Utrecht), | Trust is a good quality > ) ^ ( [EMAIL PROTECTED] | the Netherlands| for other people to have > > >-- >TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to >[EMAIL PROTECTED] . >Trouble? e-mail to [EMAIL PROTECTED] . You can get it from ftp.i-connect.com, in Incoming, and info, et al will run again. On a similar note, the package libreadlineg2 is wanted by both bc and dc, and maybe others. I have installed the package (from i-connect) but the programs still seg-fault. Also, installing this packages causes dselect to get really excited about dependencies, since everything wants the old libreadline, and the new one doesn't want to coexist. What is the solution to this sort of thing? For now, I have left the new one installed, deselected it, and stopped upgrading for a bit, but "bc" doesn't run yet. (so I use python to do arithmetic.) [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- I see the eigenvalue in thine eye, I hear the tender tensor in thy sigh. Bernoulli would have been content to die Had he but known such _ a-squared cos 2(phi)! -- Stanislaw Lem, "Cyberiad" -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: gstep-base library use boundary="==_Exmh_12596774180"
"Graham C. Hughes" wrote: >I've been interested in learning about Objective C, and figured I'd use >the gstep-base libraries packages with Debian. > >My question is fairly simple: what link flags am I supposed to use? > >A sample session is MIME-attached. Hints and suggestions would be really >appreciated, eh? ;-). > >BTW, I'm using libc6-dev here. > (clip) (Note that I have taken the liberty of clipping the "mimery" from your message. It is usually handled funny by my copy of "mh") I have found the following in the GCC info system: `FILE.m' Objective-C source code. Note that you must link with the library `libobjc.a' to make an Objective-C program work. `FILE.h' C header file (not to be compiled or linked). (... clipping info on other extensions ...) You can specify the input language explicitly with the `-x' option: `-x LANGUAGE' Specify explicitly the LANGUAGE for the following input files (rather than letting the compiler choose a default based on the file name suffix). This option applies to all following input files until the next `-x' option. Possible values for LANGUAGE are: c objective-c c++ c-header cpp-output c++-cpp-output assembler assembler-with-cpp NOTE that this wasn't easy to find from "emacs", since emacs won't search more than the current text chunk with Control-S. In info, it is easier, but info does not work anymore due to not finding curses or something, so "zcat /usr/info/gcc* | less." So I would guess you want to use something like: gcc -o outputfile -x objective-c filename.c -lobjc.a (I have never done any objective-c myself.) > >Graham Hughes <[EMAIL PROTECTED]> MIME OK, PGP preferred > >from stddisclaim import footer >pgp_fingerprint = "E9 B7 5F A0 F8 88 9E 1E 7C 62 D9 88 E1 03 29 5B" > >Graham Hughes <[EMAIL PROTECTED]> MIME OK, PGP preferred > >from stddisclaim import footer >pgp_fingerprint = "E9 B7 5F A0 F8 88 9E 1E 7C 62 D9 88 E1 03 29 5B" > [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- I'm an artist. But it's not what I really want to do. What I really want to do is be a shoe salesman. I know what you're going to say -- "Dreamer! Get your head out of the clouds." All right! But it's what I want to do. Instead I have to go on painting all day long. The world should make a place for shoe salesmen. -- J. Feiffer -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Location of "sc" spreadsheet
Brian White wrote: >Does anybody know if the "sc" spreadsheet is available for Debian? I >cannot find it anywhere. > > Brian > ( [EMAIL PROTECTED] ) > >--- > Friends are relatives that you make for yourself. > > >-- >TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to >[EMAIL PROTECTED] . Trouble? >e-mail to [EMAIL PROTECTED] . Sorry this is so late, I am behind on reading, so ignore it if it isn't helpful anymore, or you already found out. I have sc here, installed from a package "slsc" I believe, which is a slang-ified sc. [EMAIL PROTECTED]/GNU--1.3---Linux--2.0.30--- <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> -- Any sufficiently advanced operating system is indistinguishable from Un*x. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: g++ file doesn't run
Jeff Shilt wrote:
>Thanks for the help - it does compile with g++ instead of gcc, but the
>executable produced isn't d
>oing anything. Here's what i'm doing:
>
>//test.c
>#include
>
>main(){
> cout << "Hello there.";
>}
>
>The test file doesn't print out anything when I run it.
>
>Also, I was wondering if this was a descrepency in versions of gcc. The
>installaition I had befor
>e (using the main parts of the Slackware dist.) compiled both c and c++. Like
>somehow it knew whic
>h way to process it.
>
>
Don't call it "test" and don't end it in ".c". Try .cc or .CC or
even .C because the compiler will look at the extension.
g++ -o filename filename.cc
will name the exec. the same as the input w/o the .cc extension.
[EMAIL PROTECTED]/GNU--1.2---Linux--2.1.25---
<[EMAIL PROTECTED]> <[EMAIL PROTECTED]> <[EMAIL PROTECTED]>
--
A computer lets you make more mistakes faster than any other
invention, with the possible exceptions of handguns and Tequilla.
-- Mitch Ratcliffe
Re: shared library tutorial?
Dale Martin wrote:
>Hello,
>
> I have built a PCCTS source package - PCCTS is the "Purdue
>Compiler-Construction Tool Set" - it produces LL(K) parsers. I'm
>using it in a project which I will eventually Debianize. The PCCTS
>package is close to ready to upload, except it has some libraries in
>it, and I would like to compile them as shared libraries and don't
>know how. (I'd also like to use shared libraries in my own project.)
>
>Can anyone point me to an online reference on how to compile and use
>shared libraries? Note that I'm also interested in the portability of
>the solution - my project also is working with Linux/Alpha, and
>Solaris machines...
>
>Thanks for any info!
>
> Dale
>
I don't know anything about the portability side but this seems
to work for me:
sophis.c:
/* a sophis-tercated example */
extern void grunt(char *);
int main()
{
grunt("Hello world.\n");
}
tercated.c:
/* a sophis-tercated example */
#include
void grunt(char * foo)
{
printf(foo);
}
And then:
$ gcc --shared -o libgrunt.so tercated.c
(as root)
# cp libgrunt.so /usr/local/lib/
# ldconfig -v
$ gcc -o sophis sophis.c -lgrunt
$ ./sophis
Hello world.
$
I suppose there is more to it than this -- depending on your
application -- but I don't know what. The path /usr/local/lib
has to be in the configuration in /etc/ld.conf and should not
be /usr/local/lib if you are actually making a ".deb".
[EMAIL PROTECTED]/GNU--1.2---Linux--2.1.25---
<[EMAIL PROTECTED]> <[EMAIL PROTECTED]> <[EMAIL PROTECTED]>
--
A computer lets you make more mistakes faster than any other
invention, with the possible exceptions of handguns and Tequilla.
-- Mitch Ratcliffe
Re: Fetchmail/procmail
Jason Killen wrote: >Thanks for the pointer on fetchmail. > >The part about having procmail send the mail to an intermediate mail box >is intresting and a little over my head (I don't know much about procmail, >just a sweep of the man pages.) but I will try to work things out. > >Thanks a lot. > >On Wed, 12 Feb 1997, Behan Webster wrote: > (clip) > >Jason KillenQuestion Stupidity >Monolith : driven by inner daemons "My thoughts of despair are getting >[EMAIL PROTECTED] loud." --Social D. > If you are loosing mail, you might want to look in the /var/spool/mail directory. I don't remember what I did wrong, but earlier this year I had a configuration problem and my mail ended up there with the user name of "nobody". I wish I could remember what I did to fix it (it works now). I think I had a problem with the ".procmailrc" and I scrapped it and wrote another one (the old one had gotten a lot of junk in it anyway.) After this everything went fine. [EMAIL PROTECTED]/GNU--1.2---Linux--2.1.25--- <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> -- It is difficult to produce a television documentary that is both incisive and probing when every twelve minutes one is interrupted by twelve dancing rabbits singing about toilet paper. -- Rod Serling -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
RE: mailing list
On Wed, 28 Aug 1996, Ninoles, Fabien: DGSE wrote: > More of that, a news group is more organized. Thread > are follow up and it's easy to jump over some who > don't interested yourself. This will help people to > got an answer more easily and more quickly, and to > free a bit the traffic on this mail list (Debian-users). > (clipped) If you haven't yet, I suggest you look at pine. Pine can be set to sort on "ordered subject" and will behave like a threaded news reader. (I think it isn't "actually threaded" but I don't know the exact difference.) Pine makes newgroups and mailing lists almost equivalent. (some people don't like that design, though.) [EMAIL PROTECTED]/GNU--1.1---Linux--2.0.14--- Develop free apps? Look at http://www.ied.com/~honza/fslu/ - Different all twisty a of in maze are you, passages little.
Re: LinuX Book
On Tue, 27 Aug 1996 [EMAIL PROTECTED] wrote: > Hi guys, > I finally decided to buy a book on LinuX and guess what? There are so > many, I am at a loss for which one to choose (and some are pricey too). Could > anyone out there point me to what is considered the 'defacto' definitive guide > (Advanced!!) to LinuX > > Thanks > > Jonathan Lawson What is LinuX? Is that like Linux? Look in /usr/doc/HOWTO and /usr/doc/HOWTO/mini for the BookList-HOWTO (name may be inexact.) [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.14___ We believe that after death comes The Nothing because when you ask the dead what happens they say Nothing. If death is not the end, if the dead have lied, then it's compulsory heaven for all excepting perhaps Hitler, Stalin and Genghis Kahn. -- Steve Turner, "Creed"
Re: Irqtune: some stats (was Fix for your serial/PPP problems)
On Mon, 26 Aug 1996, Gerry Jensen wrote: > On Mon, 26 Aug 1996, Philippe Troin wrote: > > > Test I was done on a 2.0.14 kernel with a 100kb nul file (dd > > if=/dev/zero...) > > Test II was done on a 2.0.14 kernel with irqtune and a 100kb nul file. > > Test III was done on a 2.0.15 kernel with a 100kb nul file. > > Test IV was done on a 2.0.15 kernel with irqtune and a 100kb nul file. > > Test V was done on a 2.0.15 kernel with a 100kb random file. > > Test VI was done on a 2.0.15 kernel with irqtune and a 100kb random > > file. > > > I IIIII IV V VI > > Count Time Rate Time Rate Time Rate Time Rate Time Rate Time Rate > > -- > > Avg. 59.17 1.7 36.66 2.7 58.53 1.7 34.60 2.9 57.09 1.7 26.93 3.7 > > Dev. 10.32 5.18 9.15 6.31 11.58 0.05 > > > > The figures are quite clear. Irqtune really improves my serial > > performances, even with kernel 2.0.15 (see previous messages, 2.0.15 > > included a patch which might have helped serial performance). > > What I don't understand is why a file consisting of zeros only do not > > get the same transfer rate as a random file. > > If I understand your data, you're saying you get a much faster > transfer rate with the random file over the all-zeros file, right? > This is very bizarre. The random file and a file consisting of all > zeros should *not* get the same transfer rate on any modern modem > which does hardware compression (most do). The all zeros file should > get a much faster transfer rate than the random file. The reason is > that it is not possible to compress a random file so the best transfer > rate possible would be the line speed (28.8 kbs for a 28.8 modem). A > file of all zeros is easily compressible so you should do much better > (theoretically 115200 kbs for a 28.8 modem which does hardware > compression but in practice probably much less). Just at a guess (and I don't know ppp internals that well), it might be PPP escaping the zero as a control character. If the software had to handle the escape sequence for "N thousand" nulls, it might bog it down a bit, and those escape sequences might themselves take up more space, as they often do. Really, though, I can't see that accounting for the difference on a modem that does compression. The compression should be more of a factor than escaping every character even, assuming it IS a compressed link. I don't know if PPP even escapes nulls. (clip) > > Gerry [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.14___ ... although "AI" research had centered on the idea of creating thousands of lines of LISP code which would then somehow become intelligent ... meanwhile in the darkness, the "EMACS" grew and grew, unrecognized even by the AI community, fifty, one hundred times the size of other text editors. In retrospect the general lack of suspicion is [perhaps] the most convincing proof -- "Why they didn't see it coming", New Brittanica 2136
Re: Bugs
On Mon, 26 Aug 1996, Ian Jackson wrote:
> Kevin M. Bealer writes ("Bugs"):
> > I've submitted these two bugs to debian bugs, and got no
> > response; I submitted them again, and still got no response.
> > (Has the list address changed recently?) I could try again, but
> > then I would have to wait again, and so on, so... (In
> > otherwords, sorry for the offtopic.)
>
> Kevin: please send them to [EMAIL PROTECTED], if you haven't
> already.
(done).
> Bruce: this is very bad. Could you change the Pixar alias to point to
> [EMAIL PROTECTED] ? I've made this an alias for
> [EMAIL PROTECTED] We know that Pixar can deliver mail here.
>
> It is definitely very bad if messages are just disappearing and not
> even bouncing. If all else fails the alias at Pixar should be
> removed.
>
> Thanks,
> Ian.
>
>
[EMAIL PROTECTED]/GNU__1.1___Linux__2.0.14___
We believe that each man must find the truth that is right for
him. Reality will adjust accordingly. The universe will
readjust. History will alter.
-- Steve Turner, "Creed"
Re: "man" crashes
On Sat, 24 Aug 1996, Christoph Lameter wrote:
> On Sat, 24 Aug 1996, Dale Scheetz wrote:
>
> dwarf>On Sat, 24 Aug 1996, Christoph Lameter wrote:
> dwarf>
> dwarf>> Seems that the man package cannot cope with some screwups of the dpkg
> dwarf>> updates.
> dwarf>>
> dwarf>Sounds like you upgraded to an experimental version of dpkg. Which
> version
> dwarf>do you have? I'm running 1.3.6 with none of these problems.
> 1.3.9 as of today.
>
> Any way to fix those problems?
>
> {}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}
> {}Snail Mail: FTS Box 466, 135 N.Oakland Ave, Pasadena, CA 91182
> {}
> {}FISH Internet System Administrator at Fuller Theological Seminary
> {}
> {}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}
> PGP Public Key = FB 9B 31 21 04 1E 3A 33 C7 62 2F C0 CD 81 CA B5
I would suggest removing any of the files that it lists as
problems, then remove and reinstall the manpages package, and
possibly the actual "man" utility as well.
[EMAIL PROTECTED]/GNU__1.1___Linux__2.0.14___
"Don't worry about people stealing your ideas. If your ideas
are any good, you'll have to ram them down people's throats."
-- Howard Aiken
Bugs
I've submitted these two bugs to debian bugs, and got no response; I submitted them again, and still got no response. (Has the list address changed recently?) I could try again, but then I would have to wait again, and so on, so... (In otherwords, sorry for the offtopic.) Package: kbd Version: 0.91-3 The kbd package config utility has a script called kbdconfig which can be run to configure keyboard mapping and soft fonts. These settings are used on boot by a script /etc/rc.boot/0kbd from a config file /etc/kbd/config. Some of the fonts have multiple sizes (972.cp for example). When such a font is selected, it is necessary to select which size is desired, for example (setfont /usr/lib/kbd/consolefonts/972.cp -16) would specify size 16. When such a font is selected by kbdconfig, no mechanism is provided to set the font size in the boot script, so the font is not loaded. # setfont /usr/lib/kbd/consolefonts/972.cp This file contains 3 fonts: 8x8, 8x14 and 8x16. Please indicate using an option -8 or -14 or -16 which one you want loaded. # I am using kernel 2.0.0, Debian 1.1 (unstable tree circa Jun 22 1996), and libc5.2.18-8 and libc4.6.27-15. I have made changes to the /usr/sbin/kbdconfig script to poll the user for the font size if it is needed (and to save it in the config file), and to ask the user if they want to enable the configuration when the script completes. I have also changed the /etc/rc.boot/0kbd file to use the font size value. These are the patches ... I'm not familiar enough with diff/patch to give a full-directory patch of the changes, so I am diff'ing them seperately. --> /etc/rc.boot/0kbd --- 0kbd.oldTue Jun 25 23:44:33 1996 +++ 0kbd.newTue Jun 25 23:44:11 1996 @@ -16,6 +16,7 @@ TERM=linux KEYMAP="" SOFTFONT="" + SIZE="" fi echo "Console setup:" @@ -35,5 +36,5 @@ if [ -n "$SOFTFONT" -a "$SOFTFONT" != "NONE" ] then echo -n " " - setfont $SOFTFONT + setfont $SOFTFONT $SIZE fi -> /usr/sbin/kbconfig --- kbdconfig.old Tue Jun 25 23:45:22 1996 +++ kbdconfig.new Wed Jun 26 00:00:50 1996 @@ -88,24 +88,101 @@ fi fi + +while [ x$fontok != xyes ] +do + cat << EOT Please choose one of the following display fonts for the VT devices: EOT -(cd $LIBDIR/consolefonts; ls -x) + (cd $LIBDIR/consolefonts; ls -x) + + echo -e "\nOr answer NONE to skip the loading of a consolefont." + echo -n "What consolefont to load? [$SOFTFONT] "; read softfont + if [ -z "$softfont" ] + then + if [ -n "$SOFTFONT" ] + then + softfont=$SOFTFONT + else + softfont=NONE + fi + fi + +# NOTE: We have no way of detecting whether the font makes the screen +# unreadable or correctly loads; I know of no way to test if the +# font requires a size parameter except to test the program for +# failure, so we must test for existence of font. -- KMB -echo -e "\nOr answer NONE to skip the loading of a consolefont." -echo -n "What consolefont to load? [$SOFTFONT] "; read softfont -if [ -z "$softfont" ] -then - if [ -n "$SOFTFONT" ] + bname=$LIBDIR/consolefonts/$softfont + + # Add a default ending if necessary + + [ ! -e $bname -a -e $bname.cp ] && softfont=$softfont.cp + [ ! -e $bname -a -e $bname.psf ] && softfont=$softfont.psf + + if [ ! -e $LIBDIR/consolefonts/$softfont ] + then + if [ $softfont != NONE ] + then + echo + echo Font $softfont not found, setting to NONE. + softfont=NONE + fi + fi + + if [ $softfont != NONE ] then - softfont=$SOFTFONT + echo ; echo Testing font ... ; echo + sleep 2 + setfont $softfont || SIZE=err + + if [ $SIZE != err ] + then + sleep 5 + setfont &> /dev/null + echo + echo If that font is unacceptable, please + echo rerun this script \($0\). + fi else - softfont=NONE + SIZE=NONE fi -fi + + if [ $SIZE = err ] + then + echo + echo -n "Please enter the font size: " + read SIZE + if [ $SIZE != NONE ] + then + echo $SIZE | grep "-" &> /dev/null || SIZE=-$SIZE + fi + + echo ; echo Testing font ... ; echo + sleep 2 + setfont $softfont $SIZE || SIZE=err + + if [ $SIZE != err ] + then + sleep 5 +
Re: gs 3.53 and up
On Thu, 15 Aug 1996, Ervin D. Walter wrote: > I have had a problem using ghostscript verisons 3.53 up through 4.00 > in the standard distributions. Specifically, when I try to use gs > with *any* device be it the console, X, or a printer, I get errors > like the following with the standard distribs and compiled versions: > > Error: /limitcheck in --moveto-- > Operand stack: >1 0 0 0 > Execution stack: >%interp_exit () --nostringval-- --nostringval-- > --nostringval-- false --nostringval-- --nostringval-- > --nostringval-- --nostringval-- > Dictionary stack: >--dict:645/653-- --dict:0/20-- --dict:42/200-- > --dict:117/250-- > Current allocation mode is local > Current file position is 311980 > > Now, the wierd parts. First, these errors don't occur every time, and > they don't occur in the same place every time. Second, the exact save > .ps file sent through gs 2.whatever or gs >= 3.53 *compiled with > optimizations turned off* works perfectly. This evidence leads me to > believe that there is nothing wrong with my postscript source (not > that I wrote it by hand. I came straight from dvips.). I have solved > the problem personally by downloading gs 4.00 and compiling it with no > gcc optimizations. However, I have to install it in /usr/local/ > because I didn't have a functioning debianized version. This means I > either have to keep a broken gs 3.x or working gs 2.x installed in TeXtpert I'm not, but; [ This is only a guess ] Turning off the optimizations and getting _any_ kind of different end results, except for a speed or size difference should probably be considered a bug in gcc. That having been said, you do know there are two different "gs" interpreters? gs is an Alladin product, donated to GNU (accord. to the Printing HOWTO). As such, the GNU version is just an older Alladin version, presumably the newer Alladin version has more features, and different bugs. So if you have source from one and a binary from the other, you may be hunting bugs that are no longer with us... > P.S. I did try the recommended bug-search algorithm with gs 4.0 to no > avail. ? ... you mean trying to find the "bad patch?" by walking thru the versions? > Thanks, > Ervin Walter > [EMAIL PROTECTED] > > -- > ,,, > (o o) > -oOOo-(_)-oOOo- > Different all twisty a of in - [EMAIL PROTECTED] > maze are you, passages little. - [EMAIL PROTECTED] > > PGP Key fingerprint = A5 AB 25 7D 7A FD 4D FE BE 21 47 60 0C DC 67 9E > > > [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.12___ Programming is always harder than doing the same task manually. It's hard because you must completely understand the problem, take everything into consideration, and protect against every possible flaw. Then you never have to do it manually again. The great driving force behind programming is the fervent devotion to laziness: you wage a war to save typing later.
Re: SCSI and EIDE
On Tue, 13 Aug 1996, Sherwood Botsford wrote: > > This is known as elevator seeking and should be done at the OS level. > > The order that data is written out is very important for data > > reliability, and for this reason, I don't think any hard disk change > > the order of the writes. > > There are several different algolrithms for doing this. Elevator seek > is the simplest of them. Some put a higher priority on reads over > writes. Some re-priortize the queue so that requests that have been > outstanding for a while get bumped up the queue. Linux currently uses (accord. to the kernel lists) something called sawtooth ordering. This means the head moves in the same direction until no more is left to read in that direction, then does the same in the other direction... This provides more "fairness" ie applications with data at the extreme ends of the disk don't stall for a long time. > I agree that you can do this in the OS, but I don't think that it > *should* be done in the OS. > 0.In general smarts should be at the point they are used. We had a > VAX that was about as speedy as a 12 MHz 286 with 287 co-processor. > However, that vax could handle 8 simultaneous 19 KB terminal lines. > How? Smart serial cards. I agree for most hardware, but some issues of hard drive usage, ie reading ahead, really should have support, not only in the kernel but in the application, for best performance. The application knows more about it's usage patterns, which could help in speeding up hardware. (for example, "file" only reads a small chunk of a file, but grep always scans the entire file.) > The CPU should do those tasks that are either too general for > dedicated hardware, or that don't happen often enough to take up much > time. > 1.The OS shouldn't have to care about the layout of the disk. > 2.I don't understand why the order matters for reliability -- > unless you're talking about power failures. This is a low order failure > mode -- which is why you do backups -- for those rare occasions when the > disk does lose a file when the lights go out. The disk is smart enough > to recognize that a new request to write out sector 12345 makes the > previous request to write out sector 12345 invalid. The reasoning is that the filesystem should be in an "inconsistent" state as little as possible. (ideally...) This is for power failures, kernel oops'es, etc. > 3.The kernel of any unixen caches metadata for most of the file > system for long periods of time. This is the whole point of the sync > daemon -- to flush this out to disk. > > > Sherwood Botsford |Unsolicited email that advertises commercial > Physics Dept |activities will consitute a request for > U of Alberta |spellchecking of all words of less than three > Edmonton, AB, |characters. I charge $US500 for this service. > T6G 2J1 |There is no warranty of correctness of this service. I think the OS's job should be efficiency, reliability, etc. Hardware is nice... but doing it in software is cheaper. If Linux can do well on cheap flaky hardware, more power to it. [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.11___ THE LESSER-KNOWN PROGRAMMING LANGUAGES #12: LITHP This otherwise unremarkable language is distinguished by the absence of an "S" in its character set; users must substitute "TH". LITHP is said to be useful in protheththing lithtth.
Re: UncorrectableError from two disk sectors
On Mon, 12 Aug 1996, Douglas Bates wrote:
> `dmesg' shows a lot of messages about two sectors on my disk. For
> example
> hda: read_intr: error=0x01 { AddrMarkNotFound }, LBAsect=1097170,
> sector=48787
> hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
> hda: read_intr: error=0x40 { UncorrectableError }, LBAsect=1097170,
> sector=48787
> end_request: I/O error, dev 03:05, sector 48787
> hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
> hda: read_intr: error=0x40 { UncorrectableError }, LBAsect=1097170,
> sector=48787
> end_request: I/O error, dev 03:05, sector 48787
> hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
> hda: read_intr: error=0x40 { UncorrectableError }, LBAsect=1044268,
> sector=81
> end_request: I/O error, dev 03:03, sector 81
>
> Should this be a concern? What action would be recommended?
Aug 13 01:16:15 brando kernel: hda: write_intr: status=0x51 { DriveReady
SeekComplete Error }
Aug 13 01:16:15 brando kernel: hda: write_intr: error=0x04 { DriveStatusError }
Aug 6 18:50:42 brando kernel: hda: write_intr: status=0x51 { DriveReady
SeekComplete Error }
Aug 6 18:50:43 brando kernel: hda: write_intr: error=0x04 { DriveStatusError }
Aug 6 23:20:24 brando kernel: hda: write_intr: status=0x51 { DriveReady
SeekComplete Error }
Aug 6 23:20:24 brando kernel: hda: write_intr: error=0x04 { DriveStatusError }
This are skimmed from my /var/log/messages ...
My system is a Maxtor 540 (Bios calls it a 528) MB drive, IDE
and a 486-33.
BTW: I posted about these messages on the linux-kernel list a
few days ago, it is being discussed there. This is almost
certainly not Debian specific.
[EMAIL PROTECTED]/GNU__1.1___Linux__2.0.11___
THE LESSER-KNOWN PROGRAMMING LANGUAGES #12: LITHP
This otherwise unremarkable language is distinguished by the
absence of an "S" in its character set; users must substitute
"TH". LITHP is said to be useful in protheththing lithtth.
Re: printing and .profile problems
On Wed, 7 Aug 1996, chris beamis wrote: > > Leszek, > Sorry, I should have mentioned that it is the same .profile which I've > always used successfully with Slackware releases. I was going from memory > which is the only reason I got the quotes wrong. There is something else > going on causing bash or something to ignore my .profile. > > Thanks anyway. > > Chris Beamis Is it named .profile? I have a .profile in /etc/ but in my home directory I use ".bash_profile" and ".bash_rc". [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.11___ "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: what files does dselect/dpkg use to discern choices?
On Tue, 6 Aug 1996, David C Winters wrote: > > Now, to explain the question, since I can't understand the > Subject: line and I wrote it myself... > > I've got approximately 50 machines I need to build. My best > option for the initial system build, unless I've missed something, would > be to run deselect on one machine, then take a deselect-generated file > containing my selections and exporting that file to all the other > machines, so I can just start up dselect and choose "Install" without > having to go through "Select" on each. > How can I accomplish this? Which file(s) do I need to export in > order to make this happen? And, is there a better way to achieve the > same results? > > Now, a meta-issue. I've been extremely pleased with the Debian > distribution, as well as the support I've received from this list. > Thanks, all! > > > David [EMAIL PROTECTED] aka [EMAIL PROTECTED] > Office: 3503 WeH, x86720 Let me suggest (in addition to my last suggestion) the use of "afio" (or cpio) with the utility called netcat. Send the compressed volume through a pipe to netcat, and do the opposite on the other end. nc just acts as a pipe between two systems. Note that afio may do bad things if it overwrites running programs, so look out for that. --> you may want to create a 15 MB or so partition on each machine to run Debian until you get it set up minimally (base system + some networking). When you are done, you can use the bootstrapping partition as an emergency repair system, or just add it on as more swap space. [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.11___ We believe that man is essentially good. It's only his behavior that lets him down. This is the fault of society. Society is the fault of conditions. Conditions are the fault of society. -- Steve Turner, "Creed"
Re: Two Questions
On Tue, 6 Aug 1996, Dan Bergman wrote: > 1. How do I set up a boot so I can boot Debian or Win95? > Now I have to use a boot disk to get debian to boot... > > 2. How do I change the diplay mode in XF86 from 8bit to 16bit > color? I have a Diamond Stealth 64 Video VRAM w/ 2MB and use > 16bit color in win95 so there shouldn't be any probs to get > it in X too.. generally, you type: startx -- -bpp 16 You need XF86Config set up correctly (is it possibly already correct?) You can also edit the Xservers file (type "locate Xservers" it should be in /etc/X11 somewhere) to use 16 bit. Mine has: :0 local /usr/X11R6/bin/X -gamma 1.5 -bpp 16 as the active ingredient. [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.11___ We believe that man is essentially good. It's only his behavior that lets him down. This is the fault of society. Society is the fault of conditions. Conditions are the fault of society. -- Steve Turner, "Creed"
Re: what files does dselect/dpkg use to discern choices?
On Tue, 6 Aug 1996, David C Winters wrote: > > Now, to explain the question, since I can't understand the > Subject: line and I wrote it myself... > > I've got approximately 50 machines I need to build. My best > option for the initial system build, unless I've missed something, would > be to run deselect on one machine, then take a deselect-generated file > containing my selections and exporting that file to all the other > machines, so I can just start up dselect and choose "Install" without > having to go through "Select" on each. > How can I accomplish this? Which file(s) do I need to export in > order to make this happen? And, is there a better way to achieve the > same results? > > Now, a meta-issue. I've been extremely pleased with the Debian > distribution, as well as the support I've received from this list. > Thanks, all! > > > David [EMAIL PROTECTED] aka [EMAIL PROTECTED] > Office: 3503 WeH, x86720 I don't think you can do much in terms of recording input to dselect. I would look at something like a tape-backup technology, or using a null modem to transfer the data. (I assume if you have 50 machines you have more advanced networking than a null modem). If you recompile the kernel, there is an option for "Root filesystem on NFS." If you could get an NFS on one drive to export, you could boot all the machines with NFS and then copy stuff en mass from machine to machine. I'm not sure how to actually send the data, but you could use something like afio to packages the filesystem up and send it through the NFS. A neat idea: look into dbackup... (but it isn't done yet, AFAIK) [EMAIL PROTECTED]/GNU__1.1___Linux__2.0.11___ We believe that man is essentially good. It's only his behavior that lets him down. This is the fault of society. Society is the fault of conditions. Conditions are the fault of society. -- Steve Turner, "Creed"
Re: DEBIAN-Re: logging in takes ages
On Tue, 18 Jun 1996, Paul Wade wrote: > Jim Rush wrote: > > > > My problem, since upgrading is similiar. I get the prompt, enter username > > and password and then wait. It seems to be account related. I can go to > > other virtual terminals and log in as other users, but if I try the same > > user, that > > terminal will also hang. Typicall hang times seem to be as bad as 30 > > seconds. > > > > Note, I almost always log in as root, so my problem may be root specific. > > > > Jim > > > > Let me clarify what I did: > > I downloaded the new kernel from Finland before it was mirrored > in the US. I built it, booted it and got the new delays. I usually log > in via telnet from a Windoze system and then su to root with no delay. > If you have an user account that responds fast, log in with it and then > su root. It will be interesting to see if you get the delay on root that > way. > > Paul Wade - Greenbush Systems > > *** > * http://www.wtop.com/ - THE OTHER WEBSITE forever under construction * > *** This looks exactly like a well-known bug .. possible in syslogd or sysklogd. It no longer happens for me -- perhaps upgrade things some. I keep up with the mirrors so I am usually only a week or so behind the distribution. My current Kernel is 2.0.0 and my sysklogd package is 1.3-6. My libc is 5.2.18-8 and sysvinit is 2.62-1. (hope this helps.) [EMAIL PROTECTED] Error: this signature should not appear. If you see this signature, ...
Re: critical :) nethack question
On Sun, 16 Jun 1996, Dale Scheetz wrote: > On Sun, 16 Jun 1996, Kevin M Bealer wrote: > > > A good place to discuss this (or anything really) is debian-talk... That's > > what it is for... (anything which is not debian only but is debianated ;) > > > So, how do we subscribe to this list Kevin? > > Dwarf > > -- > > aka Dale Scheetz Phone: 1 (904) 877-0257 > Flexible Software Fax: NONE > Black Creek Critters e-mail: [EMAIL PROTECTED] > > If you don't see what you want, just ask -- Send a message containing "subscribe" in the body to: [EMAIL PROTECTED] [EMAIL PROTECTED] Two men looked out through prison bars; One saw mud, the other stars.
Re: critical :) nethack question
On Sat, 15 Jun 1996, Syrus Nemat-Nasser wrote: > On Sat, 15 Jun 1996, Rick Hawkins wrote: > > > Is there a way to open the chests other than kicking them? would this > > stop the potions from shattering? > > This is the wrong forum, but you can use a skeleton key or credit card. > (clip) > > Syrus. > > -- > Syrus Nemat-Nasser <[EMAIL PROTECTED]>UCSD Physics Dept. A good place to discuss this (or anything really) is debian-talk... That's what it is for... (anything which is not debian only but is debianated ;) "If you want to learn about computer programming, the best book to read is Lewis Caroll's "Alice in Wonderland" -- but that's just because it's the best book to read on _any_ subject." -- fortune cookie [EMAIL PROTECTED] Quidquid latine dictum sit, altum viditur. (Whatever is said in Latin sounds profound.)
Re: kernel-package and Lilo
On Sat, 15 Jun 1996, Craig Sanders wrote: (clipped own stuff) > > (I'm thinking of the equiv. of "lilo (options) || echo Warning! blah > > blah") > > > > and/or could LILO be run in "verbose" mode? > > > > I suggest this because it would have saved me a lot of time.. > > > > In other words: (please idiot proof this for my benefit :) ) > > Before I switched to using the kernel package scripts I had my own > 'makelinux' and 'addlilo' scripts which performed a similar function. > > addlilo added an appropriate entry to the end of /etc/lilo.conf and then > ran: > > lilo -t && lilo > > i.e. 'lilo -t' tests the lilo config. if it's OK then lilo is run. > > maybe something like this (but expanded to provide useful warning/error > information) should be done in the debian kernel package script? > > Craig After using "perforate" I wasn't sure so went to copy the kernel image around to unperforate it and was going to run lilo again. But, I accidentally deleted the kernel image. So I ran dpkg -i (the kernel image) and it finished without incident -- Lilo gave an error, but no being verbose, only said "OldLinux not found" and I assumed (<- the problem) that the found images were used, rebooted, then spent several hours trying to get a clean boot into the system so I could run lilo again. lilo -t && lilo wouldn't have helped (me anyway) I think -- if LILO fails it won't install an MBR anyway (as far as I know). It might give a more verbose error message, tho. [EMAIL PROTECTED] "The C Programming Language -- A language which combines the flexibility of assembly language with the power of assembly language."
kernel-package and Lilo
I love the kernel-package thing; but the other day I used it to install a kernel and the vmlinuz.old kernel was not in boot (I deleted it accidentally). The system would not boot -- may I humbly suggest that the install process for the generated package checks if "lilo" returns an error and throws a warning if it does? (I'm thinking of the equiv. of "lilo (options) || echo Warning! blah blah") and/or could LILO be run in "verbose" mode? I suggest this because it would have saved me a lot of time.. In other words: (please idiot proof this for my benefit :) ) Thanks, [EMAIL PROTECTED] "The C Programming Language -- A language which combines the flexibility of assembly language with the power of assembly language."
Re: Re[2]: DEBIAN Linux on floppy disks
On Wed, 12 Jun 1996 [EMAIL PROTECTED] wrote: > use a cable to transfer the files to the laptop from the machine with > the cdrom. Dont forget to use a proper cable setup... > > It'll save your ass a ton-o-time. and you wont have to be splitting > files across floppies. > > I use xtgold's "xtlink" which works great... > > BUT - if you dont have xtlink... > > MS-DOG provides the software called "interlnk", which does the same > thing - more or less. (clip) The linux kernel has PLIP (parallel line ip) which can do this once you get a base system installed... look in /usr/src/linux/Documentation or in the /usr/doc/HOWTO/PLIP-HOWTO (names may be garbled)... you need a laplink cable and possibly a version 2 kernel on both ends. This will provide an IP connection I believe ... you could then NFS mount or FTP the stuff across. If you NFS mount you can just install from those directories... *warning* but I haven't tried any of these things. [EMAIL PROTECTED] "The C Programming Language -- A language which combines the flexibility of assembly language with the power of assembly language."
Re: DEBIAN Linux on floppy disks
On Thu, 13 Jun 1996, Rick Hawkins wrote: > > > Why are there always utilities to do this? > > > > What would happen if I wrote a simple program to output the file in volumes > > of a certain size, and then used "cat" to stick em end to end? > > that works. I had to do this a few times. I actually have a similar problem myself... I would like to back up by debian system with my non-linux-compatible tape drive. My theory is that I can dump it in say 20 MB volumes to a file on a DOS partition, and use dos to throw it onto the tape. The tape can hold about 400 MB --> the size limit will come from how much space I can free in my Win95 partition which is only 100 megs and is nearly full. I have read the manpage for afia -- it supports volumes. Is there any way I can tell it to output only "volume 3" once and next time only "volume 4"? I am thinking I should have it dump to a symlink and make the symlink point to /dev/null for all but the volume I want each time... is this feasible? Is there a better way? (Of course the _right_ way would be to write a driver for my tape drive.) (I expect to mount my system readonly to do this, but I could mount a ramdisk to hold the symlink... still a pita probably.) [EMAIL PROTECTED] "The C Programming Language -- A language which combines the flexibility of assembly language with the power of assembly language."
Re: DEBIAN Linux on floppy disks
On Wed, 12 Jun 1996, Steve Preston wrote: (clip) > > You need to 'split' the .deb files that are too big. This requires > dpkg-split, but I am not aware of a WIN95 version of this. If you > created a (relatively minimal) linux partition on your cdrom machine, > then you could split the .deb files and copy them to floppies. > > -- > Steve Preston ([EMAIL PROTECTED]) Why are there always utilities to do this? What would happen if I wrote a simple program to output the file in volumes of a certain size, and then used "cat" to stick em end to end? [EMAIL PROTECTED] You have new mail in /dev/null
Re: Debian 1.1beta problems
On 12 Jun 1996, Rob Browning wrote: > David Gaudine <[EMAIL PROTECTED]> writes: > > > This does work. Since I don't know how to find the configuration > > programs for some package, I use dselect to remove the package and > > then to reinstall it. I assume there's a better way, but this does > > work. > > There's nothing too terribly wrong with that, but often a package will > tell you about it's config programs during the install process. > Failing that, or if you forget what it told you, you can also either > try reading the docs in /usr/doc/packagename, or more directly, try > "dpkg --listfiles packagename". > > -- > Rob If you have the package on your drive, you can just dpkg -i (package). Dselect uses options to prevent it reinstalling packages that are already in... A "reinstall" options along with install;remove;purge would be nice... (but it's not real critical). dpkg -i (already there package) is good when you are dealing with "essential" packages and so on. (the manual attention is good for base packages anyway :) ) --- BTW I'd like to say I really like the new "[S]elect" screen in dselect... Especially the new/broken packages at the top. === [EMAIL PROTECTED] You have new mail in /dev/null
Re: loss of routing info with 2.0.0
On Tue, 11 Jun 1996, Rick Hawkins wrote: > > I have compiled the 2.0.0 kernel (successfully), but lost my routing > info in the process (coming from 1.3.95). "route" only shows the > machine itself. > > If i manually add the missing router & gateway, it works fine, but i > lose this on reboot. > > help! :) > > rick I had to add a default gateway to one of the files in /etc/init.d also: route add default gw (my gateway's ip) NOTE: I don't know if this has been fixed or where the problem is .. I'm not sure how up-to-date my system is. Also this may not be the correct fix. This happened around the same Kernel versions you mention. However at the same time I also had to reinstall the entirety of Debian due to disk corruption... (fsck actually hanged the system) so I thought it was just me. [EMAIL PROTECTED] Pascal, n.: A programming language named after a man who would turn over in his grave if he knew about it.
Re: Quotas problems
On Sun, 26 May 1996, Michael Meskes wrote: > Kevin M Bealer writes: > > Just for reference, I had to change the fstab to have > > "usrquota,grpquota,quota" in it for the fs. After that, edquota gave a nice > > All three? It should suffice to use 'usrquota' if you want user quotas and > 'grpquota' for group quotas. Just 'quota' shouldn't be needed. > > Michael > > -- > Michael Meskes |_ __ > | / ___// / // / / __ \___ __ > [EMAIL PROTECTED] | \__ \/ /_ / // /_/ /_/ / _ \/ ___/ ___/ > | ___/ / __/ /__ __/\__, / __/ / (__ ) > Use Debian Linux! | //_/ /_/ //\___/_/ // > I started with just "quota" and it didn't work. Using all three it worked, I haven't tried any other combinations. [EMAIL PROTECTED] "You are in a maze of twisty little passages, all alike."
Re: Quotas problems
On 22 May 1996, Hugo HAAS wrote: > Hi. > > I've tried to install quotas but I've got a little problem. > When I make an 'edquota someone', I get : > > Quotas for user someone: > > and I don't know what to write after this. > > I've found no informations in the man, or in the docs. So, if someone has > managed to do that or possesses a good documentation, could he send me all > the useful informations. > > Thanks. > > +-+ > | Hugo HAAS |E-Mail : [EMAIL PROTECTED] > | > | IRC : Bartman| WWW : http://www.cti.ecp.fr/~haash8/ > | > +-+ Just for reference, I had to change the fstab to have "usrquota,grpquota,quota" in it for the fs. After that, edquota gave a nice block like this: Quotas for user someone: /dev/hda2: blocks in use: 78554, limits (soft = 0, hard = 0) inodes in use: 3657, limits (soft = 0, hard = 0) When I edit the numbers, and quit, it works. [EMAIL PROTECTED] Error: this signature should not appear. If you see this signature, ...
Re: kernel headers
On 20 May 1996, Manoj Srivastava wrote: (clip) > The kernel-source package is a superset of the kernel-headers > package, so the headers have not been "separated" from the rest of > the source. (clip) > manoj > -- > Everyone has a purpose in life. Perhaps yours is watching television. > -- David Letterman %% > Manoj Srivastava Systems Research Programmer, Project Pilgrim, > Phone: (413) 545-3918A143B Lederle Graduate Research Center, > Fax: (413) 545-1249 University of Massachusetts, Amherst, MA 01003 > <[EMAIL PROTECTED]> http://www.pilgrim.umass.edu/%7Esrivasta/> I have heard discussion on this list before about why debian does this; I don't want to argue why;; But will it break anything major if I don't follow this guideline, and esp. is there a temporary way to set things up 'the old way'? Most of what I compile right now wants kernel headers so it can be compatible with the current kernel (ie kernel utilities and patches.) For example I have kernel utilities which use #include and I keep catching them raiding the /usr/include directory. [EMAIL PROTECTED] "The C Programming Language -- A language which combines the flexibility of assembly language with the power of assembly language."
Re: dselect complaints
On Sun, 19 May 1996, William S. Gribble wrote: > Ian Jackson wrote: > > I'm not interested in hearing any more complaints or even extensive > > suggestions for improvement, unless the person complaining is > > volunteering to do the work on a new interface. > > If you don't want feedback about the tools, might I suggest that you (junk clipped) As someone who has suggested a rewriting of dselect, let me vote for Ian on this one. Dselect is hard for new users, but I like it overall... It is extremely functional. I was going to come down on Mr. Gribble for his post, but that work has already been done, so I'm as of now downloading the source to dselect and I'm going to head-butt it until I understand it or it defeats me. (inside tip: bet heavily on dselect ;) [EMAIL PROTECTED] "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: Root login is waiting
On Tue, 21 May 1996, Fundamental wrote: > On Mon, 20 May 1996, Erick Branderhorst wrote: > > > I have a problem with root login again. All other logins are fine, > > but root login is waiting after I typed the password. Even > > su isn't working anymore. I updtated some packages, including > > a few from Incoming this morning, after reboot, this behaviour > > is present. Sound familiar, a solution? > > > this happened to me eactly (accept on a solaris machine - but i have a > debian machine also) > > setting quotas incorrectly is what stuffed me up .. so i booted the > machine up off the installation boot disk and hacked out the quotas .. (clip) There was a discussion before on linux-kernel about this---it seems like the xconsole (and maybe other) pipes fill up and since root logins are tracked by the system logs, the login hangs due to the 'filled pipes' (call a plumber?) If quota is set to verbose, it displays that [-] [\] [|] [/] rotating bar... perhaps this clogs it faster... I don't think this is generally curable :( perhaps fiddling with the code that logs root logins to the logs might help... or course a security hole could be created if one was not careful. You can try to make your login less noisy... > > \\ > > ONE SENSE Of BEAUTY > On white plum-petals that were pure and sweet, > The Nightingale now wipes its muddy feet. > \\ > The Australian Internet Company ISP Par Exceliance [EMAIL PROTECTED] "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: random post
On Thu, 16 May 1996, Guy Maor wrote: > On Wed, 15 May 1996, Kevin M Bealer wrote: > > > Is there a reason debian does not ship with /dev/random and /dev/urandom > > mknodded? (mknod /dev/random c 1 8 ; mknod /dev/urandom c 1 9 and they work > > fine however) > > No. Report it as a bug against base. Or better yet, write an updated > MAKEDEV script from devices.tex. > > > Guy OK. here's the diff to the MAKEDEV script: --- MAKEDEV.old Fri May 17 00:29:18 1996 +++ MAKEDEV.new Fri May 17 00:31:26 1996 @@ -226,6 +226,8 @@ makedev zero c 1 5 $public symlink core $procfs/kcore makedev full c 1 7 $public + makedev random c 1 8 $public + makedev urandom c 1 9 $public makedev ram b 1 1 $disk makedev tty c 5 0 $tty ;; I believe this is the only change necessary, as /dev/zero is mentioned nowhere else in the script. [EMAIL PROTECTED] error
Re: random post
On Fri, 17 May 1996, Guy Maor wrote: > On Fri, 17 May 1996, Kevin M Bealer wrote: > > > > No. Report it as a bug against base. Or better yet, write an updated > > > MAKEDEV script from devices.tex. > > > here's the diff to the MAKEDEV script: > > Hehe, that's not what I meant. Take a look at devices.tex. There's > LOTS and LOTS of stuff that's not in MAKEDEV, and consequently, not in > the base package. And there's lots of links and stuff that MAKEDEV > creates that shouldn't be created. I meant for you to write a brand > new MAKEDEV based off a recent devices.tex. > > That's a bit more than a 2 line patch. > > > Guy Yes that would make more sense, but I'll have to leave it to someone more competent for now... I've only understood what mknod is used for about a week now. :) [EMAIL PROTECTED]
Re: X Windows
On Thu, 16 May 1996, Dale Scheetz wrote: > Well, I found the problem. I was running procps-0.99-1, which clearly has > some problems. When I upgraded to procps-0.99-3, 'ps a' produces a much > fuller list, including the entry for xdm. Kill on the proper pid does the > trick. I'm still not sure why the start/stop script didn't work. Maybe it > will work now that I have a good ps. > > Thanks, > > Dwarf > > -- > > aka Dale Scheetz Phone: 1 (904) 877-0257 > Flexible Software Fax: NONE > Black Creek Critters e-mail: [EMAIL PROTECTED] > > If you don't see what you want, just ask -- Just for reference, you can kill by name with -killall-. "killall xdm" will do in a pinch. [EMAIL PROTECTED] error
Re: X Windows
On Thu, 16 May 1996, Craig Sanders wrote: (clip) > Is it possible to get xdm to put a login prompt on more than one vc? > > e.g. have xdm login screens on vc 3,4,5,6 > > so i can login on vc3 but let other people share my machine in X windows > without having to log out...just press ctrl-alt-F4 to switch to another X > login screen. > > Anyone done this before? if xdm can't do it, is there another tool that > can? > > Craig You can start more than one X server, (add an entry to etc/X11/xdm/Xservers /using :0.1) but I'm not sure how well logins will work. I tried once to have two X-servers, but it was memory-intensive and messy to use so I didn't bother ironing out the details. Usually the reason to do this would be to have a 'true color' and a '8 bit color' screen at the same time. I think you have to tweak your environment just right, or else give parameters to every X program telling it which server to use. [EMAIL PROTECTED] THE LESSER-KNOWN PROGRAMMING LANGUAGES #12: LITHP This otherwise unremarkable language is distinguished by the absence of an "S" in its character set; users must substitute "TH". LITHP is said to be useful in protheththing lithtth.
Re: Unidentified subject!
On Wed, 15 May 1996, Gerry Jensen wrote: > Does anyone know of an easy way to make it so that > cron.daily/weekly/monthly jobs get executed even if the system is not on > at the specified time? (clip) > > For instance, I typically only run my home system an hour or two per > day. How could I make it so that the first time I boot my system each > day, /etc/cron.daily jobs gets run, the first time I boot each week that > /etc/cron.weekly gets run, and the first time I boot at the beginning of > a new month that /etc/cron.monthly gets run? As it is now, these almost > never get run unless I do it manually. > > Thanks, > > Gerry > [EMAIL PROTECTED] Try running a script like this out of your crontab (say once an hour). It should run things every day. Even better, " date +%s > /var/local/today " could be put in a script in /etc/cron.daily and this would then run only if the regular mechanisms didn't. Week and month scripts are left as an exercise :) #!/bin/bash if [ ! -f /var/local/today ] then # this will cause it to run if the time-file disappears. date -d "2 day ago" +%s > /var/local/today exit 0 fi if [ `date -d "1 day ago" +%s` -gt `cat /var/local/today` ] then date +%s > /var/local/today run-parts /etc/cron.daily fi [EMAIL PROTECTED] So you think you know the real meaning of fear? Yeah, you think you do know, but I doubt it. When you sit in a shelter with bombs falling all over. And the houses around you are burning like torches. I agree that you experience horror and fright For such moments are dreadful, for as long as they last, But the all-clear sounds--then it's okay-- | -- Ilya Selvinskiy You take a deep breath, the stress has passed by. | (Taken from "The Sum But real fear is a stone deep down in your chest. | of All Fears" by You hear me? A stone. That's what it is, no more. | Tom Clancy pg. 182.)
random post
Is there a reason debian does not ship with /dev/random and /dev/urandom mknodded? (mknod /dev/random c 1 8 ; mknod /dev/urandom c 1 9 and they work fine however) [EMAIL PROTECTED] The greatest dangers to liberty lurk in insidious encroachment by men of zeal, well-meaning but without understanding. -- Justice Louis D. Brandeis
Re: 1.1 setup for /dev/xconsole?
On Tue, 14 May 1996, Maarten Boekhold wrote: > > >If you try bringing up xconsole and see nothing, then it's possible that > > >you've already been logging messages for a while and the pipe filled up, so > > > > Has anyone noticed that with recent kernels, the pipe to xconsole fills > > up on boot up before boot up is finished, which causes all kinds of > > problems. This started around 1.3.65 and still is a problem, so I > > stopeed using xconsole. > > I'm using 1.3.71, no probs with xconsole, system is a (to 1.1) upgraded > 0.93R6. > > Maarten > > ___ > | Maarten Boekhold, Faculty of Electrical Engineering TU Delft, NL | > | [EMAIL PROTECTED] [EMAIL PROTECTED] | > | Take life as a party! | > --- > > > Yes I noticed that but could never figure out what to do. So now I just (as root) /etc/init.d/syslogd stop and then /etc/init.d/syslogd start. (Thanks folks!) [EMAIL PROTECTED] "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: X Windows
On Mon, 13 May 1996, Dale Scheetz wrote: > On Mon, 13 May 1996, Susan G. Kleinmann wrote: > > > One of the joys of getting X up is that you'll get Netscape > > up and then you'll be able to use DejaNews to answer every question under > > the Sun! > > Switching to a Cyrrus Logic card solved the major problem. Now I only have > a hand full of little ones :-) > First, when I try and install the netscape package, I get: > > (Reading database ... 13477 files and directories currently installed.) > Preparing to replace netscape (using netscape-2.01-2.deb) ... > Unpacking replacement netscape ... > Setting up netscape ... > ERROR: The Netscape archive must be in /tmp under the name >/tmp/netscape-v201-export.i486-unknown-linux.tar.[Z|gz] > dpkg: error processing netscape (--install): > subprocess post-installation script returned error exit status 1 > Errors were encountered while processing: > netscape You get it from netscape... go to ftpX.netscape.com where X is a number... ie try ftp2.netscape.com. > I realize that I need to get this archive before I can do the > installation. Where do I find this beast? How come the installation > doesn't point me to it? I looked in contrib, non-free, etc... and didn't > find anything. > > > I think in addition to the packages you mentioned, you'll want to get > > xpm4.7 (for fvwm2 -- I think you want fvwm2 rather than fvwm). I also > > find myself always getting xfntpex. > > Fvwm2 installs fine once you install xpm4.7. However, since fvwm2 is not > in the /etc/window-managers file. Shouldn't this file contain all of the > window managers that are available under Debian X? > > > > > Then there's the issue of really doing something with X. What you need > > then > > depends on what you want to do. > > > When I decide what I want to add, how do I hang it on the menu system? Is > there a good user tutorial? In particular, I can start the xserver either > as root or as a user, but I can't seem to do both. Once root cranks up the > xserver, how does a user gain access to the server? copy /etc/X11/fvwm2/system.fvwmrc or /usr/doc/examples/fvwm2/system.fvwmrc to your home dir as .fvwm2rc And then edit it into the ground :) For a real-easy-to-use menu system try the 9menu thing. You actually can type: 9menu hello sync xterm to get a menu with these three commands. > TIA, > > Dwarf > > -- > > aka Dale Scheetz Phone: 1 (904) 877-0257 > Flexible Software Fax: NONE > Black Creek Critters e-mail: [EMAIL PROTECTED] > > If you don't see what you want, just ask -- > > > [EMAIL PROTECTED] Error: this signature should not appear. If you see this signature, ...
Re: X Windows
On Tue, 14 May 1996, Carlo U. Segre wrote: > Unfortunately, you mus edit the /etc/X11/fvwm/system.fvwmrc file yourself > to add the menu items. I don't know of any good menu editor but there > may be one for all I know. > > Cheers > > Carlo > > *** > *Carlo U. Segre * > * Department of Biological, Chemical and Physical Sciences * > *Illinois Institute of Technology, Chicago, IL 60616 * > * Voice: (312) 567-3498 FAX: (312) 567-3494* > * [EMAIL PROTECTED] * > *** It is recommended you copy system.fvwmrc to ~/.fvwmrc to have a private copy and edit that. [EMAIL PROTECTED] "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: X Windows
On Tue, 14 May 1996, Dale Scheetz wrote: > > Unfortunately, you mus edit the /etc/X11/fvwm/system.fvwmrc file yourself > > to add the menu items. I don't know of any good menu editor but there > > may be one for all I know. > > > Vi works for me :-) 'pico' is good for a unix newbie. If you are used to the ancestral wordstar command set that many dos editors use (even windoze borrows from it ^V=paste etc.), "joe" (which is my editor of choice) is good. joe is one executable with 5 names: joe, jstar, jmacs, rjoe, jpico. Joe is the author's choice, the others are partial/complete emulations. > > Thanks, > > Dwarf > > -- > > aka Dale Scheetz Phone: 1 (904) 877-0257 > Flexible Software Fax: NONE > Black Creek Critters e-mail: [EMAIL PROTECTED] > > If you don't see what you want, just ask -- [EMAIL PROTECTED] "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: Dselect proposed interface (was Re: 1.1 installation notes.)
On Tue, 14 May 1996, Christian Hudon wrote: > > Looks OK (we can include a mouse interface). But what is required more then > > a new look is the ability to list only the 'installed' packages, or select a > > singe package to be installed/removed, and a method of seeing which package > > (installed) is older than its archive version (and will therefore be > > updated). > > About the interface, it'd be nice to have something mc-ish. (mc = Midnight > Commander; get it and take a look if you don't have it, there's a Debian > package for it). It's got pull-down menus and dialog boxes... it's very > nice, supports gpm mice and is something DOS people will recognize > instantly. > > And since mc is GPL'ed, we could borrow their low-level > interface code. Would save quite a bit of coding work and would make for a > nice interface, IMHO at least. > > Christian > There's libraries for this already, for example ncurses. OTOH I understand they are quite large (and would have to be linked static of course). Which does not bother me personally. [EMAIL PROTECTED] Parents: this version of Linux contains small components that new users may pick up and put in their mouth.
Re: problems with mailagent
On 13 May 1996 [EMAIL PROTECTED] wrote: > Ahh, I see. [I am copying this to debian user, since the answer maybe > useful to others as well.] > > People using mailagent with popclient may want to run > mailagent by itself after downloading mail from the server9since > popclient delivers directly yo the spool file, not honoring .forward > files), like so: > % mailagent -f /var/spool/mail/ > > Fron the manpage: > -f mailfileUsing mailfile as a UNIX-style mailbox (i.e. one where each > mail is preceded by a special From line stating the sender > and the date the message was issued), extract all its mes- > sages into the queue and process them as if they were > freshly arrived from the mail delivery subsystem. > > manoj > > -- > "When, however, the lay public rallies round an idea that is denounced > by distinguished but elderly scientists and supports that idea with > great fervor and emotion--the distinguished but elderly scientists are > then, after all, probably right." --Isaac Asimov %% > Manoj Srivastava Systems Research Programmer, Project Pilgrim, > Phone: (413) 545-3918A143B Lederle Graduate Research Center, > Fax: (413) 545-1249 University of Massachusetts, Amherst, MA 01003 > <[EMAIL PROTECTED]> http://www.pilgrim.umass.edu/%7Esrivasta/> However, it does honor /etc/aliases. [EMAIL PROTECTED] "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
Re: X Windows
On Mon, 13 May 1996, Dale Scheetz wrote: > When I decide what I want to add, how do I hang it on the menu system? Is > there a good user tutorial? In particular, I can start the xserver either > as root or as a user, but I can't seem to do both. Once root cranks up the > xserver, how does a user gain access to the server? xdm gives a login prompt. > > TIA, > > Dwarf > > -- > > aka Dale Scheetz Phone: 1 (904) 877-0257 > Flexible Software Fax: NONE > Black Creek Critters e-mail: [EMAIL PROTECTED] > > If you don't see what you want, just ask -- [EMAIL PROTECTED] Error: this signature should not appear. If you see this signature, ...
Re: X Windows
On Sun, 12 May 1996, Dale Scheetz wrote: > I found some time this weekend to finaly set up X windows and was able to > learn enough to put it off again. > The graphics card in my machine is a Trident TGUI 9440 AGI which the docs > say is not yet supported (as of June of 95) but that someone was working > on it. Has there been any news? > I have a Cyrus Logic card in the shop I can swap. I suspect that will > happen before the new driver gets to me :-) > I am somewhat put off by the hodge podge of packages in the x11 section. I > know I need: > xserver_x I chose xserver_svga > xlib > xbase > xfntbase > xmanpages > fvwm2 > Is there anything else I need for a minimal X installation? How do I tell > (asside from examining dependencies and description fields) which of the > other packages in this section are desirable, necessary, objectionable, > etc... > I'd love some pointers on what's left as well. > > TIA, > > Dwarf > > -- > > aka Dale Scheetz Phone: 1 (904) 877-0257 > Flexible Software Fax: NONE > Black Creek Critters e-mail: [EMAIL PROTECTED] > > If you don't see what you want, just ask -- I don't think you technically need fvwm2, (but it's nice.) Also rxvt is supposed better on mem than xterm, and xcontrib probably has neat stuff but I've never uninstalled it to see what disappeared. Most people will use xkeycaps eventually since every person who installs X seems to have to fiddle with bkspc/delete and fix them. You will want some other stuff if you intend to _compile_ anything which is X'ed. As for your graphics drivers, I would strongly recommend taking that to an X11 newsgroup if you have access to one... not that this list is overtrafficed, just you will get more responses; some of the hardware X11 lists are 100 messages or so / day last I checked :) BTW, what does TIA mean? [EMAIL PROTECTED] Hey guys, I just heard something that really worries me... (wringing hands) What is the correct side of the bed to get up on?
Dselect proposed interface (was Re: 1.1 installation notes.)
On Fri, 10 May 1996 [EMAIL PROTECTED] wrote: > > From: Steve Preston <[EMAIL PROTECTED]> > > Subject: 1.1 installation notes. > > > > ... > > > > One trouble is that I find the dselect "Select" screen confusing. > > Admittedly, It is not immediately obvious to me which line is the > > "selection". Also I was surprised, the first time I used dselect, to > > see each section (admin, doc, text, etc.) show up multiple times. > > > > I would prefer not starting in a split-screen mode here. > > > > Another trouble was when dselect throws you into its > > conflict/dependency sublists. I was afraid to hit 'Enter' (to get > > back to the main package list), for fear it would end the whole > > package selection process. Of course, I didn't know that I could go > > back to the "Select" screen even if I did accidentally leave it. Thinking about the dselect interface... What we need is a simpler job, yes power is nice, by the word here is interface, and we want an interface that any 'dos' user can understand. This is just a concept mockup -- it should be larger, not be made of ascii, etc. +-+ | GNU Emacs -- v1.0.3-2 |1| Package: X.XX MB Installed: X.XX |3| | | The extensible self documenting |#| Installed but not configured. |#| | text editor | | Requires: (name1), (name2)| | | | | Conflicts: (name3)| | | | | Suggests: (none) | | | | | Provides: (editor)| | | | | | | | | | | | | | | | | |_|2|___|4| |5< admin base comm devel doc [editor] electronics ham mail ... >6| |-+ +--+-Installed-+-| | Important Editor: | |7| | ed -- The classic Unix line editor | Yes | | | nvi -- 4.4BSD Reimplementation of vi | No | | | Standard Editor: | |#| +--> Emacs -- The GNU Emacs Editor.| Partial | | | Optional Editor: | | | | elisp-manual -- Emacs Lisp Reference Manual | | | | elv-vi -- elvis, vi, view, input - The editor | |8| +---+-+ |9< (+)Add package (H) General Help (D) Done (save results) >0| |9< (-)Remove package (C) Command Help (Q) Quit (without saving)>0| +-+ All of the '012..9' keystrokes would move the scrollboxes as if you were pressing an arrow. All of the keystroke commands could be listed in a box at the bottom with a scrollbar, the sections likewise in the middle. (shift could scroll bigger stuff a page at a time.) Just bang on the buttons and you scroll it... We can have all the letter commands we want and document as much as we want. I know it looks comical, but anyone who saw this would be able to grasp it -instantly-, leaving them to worry about more important things. The dpkg --info text blurb should is in the upper left hand block, and all the size/depends/status stuff can be in the upper right, including the (hold/installed/old/selection) status which is also given a one-word summary next to the package-name scrollbox in the center. Sections listed in the center, of course change the order with 'o' or whatever, but that can be dealt with later. The current dselect screen isn't bad -- it's efficient, etc. But it is too 'unix' ... which is to say, you're expected to think. At this stage the first time user has ~ 400 packages to deal with. All the power in the world can be hidden _just_below_ the surface, but the steering wheel and brakes have to be easy to find. Also someone suggested there be an installed size parameter in the packages file... this probably wouldn't be a bad idea. [EMAIL PROTECTED] So you think you know the real meaning of fear? Yeah, you think you do know, but I doubt it. When you sit in a shelter with bombs falling all over. And the houses around you are burning like torches. I agree that you experience horror and fright For such moments are dreadful, for as long as they last, But the all-clear sounds--then it's okay-- | -- Ilya Selvinskiy You take a deep breath, the stress has passed by. | (Taken from "The Sum But real fear is a stone deep down in your chest. | of All Fears" by You hear me? A stone. That's what it is, no more
Re: 'unsupported packages' (was Re: uugetty?)
On Tue, 7 May 1996, Steffen Mueller wrote: > On Tue, 7 May 1996, Craig Sanders wrote: > > Hi Craig, > > > As an added bonus, you also get the satisfying feeling of putting > > something worthwhile back into debian! > > 8-) > > At least some preservation of config files should be done additionally > to the simple control file setup. > > > Maybe there should be an 'unsupported' directory for packages which > > have been released but don't have an active maintainer...or is this > > what 'contrib' is for? Also, allow any package in 'unsupported' can be > > adopted by anyone who takes the time to actively maintain it. > > That's okay so far but it doesn't guarantee bug maintainance. Assume all > those folks going to install some of these packages and run into > trouble > > No problem for the experienced user but quite annoying to all the > members of the mailing lists dealing with supported packages. > > > encourage this? Does the benefit of having lots of extra packages > > outweigh the disadvantage of having bug reports about unsupported > > packages? > > That's what really could be a problem. > > Generally packaging "unsupportet" files would be okay as long as those > packages are marked UNSUPPORTET or orpahned. If someone adopts it it can > make its way into one of the common directories with a small comment > "adopted 96 by John Doe" + "Maintainer : " > > Any suggestions ? > > Greetings, > > Steffen > - > Steffen R.Mueller __ ___ _ _ email: [EMAIL PROTECTED] > NTG Netzwerk und Telematic GmbH \ \/ / (_)_ __ | | __ fax : +49 721 9652 210 > Geschaeftsbereich Xlink \ /| | | '_ \| |/ / phone: +49 721 9652 211 > Vincenz-Priessnitz-Str. 3/ \| | | | | | < RIPE : SM25-RIPE > D-76131 Karlsruhe, Germany /_/\_\_|_|_| |_|_|\_\ WWW.Xlink.net/~steffen How about a debian-hacker@ or debian-guru@ mailing list where programmers debian can all take a crack at problems as they arise for unsupported packages? I think a lot of people would contribute who otherwise don't have the flexibility/confidence to fully maintain packages... and people on the list would probably, in time, start 'graduating' into maintaining packages. [EMAIL PROTECTED] So you think you know the real meaning of fear? Yeah, you think you do know, but I doubt it. When you sit in a shelter with bombs falling all over. And the houses around you are burning like torches. I agree that you experience horror and fright For such moments are dreadful, for as long as they last, But the all-clear sounds--then it's okay-- | -- Ilya Selvinskiy You take a deep breath, the stress has passed by. | (Taken from "The Sum But real fear is a stone deep down in your chest. | of All Fears" by You hear me? A stone. That's what it is, no more. | Tom Clancy pg. 182.)
Re: smail ignores visible_name and from_field config settings
On Mon, 6 May 1996 [EMAIL PROTECTED] wrote:
> I have a home machine on which I would like to send mail to local users as
> well as the Internet (when a PPP link is up). I used to do this by telling
> smailconfig that I am an "Internet site" with visible name "primenet.com"
> (my Internet Service Provider's domainname).
I use the visible name now, I have tried in the past to use 'psu.edu' as my
domain (or some similar setting) and brando as the 'hostname'... Which was a
mistake My ISP (Penn State Univ. PA, USA) started trying to route
messages through my machine. Only a few, when the user misconfigured
their mailer... The messages got stuck in a loop on my machine, and built up
header info for a while till they died in transport.
> This produced from fields like @primenet.com, which is great for
> me ([EMAIL PROTECTED] is my account with my ISP) but not so hot for my
> wife ([EMAIL PROTECTED] is not an actual account) or for root
> ([EMAIL PROTECTED] is not the right account). I fixed this by adding
>
> from_field="From: Matt Birkholz <[EMAIL PROTECTED]>"
>
> to the end of /etc/smail/config.
>
> After upgrading to Debian 1.1beta (smail 3.1.29.1-22), this arrangement
> does not work. The from_field setting seems to be ignored. All of my mail
> is "From: <[EMAIL PROTECTED]>" ("dagobah" being my local hostname).
>
> Am I the only person who was / is trying to do this? Has anyone succeeded?
>
I think I'm set up this way, but I did not get the setup from anywhere, it
does seem to work though.
> Matt Birkholz <[EMAIL PROTECTED]>
> Finger [EMAIL PROTECTED] for PGP 2.6.2 Public Key ID = 74305425
> Key Fingerprint = B3 34 FB 3E 3C FE E8 57 AA B4 B2 95 A7 C0 1E AF
My latest attempt to 'stabilize' stuff: I use a script to check my mail, an
I just put code in to not download mail while 'pine' is up... Otherwise
everyone (might) have their hands in the mailbag at once. I don't know if
this is necessary.
[EMAIL PROTECTED]
Hey guys, I just heard something that really worries me...
(wringing hands) What is the correct side of the bed to get up on?
Re: base system: /dev/psmouse not supported by kernel?
On Mon, 6 May 1996 [EMAIL PROTECTED] wrote: > Package: base > Date: 3 May 1996 > > X11 cannot find my PS/2 mouse. I assume that it should be /dev/psmouse > but it appears to be misdefined. It is defined as character type 10, > but 'cat /proc/devices' does not show a type 10 character device! > > Was this not compiled into the kernel (1.3.95) I picked up from > the installation disks? I am now using my own 1.3.88 kernel > > Derek Lee I believe I am using /dev/ps2mouse. [EMAIL PROTECTED] Hey guys, I just heard something that really worries me... (wringing hands) What is the correct side of the bed to get up on?
Re: status of dftp and dpkg-ftp
On Tue, 30 Apr 1996, Bruce Perens wrote: > From: Scott Barker <[EMAIL PROTECTED]> > > What's the current status of dftp and dpkg-ftp? Are either of them > > "official" > > I would like dpkg-ftp to become part of the "base system", which is the > stuff that is a required part of Debian. It's necessary so that > mostly-net installs would work. Dpkg-ftp has some bugs to be shaken out > before that happens. Also, the installation menu in 1.1 doesn't set up > PPP and SLIP for you, which is necessary for many network > installations. So expect this in 1.2 . The target date for that is 3 > months after 1.1 is released. > > Thanks > > Bruce By way of a minor bug report.. when version numbers change, the files are not always seen by dpkg to be newer, for example, I have a copy of lynx, which is listed by dpkg -l as '2.4.2-2' but the version in /unstable/* is listed as 'lynx-2.4-FM-960316-1.deb' ... now dpkg-ftp will fetch the file, but dpkg thinks it is downgrading and balks... I assume the lynx with the letters in the name is the newer... there have been a few such packages in the past. '-force' usually works of course. What about the idea of having an 'expert' option on dpkg that would prompt you for 'y/n' if a particular force-option was necessary? [EMAIL PROTECTED]/GNU__Linux__1.3.77___ "Love is a snowmobile racing across the tundra and then suddenly it flips over, pinning you underneath. At night, the ice weasels come." -- Matt Groening
