Re: Can run testing with slow net connect?
On Sat, Jul 21, 2001 at 02:54:11PM -0500, David A. Rogers wrote: | I've got a 56k modem on my home machine. Is it feasible/reasonable for me | to run testing on this machine? I've got 2.2r2 CDs. Any guesstimates as to | how long it will take for the initial upgrade to testing? A few weeks ago I upgraded from 2.2r3 (I have 2.2r2 cds and had upgraded via modem to r3) to testing. I have a 33.6 modem. The estimate was ~16 hours for the download. The actual was fairly close -- I would use ^C to stop apt-get when I had to disconnect the modem, then apt-get would resume where it left off next time (even mid file, very cool!). It is certainly feasible, especially if you don't mind spending a few late nights (ie do the downloads after the point when no one calls). -D
Re: Email line-length defaults to about 76; how to increase?
On Sun, Jul 22, 2001 at 10:22:07PM +0200, Matthias Richter wrote: | Andy Mott wrote on Sun Jul 22, 2001 at 07:45:14PM: | Bud Rogers ([EMAIL PROTECTED]) wrote: | | Ah, yes. I didn't think of MUA's that call external editors. I've | used kmail for the last year or so, and gnus for years before that, | so I tend to think of the reader and the composer as one and the | same. Are there not some settings in .muttrc that are passed to vi | or whoever? | | | Having been slapped for having a too-long line-length, this was a | question I was asking myself (actually, has been for a while). I also | use mutt, and spent quite a while searching the documentation for the | variable to change it. It turns out the setting is in the editor (I | use vim) - as soon as I added 'set textwidth=70' to my .vimrc I had | the answer | | You can of course set this variable in your .muttrc as well, so it is | only used when vim gets called from mutt: | set editor=vim -f \set textwidth=70\ Better yet in your .vimrc put au FileType mail set tw=70 When you are writing a mail, vim will set the textwidth to 70 and you can have a different setting for other stuff (I usually use 80 for code and plain text documents). The autocommand stuff is really cool. -D
Re: [OT] HUB question
On Sun, Jul 22, 2001 at 05:42:32PM -0400, Sunny Dubey wrote: | hey, | | I have two hubs, and I'd like to be able to connect them to each other. Both | have a port called uplink port, do I need to use a cross over cable to | connect both hubs using their uplink ports, or should I use a normal cable to | connect the both of them? You can : o use a normal cable to connect the uplink of one to a normal socket of the other o use a crossover cable to hook up the two uplink sockets (maybe) o use a crossover cable to hook up two of the normal sockets (maybe) What is needed is for the send of one interface to hook up to the receive of the other. This is what crossover cables are for in the presence of identical connectors. Manufacturers typically provide a socket labeled uplink that is crossed over internally so that you don't need an external crossover cable. HTH, -D
Re: dist-upgrade from potato to woody
On Sat, Jul 21, 2001 at 09:17:27PM +0100, Graham Ward wrote: | I just tried to upgrade my system from potato to woody. I believe | these are the correct steps: | | (1) replace potato with woody everywhere in /etc/apt/sources.list | (2) apt-get update | (3) apt-get dist-upgrade. These are good. | When I do step (3), I see (among other things) the message | | | WARNING: The following essential packages will be removed | This should NOT be done unless you know exactly what you are doing! | sysvinit util-linux (due to sysvinit) | 544 packages upgraded, 87 newly installed, 36 to remove and 6 not upgraded. | Need to get 348MB of archives. After unpacking 173MB will be used. | You are about to do something potentially harmful | | | On the face of it, removing sysvinit looks like a bad idea, so I | stopped at this point. Has something gone horribly wrong with my | set-up, or is this in fact harmless? Yeah, you kinda want to have an init on the system :-) (without it you won't be able to boot). It is possible (but I'm not looking at packages.debian.org right now) that sysvinit isn't in testing. I recommend putting potato in sources.lst after woody. apt is smart enough to get just the latest version of the packages. This way you will get the new stuff that is in testing, but you won't lose stuff that is in potato but not woody. HTH, -D
Re: kernel 2.4: where's the FM?
On Mon, Jul 23, 2001 at 02:54:45AM +0200, Joost Kooij wrote: | On Sun, Jul 22, 2001 at 07:59:15PM -0400, dman wrote: | I've just spent some time on kernel.org and elsewhere, but I can't | find any FMs to R regarding the procedure to upgrade from kernel | 2.2.19 to 2.4.x. Is an apt-get install enough (and point grub at the | new image of course)? Where can I find an FM to R on this? | | linux/Documentation/Changes I don't seem to have that on my system right now. Wouldn't that be a part of the new package? How can I read the doc before I install the package? THanks, -D
Re: kernel 2.4: where's the FM? (found)
On Sun, Jul 22, 2001 at 09:34:34PM -0400, D-Man wrote: | On Mon, Jul 23, 2001 at 02:54:45AM +0200, Joost Kooij wrote: | | On Sun, Jul 22, 2001 at 07:59:15PM -0400, dman wrote: | | I've just spent some time on kernel.org and elsewhere, but I can't | | find any FMs to R regarding the procedure to upgrade from kernel | | 2.2.19 to 2.4.x. Is an apt-get install enough (and point grub at the | | new image of course)? Where can I find an FM to R on this? | | | | linux/Documentation/Changes | | I don't seem to have that on my system right now. Wouldn't that be a | part of the new package? How can I read the doc before I install the | package? Duh ... after a bit more thought I installed the kernel-doc package and it is there, as /usr/share/doc/kernel-doc-2.4.6/Documentation/Changes.gz FYI it says --- The latest revision of this document, in various formats, can always be found at http://cyberbuzz.gatech.edu/kaboom/linux/Changes-2.4/. --- Thanks for the pointer in the right direction. BTW, where are those qoutes found regarding kernel swap issues that were posted earlier? -D
Re: [users] Mail from OE to linux and more
On Sat, Jul 21, 2001 at 03:31:54AM -0700, Vineet Kumar wrote: | * Mark Wagnon ([EMAIL PROTECTED]) [010709 15:31]: | Do you mean with or without? I'm a little confused. Mutt has some | option to change the from header for outbound mail. I don't use it, so | | (judicious use of hedgeclippers above) | | In order to set the from address that mutt wil use, try setting the | from variable. You can have this do fun things for mailing lists FYI, the from command doesn't work for me. Instead I use the my_hdr command like : # used to make the From: header correct. my_hdr From: [EMAIL PROTECTED] (D-Man) At the moment I don't really need it (because I am sending directly from that account) but if I send stuff from my Linux box I need it or else the MTA will make my linux account the from (and thus the default reply-to) address which won't work. HTH, -D
Re: it keeps crashing
On Sat, Jul 21, 2001 at 11:00:52AM +0200, Joost Kooij wrote: | On Sat, Jul 21, 2001 at 09:43:39AM +0200, Guy Geens wrote: | Martin == Martin F Krafft [EMAIL PROTECTED] writes: | | Martin instead, it keeps crashing on me... kernel panic in pid 0 | Martin process swapper. however, memtest86 reports no errors for | Martin the RAM chip, and badblocks, run with the destructive write | Martin option, reports no bad blocks within the swap partition. | | My guess is that you simply have not enough memory. | | AFAIK 8MB is enough to boot linux, or it should be. I can verify this. I am using an old Gateway2000 i486sx, 25MHz with 8MB RAM right now to write this. I have ~32MB swap. It is running as a router for my dial-up connection, but it also has a keyboard and monitor so I am using it for ssh-ing to school to read my mail (and write this). I have kernel 2.2.19 and have no stability problems (only performance problems :-), for example when I use dpkg) -D
Re: Motherboards
On Sat, Jul 21, 2001 at 04:34:23PM +0100, Keith O'Connell wrote: | Hi, | | I want a new machine, and for fun and education I am going to build it | from scratch. I have pretty much decided on an Athlon 1.2GHz. I will run | a small partition with Windows Me on it, but it will predominantly be up | in Debian. | | I thought I would consult here as to the motherboard that will give the | least compatibility problems with the various chip sets available and | for on-board sound. I have a Gigabyte motherboard for my Duron 750 processor (it should be able to handle Athlons as well). It is based on the AMD 750 chipset and I have had no problems with it (yet anyways ;-)). The Tyan Thunder K7 board looks cool (its a SMP board for the AthlonMP processors and has a lot of stuff on-board) but has a price tag to match its feature list (~$600). On this list I have heard a lot of issues with the VIA KT133 chipset, but I have no experience or documents to back that up. HTH, -D
Re: Hardware Question
On Sat, Jul 21, 2001 at 02:06:29PM -0400, Allan M. Wind wrote: | On 2001-07-21 13:17:14, Adam Bell wrote: | | What I would like to do is have one box, running Debian, which has a | constant routable IP (via cable or some other sort-of high speed protocol) | and a normal domain name. | | You could have someone serve dns for your domain, ideally you would | want someone to do secondary dns for you anyways. Alternatively, you | could consider something like dyndns.org which will give you a | hostname in their domain for free (donations are encouraged). Thanks, I'll have to look this place up :-). | This will act as a mail server, samba server, FTP server, and | internet sharing gateway for like four or five machines sitting | behind it (and also make info on it available to the owners of these | machines while they are wandering the world via IMAP). | | I suspect that you will not be putting a heavy load on any of these, | so a low-end machine (celeron/duron) should do just fine. Yeah, it all depends on the load. If you will do a lot of work, make sure you have a fast disk first. | So my question is this: | | What kind of hardware, and in what configuration, would be best to throw | at this problem? What I'm assuming is that I'll need an interface for the | modem, an interface going into a hub, and interfaces for all the clients. | Is there a better way to do this? | | wan --- gw --- hub pc1 | \--- pc2 | \-- pc3 | | You might be able to find a switch not much more expense than a hub | these days, but you should go with 100 Mb/s network (cables, network | cards and hub/switch). I'm having good experiences with the D-Link DSS-8+. It is an 8 port (plus 1 uplink) 10/100Mbps ethernet switch. I got it from CompUSA for ~$80. NICs are fairly cheap -- you can get good Linksys/Netgear/D-Link NICs for around $20 each. No-name NICs are fine also if you know what chipset they use (I have one, its a rtl8139). Routing is trivial to setup with the 2.2 kernels -- simply 'apt-get install ipmasq'. I don't know about the 2.4 kernels (other than they use iptables instead of ipchains). -D
Re: MUA with html support
On Sat, Jul 21, 2001 at 11:56:31AM +0200, Alexander Steinert wrote: | I do it with a combination of a line in my muttrc and a couple | of lines in my /etc/mailcap | | first put this line in /etc/mailcap: | | text/html; /usr/bin/links -dump '%s'; copiousoutput; description=HTML Text; nametemplate=%s.html | | Which version of links supports -dump? The one in woody does (I don't have the number right now). Version 0.84 (in potato) doesn't. The woody version also has better color support too. | I've also got auto_view lines and corresponding mailcap lines | for postscript, tex, and M$ word!! (antiword). | | Could you send the ones for ps and tex, please? For PS I would use ggv (GNOME Ghostview). I'd like to know how you handle (La)TeX files without having the .aux .log and other temp stuff laying around. Also, what do you use for M$Word? -D
Re: MUA with html support
On Fri, Jul 20, 2001 at 03:16:55PM -0700, Duncan Watson wrote: | Hi all, | | I am a current mutt/fetchmail user and have been for two years or so. I | use Linux at home and at work and keep about 630MB of mail in various | maildirs. I use mail to keep all sorts of notes and crap as well. | | Overall I love mutt and am greatly attached to many little features of it, | BUT I am constantly getting mail that only looks good/readable in html | format (including color). Also there is some weird interaction between | fetchmail and the local exchange server that causes exchange to strip the | rtf portion of messages that I download. In other words I get exchange's | plain text conversion only. This doesn't affect on of my co-workers that | uses kmail (both as an MUA and half an MTA (download only)). | | So I am looking for either a super duper lynx-like tool that can render | html attachments pretty and colorful or a new MUA that has similiar | features. The links in woody can handle color (at least to some extent), you might want to try that. For sending HTML mail you could try something like gnp (It is an editor intended for HTML) and set the Content-Type: header yourself. | I was tempted to look at balsa but the package is currently broken | and downloading half of gnome always annoys me. Balsa has a fairly nice GUI and can render HTML mail (and images, IIRC) but it doesn't handle mailing lists (like mutt does) and I don't think it can send HTML mail. HTH, -D
Re: embarrassing X question
On Thu, Jul 19, 2001 at 04:04:09PM -0400, Mike wrote: [snip] | This might be a kinda dumb question, but does X need to be running on the | remote machine? I've tried having X running on the remote machine, but it | hasn't seemed to make a difference. X must be running on the local side, obviously. It doesn't need to be running on the remote side because you aren't trying to connect to it. If the remote side is, for example, a headless system then you will have lots of trouble trying to run X on it :-). I don't know what your problem is since you seem to have the config set properly. Hmm, maybe a firewall issue? Hopefully someone with more knowledge will provide some suggestions. -D
Re: X question
On Thu, Jul 19, 2001 at 09:40:16PM -0400, Andy Saxena wrote: | It has to do with keymapping. I am feeling too lazy to look up my notes now, | and hopefully somebody will come along to help. | | In the meanwhile use CTRL-A for Home and CTRL-E for End, if you aren't | already :). If it is a keymap thing and you don't like emacs bindings (wink) put set editing-mode vi in your ~/.inputrc file then you can use vi's ^ for beginning of line and $ for end of line (when in command mode). -D
Re: Again NE2000 network device
On Fri, Jul 20, 2001 at 10:58:56AM -0700, [EMAIL PROTECTED] wrote: | On Fri, Jul 20, 2001 at 06:03:25PM +0200, Bj?rn Fischer wrote: | Hello, | it is me again, trying to start the ne module. Now the module is actually | trying to start, but alway quits with the message Resource or device busy. | I started the ne module with the following parameters: | | modprobe ne io=0x260 irq=9 | | The io and irq are the ones windos promted when having a look at the | resources of the card. | I have had a look into either /proc/ioports and /proc/interrupts, and 0x260 | and 9 was not occupied by a different device. | | Does someone of you know what to do? | | Very thankful for an answer | | Is it a PnP or a soft-config card? If it is PnP, then you will want to turn it off. I have a couple of NetGear EA-201 (ISA PnP NE2k clones) that I first used the DOS utility they came with to disable PnP. Then I could insmod the ne module and specify the base io and it works quite well. HTH, -D
Re: Characters shown as ? in mutt
On Thu, Jul 19, 2001 at 09:55:45AM -0500, Darryl L. Pierce wrote: | On Wed, Jul 18, 2001 at 10:45:34PM -0400, D-Man wrote: | On Thu, Jul 19, 2001 at 07:58:17AM +1000, Doug Hespe wrote: | | Hi J?rgen, | | I am having difficulty reading your name in mutt where it shows up as | | J?rgen. The problem is even worse in Xemacs and in regular Emacs where | | it is shown as =?ISO-8859-1?Q?=22J=FCrgen_A=2E_Erhard=22?=. | | I don't know where to set the charset for mutt's internal pager or for | emacs because I use less as the pager and vim as the editor. To | correct the display of upper-ASCII characters in less (and get rid of | the binary file warning) I added | export LESSCHARSET=iso8859 | to my .bashrc. It wasn't a problem with vim's defaults. | | How about when telnet/ssh'ing into a machine? I sometimes read the email on my | laptop from my desktop machine at work, and most extended characters show up as | /?/ as well. Ok, to tell more of the story I really put setenv LESSCHARSET iso8859 into my .cshrc on the Solaris box at school (they have csh as default, I always run bash after I log in) where I am now reading and writing this mail. You need to tell less what charset to use. If less is running on a remote box you need to set the environment variable on the remote box. When you log in to a box remotely it is just as if you are sitting in front of it. Do the same thing there as you would do for your local box. -D
Re: embarrassing X question
On Thu, Jul 19, 2001 at 08:27:49AM -0400, Richard Black wrote: | For some reason, I can no longer remote login to another terminal and | display stuff on mine! This started happening last week (with, | possibly, the changes to gdm...) | | I have tried many different things. Typical is something like: | | [local machine] | xhost + | rlogin remote | | [remote machine] | export DISPLAY=local:0.0 | nedit | | But all I get is: NEdit: Can't open display I would use ssh instead of rlogin if you can. Also, enable the ForwardX11 option in ssh. If you do this then the display will be setup for you and it will be encrypted as well. This is also the easiest (only?) way to display stuff back on a masq'd box. -D
Re: Mutt Question
On Thu, Jul 19, 2001 at 07:17:11AM -0400, Todd V . Rovito wrote: | I have procmail set up to filter my mail into folders for the list | servers I belong to. Some days I just don't have the time to read | the mail in that particular folder (or list server), I would like to | be able to perform a mass delete in that folder. What keys can I use | in mutt to delete all the messages out of a folder? If you sort by threads, ^D is the shortcut for kill-thread. If you don't want anything you could just hold down the 'd' key until you hit the bottom. If you want to blow away everything in the folder just remove it (ie 'rm ~/Mail/this_folder') since they are just plain text files. -D
Re: a 'who called me?' variable?
On Thu, Jul 19, 2001 at 06:28:54PM +0200, Robert Waldner wrote: | | On Thu, 19 Jul 2001 12:10:12 EDT, D-Man writes: | Based on your examples there, $TERM might be helpful. For example in | my .cshrc on the school's Solaris box I have | | if ( $TERM == linux ) | bash exit | end | | Hmm, `exec bash` would be better as it wouldn´t let the old shell | running in the background ;-) That's cool, it works. Thanks. -D
Re: embarrassing X question
On Thu, Jul 19, 2001 at 01:25:58PM -0400, Mike wrote: | D-Man wrote: | | I would use ssh instead of rlogin if you can. Also, enable the | ForwardX11 option in ssh. If you do this then the display will be | setup for you and it will be encrypted as well. This is also the | easiest (only?) way to display stuff back on a masq'd box. | | How do you do this? I've been trying to do this for some time now with no | success. Every time I get: | | [EMAIL PROTECTED]:~$ ssh -f hal9000 xterm | [EMAIL PROTECTED]'s password: | [EMAIL PROTECTED]:~$ xterm Xt error: Can't open display: hal9000:10.0 | | hobbiton is the local machine here, and hal9000 is the remote system I'm | trying to connect to while wanting the xterm (in this case, anyway) to | display here on hobbiton. I've got the ForwardX11 option set to true on | both machines, both in the sshd_config and the ssh_config Is there anything | else I need to do? Other than enabling ForwardX11 in both the server and client (sshd and ssh) I don't think you need to do anything. What happens if you login and get a shell, then run xterm? The error message shows that DISPLAY was set properly (sshd creates a display on the server, 10.0, which it reads from, encrypts, and sends to the client who passes it on the local DISPLAY) but that display couldn't be opened. I'm wondering if maybe ssh is closing the connection too soon. On the Solaris box at school /etc/sshd_config has X11Forwarding yes X11DisplayOffset 10 On my Debian box in ~/.ssh/ssh_config I have in the section for the remote machine ForwardX11 yes On the client side you can use the -X option instead of the config file. I like the config file because my options become persistant. HTH, -D
Re: Urgent! everything is Connection refused!
On Thu, Jul 19, 2001 at 10:39:50AM -0700, Matheson Cameron wrote: | Hey, | | I'm kind of in a fix at work here. I installed debian | at home, and i brought the machine in here at work, | but i can't connect to any ftp sites. I've configured | everything for the LAN at work, but not everything's | happy, here's what works: | | -ping (myself, the lan, and anything on the internet) | -telnet (myself, the lan, not internet) | -ftp (myself, the lan, not internet) | | When I do try to open a connection to ftp.debian.org | or ftp.gnome.org it instantly says 'Connection | refused'. But it's not even going out over the | internet (i don't think, it seems to just die). My | other debian box is working very happily, but i can't | seem to figure out what's wrong Run traceroute to see how far the packets are getting. My guess is that they are getting lost somewhere in the LAN. Did you set a default gateway? Your box needs to know where to send packets destined for machines it doesn't have direct access to. Also, does nslookup work? -D
Re: j2dsk-doc-installer
On Thu, Jul 19, 2001 at 11:06:47AM -0700, Mike Pfleger wrote: | Hello. | | I'm running testing, and last night I tried to install j2sdk1.3.1 from | Blackdown. Everything went fine, until apt got to the doc-installer. | It started to unpack the zip, and then dies when it wants to move the | doc dir from a tmp dir it's made. It looks in the _wrong_ place, so | there's apparently something hosed in the installer deb. Uh, hmm. I installed it yesterday (I started downloaded at around noon) and had no problems. I was using the unc mirror and apt-get. (it's ~25MB and took several hours on my little dial-up) I haven't tried running it yet since I found the problem with ant so it is now working with jdk1.1 (and jikes). [ For those who are curious, my company is using features from 1.3 but the debs are only for 1.2. I had replaced the jar file, but not really -- just the symlink to the real jar. Once I replaced the real jar with 1.3 it was fine ] | Any ideas? On that note; would it have been better to have grabbed | the woody pkgs instead of the potato pkgs? Oh, I guess so. I used woody instead of potato because I'm running woody. -D
Re: How do you edit the /etc/hosts with the smtp information - fetchmail problems
On Thu, Jul 19, 2001 at 08:10:24PM +0200, Martin F. Krafft wrote: | also sprach Joost Kooij (on Thu, 19 Jul 2001 12:25:30PM +0200): |telnet 127.0.0.1 25 | | or | |telnet localhost smtp | | or | | netstat -lAinet | grep smtp | | which is better since it gives a more certain answer whether the smtp | port is bound to, or not. also, telnet is a protocol, and *if* you telnet provides console access to a port. Sure, telnet is a protocol, but the telnet protocol just negotiates terminal capabilities. If you're not connecting to telnetd you get no capabilities aside from sending a character and receiving a character. If you telnet to the smtp port then you are being your own MTA and must type all the SMTP protocol data manually. It is useful for testing the MTA you are connecting to because you see _everything_ it tells you and it is also useful for toying around with network protocols. -D
Re: a 'who called me?' variable?
On Thu, Jul 19, 2001 at 11:51:06AM -0400, Daniel Patrick Berdine wrote: | Is there an environment variable in bash that can be used to tell where | bash is running from? For example, is there a variable I can test in | my .bashrc to tell whether I am running remotly, from a tty, from | konsole, an xterm, etc.? I'v looked at the Advanced Scripting HOWTO, | but the only thing that looked promising was $BASH_ENV which seems to | be empty. Based on your examples there, $TERM might be helpful. For example in my .cshrc on the school's Solaris box I have if ( $TERM == linux ) bash exit end so when I login from my linux box (console) I automatically get bash and I don't have to exit csh after I exit from bash. -D
Re: But ....
On Tue, Jul 17, 2001 at 05:19:55PM +0200, Pietro Cagnoni wrote: | On Tue, Jul 17, 2001 at 09:23:53AM +0200, Pietro Cagnoni wrote: | | 2) reboot in maintenance mode (or whatever is called in english: in | | italian is Modalità provvisoria: 640x480 screen with very few colors); | | FYI it is called Safe Mode in English. You may need to press F5 | during startup to get the menu to select Safe Mode from. | | safe? does it mean you can't make any damage that way? or maybe it | inhibits the computer self destruction, and it'safer for you to use? or | is it the mode you need when you're closed in a safe? | | Modalita' provvisoria in italian could be translated temporary mode, | and has some of the meanings of unstable mode. | | so, where's the stable mode? | | maybe i should open a call with tech support... Hehe. :-) I think it means pick 'safe' settings for your video card and don't start up any fancy services so that, maybe, the user can fix what's broke. Remember that Windows doesn't know what a virtual console is and _always_ needs a graphics card and mouse. -D
Re: Safe File Manager to run as root ?
On Wed, Jul 18, 2001 at 12:54:19PM -0400, Hall Stevenson wrote: | Is there such thing as a GUI File Manager that any | security and safety consious Debian users would | use, as ROOT, to manage a file system (i.e. move, | copy, change permissions) ?? Is it just a better | practice to use CLI w/ suid to make those kind of | changes. | | If you're referring to gmc's pop-up warning box (do other file | managers do it also ??), I wouldn't worry about it. You can do | as much, if not more, damage at the command line as you can do | with a GUI file manager... | | Actually, in what way is gmc more dangerous than mc ?? | Starting up mc as root provokes no similar warning. I've never used mc, but I did try gmc a while back. I stopped using gmc when I tried to _copy_ some files from a floppy. Instead the default drag-n-drop action is move. Also, there was a problem writing to the destination. Instead of canceling the action, gmc proceeded to remove the originals even though the copy hadn't been made. I sitck with the CLI now. Also, suppose there is a bug in gmc. Now suppose that you trigger that bug while running as root. And suppose that the bug is very serious ... (filesystem corruption or something, maybe) I think the developers are just warning you and it is your own fault if something like that happens. -D
Re: Safe File Manager to run as root ?
On Wed, Jul 18, 2001 at 02:22:30PM -0400, Hall Stevenson wrote: | I've never used mc, but I did try gmc a while back. I | stopped using gmc when I tried to _copy_ some files | from a floppy. Instead the default drag-n-drop action | is move. | | From being a fairly experienced windows user, and apparently | one of the few who understand and use drag-n-drop (I've seen | very few people who actually do it...), I have a pretty good | grasp on the situation. It appears that the gmc developers | either copied Windows Explorer behavior, or Explorer is | copying something before it. I don't know, nor care. I've used | Sun's file manager from Open Windows on the Sparcs we used to | use at work, but don't recall it's behavior. Anyone ?? I used/use explorer a lot in windows (actually less now with a good cygwin install). | Here it goes: | | * If you drag a file from one *drive* to another, the default | is to *copy* | * If you drag a file from one directory to another directory, | the default is to *move* That sounds about right (for windows anyways). | Since the *nix world makes everything a dir, you're going to | move a file instead of copy it. Going from a floppy to your | hard drive follows the two drives logic. Yeah, it would be just a dir. | By the way, in both Windows and gmc, dragging with the right | button in Windows and the middle button in gmc prompts you as | to what you want to do. The choices are move, copy, and | link/shortcut. I actually do it this way all the time... A | good habit, I think. I do that a lot when I'm in windows, I don't remember if I tried it in gmc or not. The _real_ problem wasn't the default was move situation, but rather that it removed the originals even though there was an error and the new ones couldn't be created. That left me without _any_ copy of the files. It should have stopped, and left the originals alone, when it found out the destination files couldn't be created. This is what 'mv' does. -D
Re: Segmentation error
On Wed, Jul 18, 2001 at 10:01:56PM +0530, [EMAIL PROTECTED] wrote: | What is segmentation error and how do you solve it ? A segmentation fault occurs when a program doesn't behave with its memory. Generally this only occurs with programs wirtten in a language that provies direct access to memory like C an C++. To fix it you need a d development environment and access to the source and to learn how to program in the langauge (C/C++). Since you are obviously not a programmer, where did you see this error? It could be that you have the wrong libraries on your system or someting like that. Segmentation faults can occur with mismatched library versions too. HTH, -D
Re: Characters shown as ? in mutt
On Thu, Jul 19, 2001 at 07:58:17AM +1000, Doug Hespe wrote: | Hi J?rgen, | I am having difficulty reading your name in mutt where it shows up as | J?rgen. The problem is even worse in Xemacs and in regular Emacs where | it is shown as =?ISO-8859-1?Q?=22J=FCrgen_A=2E_Erhard=22?=. I don't know where to set the charset for mutt's internal pager or for emacs because I use less as the pager and vim as the editor. To correct the display of upper-ASCII characters in less (and get rid of the binary file warning) I added export LESSCHARSET=iso8859 to my .bashrc. It wasn't a problem with vim's defaults. HTH, -D
Re: NFS alternative
On Tue, Jul 17, 2001 at 06:53:51PM +1000, Steve Kowalik wrote: | On Tue, Jul 17, 2001 at 04:05:10PM +1000, Brian May uttered: | Well, in theory, all UDP packets could be numbered, much like TCP | packets, and you get exactly the same reliability TCP offers. This | | I doubt that. UDP isn't a connection oriented protocol, and as such, it | can't deal with a packet out of order. Actually, what Brian says makes sense. UDP itself can't deal with a packet out-of-order, but Brian says that the application using UDP can deal with it itself. Kind of like re-inventing the wheel. He said the nice thing about TCP, in this situation, is that other people have already done that dirty work so he (the app developer) doesn't need to worry about it. Thanks everyone for the comments on NFS and diskless terms. -D
Re: Kernel upgrades
On Mon, Jul 16, 2001 at 10:18:44PM -0500, [EMAIL PROTECTED] wrote: | I've noticed that progeny Debian has auto-upgrade of kernel upgrades. How do I do this with my Desktop Debian system, and what are the cons? apt-get doesn't _automatically_ change (upgrade or downgrade) your kernel if you use, for example, 'upgrade' or 'dist-upgrade'. The reason is simple -- the kernel is the most critical part of your system and care must be taken to ensure that you don't break it. Bootloaders such as LILO must be reset (ie run 'lilo') everytime you change the kernel image. Also, sometimes people make custom kernels and they would not want apt to decide that the kernel on a server somewhere is better. It also protects against some attacker from putting a bogus kernel package on an apt-gettable site so that those who use the site provide a nice wide-open door for him. You can upgrade your kernel when you want to by 'apt-get install kernel-image-'. When I did that to upgrade from 2.2.18presomething to 2.2.19 I had no trouble at all. I use grub so I didn't need to change anything there for it to boot fine. HTH, -D
Re: But ....
On Tue, Jul 17, 2001 at 09:23:53AM +0200, Pietro Cagnoni wrote: | 2) reboot in maintenance mode (or whatever is called in english: in | italian is Modalità provvisoria: 640x480 screen with very few colors); FYI it is called Safe Mode in English. You may need to press F5 during startup to get the menu to select Safe Mode from. -D
Re: boot problem
On Wed, Jul 18, 2001 at 01:12:56AM +1000, Sam Varghese wrote: | On Tue, Jul 17, 2001 at 10:43:26AM +0100, J.A.Serralheiro wrote: | well maybe you misconfigured your video card. | Run xf86config. Its a text based environment and re-do the configurations. | It may help. be sure you know what chipset your board has. I think yuo can | also choose and general chipset that should work on most cards. | If you can put it to work, then use XF86Config. | | It's XF86Setup, not XF86Config. Actually, XF86Config (note capitalization) is the name of the file where the config is stored. XF86Setup and xf86config (note capitalization) are two different configuration tools for X (3.3). If you have X4 just run 'XFree86 -configure' (then hand-tweak the result). -D
Re: zip drive?
On Tue, Jul 17, 2001 at 11:32:08AM +0100, Frank Zimmermann wrote: | Leonard Stiles wrote: | Note that you can alternatively specify auto as the file-system | type, in which case it will be auto-detected by mount. | | I realized some problemes with auto and vfat, namely long | filenames. When I tried the auto/detection it always showed 8.3 Maybe it isn't really vfat? I do remember some problems with preformatted zip disks and MSDOS vs. Win filesystem types. I liked to take a few of my disks and use Linux to make the disk vfat on partition 1. | filnames. Well, most times I use vfat foramttd ZIPs and the few | times I use something else I use | mount -t hfs /dev/hdd /zip | (why don't I need a partition number in this case?) I guess because the entire disk is 'hfs' (the partition starts at the very front) or maybe the hfs part of the mounter can figure it out? (I have no idea what 'hfs' is nor have I used it) -D
Re: seting up a gateway
On Tue, Jul 17, 2001 at 07:29:12PM +, Hereward Cooper wrote: | Is there a gateway HOWTO? I found something along that lines but none of it | applied to a debian system. It would be titled IP Masquerade and Linux Networking or something like that (on linuxdoc). | Also what do i need a 2.4.6 kernel for it to work, am i right in | thinking that it is different from 2.2.x? I know you will need iptables for a 2.4 kernel (haven't been there yet). For 2.2 kernels it is as simple as apt-get install ipmasq (which will install ipchains which does the real work) and then tell the others on the LAN to use your IP as the default gateway. -D
Re: gnucash 1.6.1-2
On Tue, Jul 17, 2001 at 11:06:28AM -0700, William S. wrote: | I used apt-get install to get the latest unstable | version of gnucash. When I try to run it I get: | | gnucash: error while loading shared libraries: | libqthreads.so.9: cannot open shared object file: No | such file or directory | | I am using kernel 2.4.5 with the latest unstable | packages. I currently have libguile9, libc5 and libc6 | installed and all the other packages apt-get deemed | necessary. Do you have libqthreads? (Is that really supposed to be a 'q' or a 'p'?) | How do I get this to run? Try 'ls -l /lib/lib*threads* /usr/lib/lib*threads*' and see if you really do have that library or not. If not, find out what package has it (http://packages.debian.org) and install it. -D
Re: seting up a gateway
On Wed, Jul 18, 2001 at 09:39:54AM +1000, Sam Varghese wrote: | On Tue, Jul 17, 2001 at 04:15:06PM -0400, D-Man wrote: | I know you will need iptables for a 2.4 kernel (haven't been there | yet). For 2.2 kernels it is as simple as apt-get install ipmasq | (which will install ipchains which does the real work) and then tell | the others on the LAN to use your IP as the default gateway. | | Do you need to specifically install ipchains? I've been led to | believe otherwise. Or does this set up your basic ipchains rules? The 'ipmasq' package depens on 'ipchains' so if you apt-get install it apt will tell you that it will install ipchains also. This is the beauty of apt's dependcy resolution. -D
Re: Installing Java
On Sun, Jul 15, 2001 at 06:08:49PM -0700, Nicole Zimmerman wrote: | | As far as java itself goes, there are blackdown java 2 packages at: Yeah, Sun didn't want to pay to make a Linux JDK so instead the Blackdown people agreed to sign NDAs with Sun for the privilege to create Linux binaries. It is non-free because the NDAs prevent the Blackdown people from releasing any of the source. I believe that if you go to Sun's site and try to download a Linux jdk you get the result of the Blackdown people's work. -D
Re: zip drive?
On Mon, Jul 16, 2001 at 12:03:13AM -0400, Jeff Maxson wrote: | On Sun, 15 Jul 2001, Petr [Dingo] Dvorak wrote: | On Sun, 15 Jul 2001, Jeff Maxson wrote: | | JM mount -t vfat /dev/hdd1 /zip | JM | JM gets me | JM | JM hdd: The drive reports both 100663296 and 100646912 bytes as its capacity | JM hdd: hdd4 | JM mount: mount point /zip does not exist | | first do mkdir /zip because it doesn't exist :) | | then try mount -t vfat /dev/hdd4 /zip | | duh. Thanks, it worked. First blush I thought the problem was the first | warning, not the second... Actually, both of them were problems. For some weird reason the pre-formatted vfat zip disks use the 4th partition as the only partition, not the first. Hence the '4' instead of the '1' in the mount command. If you want to, you can repartition the disks to use the first partition and even make them ext2 if you don't intend to put it in a windows box. You can also add a line to /etc/fstab that will let you simply say mount /zip. -D
Re: D-Link DFE-530TX+ (again)
On Mon, Jul 16, 2001 at 12:57:42AM -0400, Evan Flynn wrote: | Hi. | | I tried the Realtek 8139 driver with no parameters and unfortunetely i got | the device busy error. Please note that I am INSTALLING Debian and Hmm, that's not fun. What are the numbers on the card itself? Is there a revision number or a manufactured date? With my LinkSys card I noticed it has Rev 2.0 on it and it uses a Lite-On tulip-chip clone. It works fine with the stock tulip driver included since 2.0.36 kernels. On LinkSys' site they sayt the current revision is Rev 4 and uses a chip they made themself, thought the card has the same model number. Several times on this list people have complained about the 2.2.x tulip driver not working with their LinkSys card and others have explained that they need the new one in 2.4.x kernels. Maybe your card is in a similar situation where the chipset changed slightly, but just enough to make the driver not work. | therefore i can't compile drivers (well maybe) or use kernel 2.4. Well I | compiled the version that came with the card (they only provided source) on | a system with RedHat 6.0 on it, didn't appear to work with Debian though. Is that RH system using the same kernel? It should work as long as the kernel version (and config) is the same. | This is really becoming quite frustrating. I can imagine. | Do I just need to enter parameters, or is there some way I could | compile a driver, or perhaps it is a bad BIOS setting? Maybe I'm | just a clueless newbie. In order to compile a driver you need to have a system with a compiler installed. Kind of a chicken-vs-egg problem you have here. | Anyway, the card works fine in Windows ME. Any help at all is | appreciated, and I don't want to ditch a linux distro over a nic Glad you're not giving up so easily :-). Can you download CD isos to install from? Then you wouldn't need the NIC right away and you could try compiling different versions of drivers. Alternatively you could try borrowing a different NIC from a friend to try and get a base system installed. Then you could deal with getting this card working. Where do you live? If it is around Rochester, NY, maybe I could help out? -D
Re: Why is setting up X so arcane? (SOLVED!!!)
On Mon, Jul 16, 2001 at 03:22:09PM +0200, Sebastiaan wrote: | Hello, | | yes, finally succeeded. A guru at school assisted me in building a new | XF86Config file. Cool! | I have attached a working XF86Config (sorry, its for the greater good), | works with the SVGA driver as well as with S3. Let this configuration | help those who have the S3 Vision 964 card. That's good as it will turn up in searches of the archives :-). It's not too hard for me to delete my local copy :-). -D
Re: How to install sound on Debian Linux
On Mon, Jul 16, 2001 at 05:27:04PM +0200, Ellenkamp, Guus wrote: | I have an MAD-16 soundcard and thought I installed all the required stuff, I am not familiar with this one. | but program's don't recognize the card. How to install correctly? You need to edit a file in /etc/modutils (I think /etc/modutils/local_config is a good name) and include some lines like alias sound sb options sb io=0x500 irq=5 dma=1 and then run 'update-modules' as root. I don't have my box in front of me, but the options I have in my config are for my old system which had an ESS1869 card. From RTFM-ing I learned that it uses the 'sb' module and by checking windows I found the right options for io base, irq, and dma channel (the ones above are just random numbers). If you choose the wrong option for, eg, dma you will get sound but it will be very very crappy. So basically, RTFM to find out which module you need for that sound card and then figure out which options you need to specify for it. Either check windows (if you have it dual-booting) or maybe others can suggest ways of determining that info directly from linux. HTH, -D
Re: NFS alternative
On Mon, Jul 16, 2001 at 03:56:02PM +1000, Brian May wrote: | Jeremy == Jeremy Gaddis [EMAIL PROTECTED] writes: | | Jeremy Two possible suggestions: - tunnel the NFS traffic over an | Jeremy SSH traffic (similar to remote X sessions) | | Can you use SSH to tunnel UDP traffic??? | | (I didn't think it was possible, and last I heard, NFS over TCP was | still very experimental) Interesting ... from my understanding UDP is a connectionless protocol, and as such packets aren't guaranteed to arrive at the destination. It seems to me that for a file system, you really want all the packets to arrive. How is this not a problem? (BTW, I'm just beginning in networking programming and have no NFS experience yet, but don't be afraid to give gory programming details :-)) -D
Re: GUI for GCC
On Mon, Jul 16, 2001 at 12:56:34PM +0200, Ellenkamp, Guus wrote: | Hello Matthias, | | Thanks for the reply. | | Might be I'm looking for an ADE. I used to work with an Atari-ST, which had | a nice integrated editor/compiler/linker/debugger (edit sources, compile, | jump from errors to source, graphical debugger...) | | Which one do you recommend? I recommend using (g)vim, bash, make, gcc, gdb. This is my preferred setup. Actually, I don't really use gdb directly, but I used DDD as a GUI front-end to it. I haven't done any C/C++ projects in a while so I haven't actually used anything, but I will probably use GVD (GNU Visual Debugger) for the next one because it has a nice GTK+ interface (DDD uses Motif which I don't really like, other than that it is a nice tool). One major difference between *nix systems and others is that the entire OS is your IDE. You have great tools like diff, grep, find, etc, available to work with the source tree. There are tons of different editors available, gvim happens to be my favorite (includes systax highlighting, auto-indenting, and auto jumping to errors if you run make/gcc from within it). I then run my programs from bash, rather than using a run button in some other IDE. Try searching on gnome.org or kde.org or on Google for Linux IDEs and you'll see that there are a large variety, though most only support certain languages and are relatively new. Try several of them out and see what combination you like best. -D
Re: XWindows library
On Mon, Jul 16, 2001 at 02:27:40PM +0200, Ellenkamp, Guus wrote: | I don't have the information here, but I'm just compiling a very small | program which only opens a window, so only the 'basic' library is missing. | Header files seem ok, it just doesn't link. It gives some errors like | symbols not found. Try again. This time when you get the error message, paste it verbatim into an email, along with the command (verbatim) that you executed, and send it to the list. Unless we see the exact spelling of the error message we are just shooting at the wind. Sometimes the smallest detail or difference in wording means a different problem which has a different solution. You can (almost) never provide too much information/details. -D
Re: How to install sound on Debian Linux
On Tue, Jul 17, 2001 at 01:53:36AM +1000, Sam Varghese wrote: | On Mon, Jul 16, 2001 at 11:44:06AM -0400, D-Man wrote: | | You need to edit a file in /etc/modutils (I think | /etc/modutils/local_config is a good name) and include some lines like | | The file is /etc/modules.conf Not on Debian. What happens is every time you run update-modules the /etc/modules.conf file gets overwritten. Any changes you make are lost. The way update-modules works is it concatenates all the files found under /etc/modutils together and writes them to /etc/modules.conf (with some comments, etc). Pick any file you want (even create your own) under /etc/modutils and then run update-modules and it will work fine. You are right, however, that if you don't run update-modules you won't see any change (updates). Also, if you (and every package, etc, on your system) never runs update-modules then you can edit /etc/modules.conf directly. -D
[OT] legal restrictions on email
Disclaimer : Those without a sense of humor should not read this message. It may be dangerous to your health. | Any information transmitted by means of this email (and any | of its attachments) is intended exclusively for the addressee | or addressees and for those authorized by the addressee | or addressees to read this message. Any use by a party Ok, so debian-user was the addressee. debian-user is a public entity that consists of all the world that wishes to participate. | other than the addressee or addressees is prohibited. | The information contained in this email (or any of its | attachments) may be confidential in nature and fall under a | pledge of secrecy and the attorney-client privilege. Hmm, does this mean that the whole world has attorney-client privilege with the sender of this message? If the whole world has attorney-client privilege, then who are we keeping these attorney-client secrets from? Maybe Marvin the Martian is trying to get sound working on Linux and that would cause the demise of Earth as we know it wink. [ no offense Guus, I'm just picking on the legal department ] -D
Re: NFS alternative
On Mon, Jul 16, 2001 at 11:44:54AM -0500, Dimitri Maziuk wrote: | * D-Man ([EMAIL PROTECTED]) spake thusly: | ... | Interesting ... from my understanding UDP is a connectionless | protocol, and as such packets aren't guaranteed to arrive at the | destination. It seems to me that for a file system, you really want | all the packets to arrive. How is this not a problem? (BTW, I'm just | beginning in networking programming and have no NFS experience yet, | but don't be afraid to give gory programming details :-)) | | IIRC the idea was that you don't want to use NFS over slow and unreliable | internet links anyway, you'll only use it on the LAN. Packet loss isn't | much of an issue there; OTGH, performance penalty caused by TCP overhead | is, esp. if you're trying to e.g. swap to an NFS drive. | | It's a reasonable trade-off, given that there were no gigahertz CPUs or | gigabit ethernets. With modern hardware it's the other way around: we'd | rather have reliable transport and a [usually negligible] TCP overhead. Ok, that makes sense. How about if probability leaves us behind and a packet is lost? Does NFS provide any way to correct for that or will your filesystem be hosed? -D
Re: mkfs.vfat? where is it ??
| I was unable to create an *windows* filesistem... | | mkfs -t vfat /dev/myharddrivepartition.. | | mkfs.vfat: no such file or directory I was also wondering since I don't think it was on my system. I just checked and it is in the 'dosfstools' package. We should probably try installing that ;-). -D
Re: NFS alternative
On Mon, Jul 16, 2001 at 12:45:22PM -0500, Dimitri Maziuk wrote: | * D-Man ([EMAIL PROTECTED]) spake thusly: | ... | Ok, that makes sense. How about if probability leaves us behind | and a packet is lost? Does NFS provide any way to correct for | that or will your filesystem be hosed? | | Thankfully, I forget the details[0]. From experience, no, it won't | be exactly hosed: you'll end up with a .nfs004950384672385721380937 | file that will grow and eventually fill up the partition... nothing | an rm -rf / won't fix. And then there's negative cookies and stale | mounts that require a reboot on most unices I've seen... Ok. It sounds like it would still result in data loss :-(. | you had to remind me, didn't you? Yes. Sorry. ;-) | [0] it's built on top of RPC, so whatever error-correction mechanism | RPC uses should apply. Plus, NFS does its own caching and stuff... | IIRC. Ok, I think I have a better understanding now. I'm going to be building a couple of diskless xterms soon so I'll get a bunch of experience then. Now suppose just the right packets are lost and the RPC call ends up matching a different, existant, procedure that doesn't have the intended effect grin ... sounds like it would be a good idea to make NFS over TCP stable :-). Can I use NFS-root-over-TCP for one of the boxes (I'll have 2, the other I'll leave at regular UDP as a control system)? -D
Re: man command made easy?
[I haven't been following most of this thread, but] On Mon, Jul 16, 2001 at 10:59:23AM -0700, Paul Mackinney wrote: | Night before last I ran 'find /usr/doc -name index.html foo.txt' and | then spent 2 hours adding links from foo.txt to my local home/start 2 hours!? Wow. Learn vi(m) or some other advanced editor. If you have that foo.txt with each filename on its own line the following will take very little time to create a web page of links out of it. :%s,^(.*)$,\1\1/abr, :%s,^,a href=file:///usr/doc/, Then insert HTMLBody at the top and /body/html at the bottom. Those vi commands say (1) replace each line with itself followed by '' followed by itself followed by '/abr' It will turn texmf/latex/index.html into texmf/latex/index.htmltexmf/latex/index.html/abr (2) insert the link at the beginning of the line texmf/latex/index.htmltexmf/latex/index.html/abr becomes a href=file:///usr/doc/texmf/latex/index.htmltexmf/latex/index.html/abr Becoming familiar with commands like that (and regexes in general) can really be a time saver. | page. I found answers to 3 random trivial questions that have been | bugging me; I'll never regret spending the time on this. Yeah, it is always good to find info you are interested in in the documentation. It really helps (like learning commands like the above). HTH, -D
Re: sound card not working properly
On Mon, Jul 16, 2001 at 08:36:22PM +0200, Philipp wrote: | Hi there, | | i have a soundblaster compatible sound card which is pluged on an isa-slot. | | I installed the soundcard like this: | | modprobe sb irq=5 dma=0 io=220 | | When i look in the /var/log/syslog i see that a ESS1868 chip is detected. Oh, ok, that really is SB compatible. (I used to have an ESS1869 card). When I read the sound-howto I learned that most sb-compatible cards aren't sb-compatible | That's my soundcard. | After doing this command: | | cat /usr/share/sounds/phone.wav /dev/dsp | | i get some anoying noises, like crscrsrssrsrsrsrsrsrs but no clear sound. | | What's wrong? Could be wrong DMA channel. | Ok, when i am under my X-session and typing sed in the command line, i | can use for e.g. xmms and a clear sound comes out of my speaker, but i | can't adjust the volume. huh. (email doesn't display the tone, this is a thoughtful huh, not a confused huh) | What's wrong and what can i do to get my soundcard working properly? | Hope someone can help. With my sound card I originally had the wrong DMA channel configured. When I played a sound (with, eg, xmms) it was horrible. It sounded like a record player that was jerking forwards and backwards (the same portion of the stream would be played in slow-speed several times before it got to the next portion). Also, when I tried to stop in xmms it would take several seconds for the noise to actually stop. One time I decided it was time to solve this problem so I booted windows and recorded every single option and setting in its config. Then I went back to linux and found that I was using the wrong DMA channel. I corrected it and the sound was beautiful after that. That's about the extent of my sound card experience, but make sure you aren't using an already used DMA channel and IRQ. -D
diskless terms and NFS/alternatives (was Re: NFS alternative)
On Mon, Jul 16, 2001 at 01:43:34PM -0500, Dimitri Maziuk wrote: | * D-Man ([EMAIL PROTECTED]) spake thusly: | ... | | Now suppose just the right packets are lost and the RPC call ends up | matching a different, existant, procedure that doesn't have the | intended effect grin ... sounds like it would be a good idea to | make NFS over TCP stable :-). | | Well, RPC has its own error correction so if a right packet is | lost it will be re-transmitted. Or RPC call will time out and | return error (d'oh! remote RPC call. Automated ATM machine). | The difference is that this is handled by application-level (if | you consider RPC to be in application layer) code, not by transport | layer. | | I imagine implementing NFS over TCP would involve a re-design of | RPC state machine and a serious re-write of all related code, and | it ain't exactly broken as it is, so... (given all the things that | could [theoretically] go wrong with NFS, it is surprisingly stable). Hmm, yeah, I guess that could be hard, unless the RPC mechanism could use TCP. Or maybe a different RPC implementation could be used that would work over TCP. Or maybe it isn't really a problem in practice but just in theory. | Can I use NFS-root-over-TCP for one of the boxes (I'll have 2, the | other I'll leave at regular UDP as a control system)? | | There are other networked file systems out there, like Coda, more modern | and arguably better than NFS. If you only need to support linux, why not | use one of them? Or [e]nbd? They're both going to be Debian. Can I use root-over-$FS for those filesystems? Are they stable? -D
Re: boot disks (again)
On Mon, Jul 16, 2001 at 03:09:20PM -0400, [EMAIL PROTECTED] wrote: | thanks to everyone who helped with my mformat problem, but i am | still unable to get my boot disks working so i though i'd write | another e-mail in a little bit greater detail IMO it is much easier to make a boot disk with grub. Simply download ftp://alpha.gnu.org/gnu/grub/grub-0.5.96.1-i386-pc.ext2fs and dump it to a floppy (doesn't matter if it is formatted or not) with dd. Then boot with it. Read a the sample menu.lst file that is on the floppy and tweak it to match your system. You can even copy the kernel (use cp) to the floppy and tell grub to boot with that one. Grub's config file is mostly self-explanatory. If you have more grub questions, don't hesitate to ask. Looking at the ftp site it looks like there is a new release labeled ftp://alpha.gnu.org/gnu/grub/grub-0.90-i386-pc.ext2fs, but I am using the older (0.5.96.1) version and have been very happy with it. HTH, -D
Re: zip drive?
On Mon, Jul 16, 2001 at 03:24:06PM -0400, Peter S Galbraith wrote: | Jeld The Dark Elf wrote: | | There is a wierd problem with ZIP disks. The ones formatted for | windows have their main partition at /dev/hdd4 while the ones | formatted with ext2 are /dev/hdd1. Don't know why. | | I've never seen any formatted with ext2 unless I had done it | myself, in which case I did whatever I wanted. Ditto. | /usr/share/doc/HOWTO/en-txt/mini/ZIP-Drive.txt.gz advised people | to do this in /etc/fstab: | | /dev/sda1 /zip ext2 noauto,rw,user,nosuid,sync | /dev/sda4 /zipdos vfat noauto,rw,user,nosuid,sync,mode=0777 This works. Alternatively you could make sure you use the same partition for both the (pre-made) vfat and ext2 disks and use auto for the fs type. Then mount will figure out if it is ext2 or vfat for you and you can use the same mount point. -D
Re: NFS alternative
On Mon, Jul 16, 2001 at 12:13:10PM -0700, Francois Gouget wrote: | On Mon, 16 Jul 2001, D-Man wrote: | On Mon, Jul 16, 2001 at 12:45:22PM -0500, Dimitri Maziuk wrote: | | * D-Man ([EMAIL PROTECTED]) spake thusly: | | ... | | Ok, that makes sense. How about if probability leaves us behind | | and a packet is lost? Does NFS provide any way to correct for | | that or will your filesystem be hosed? | | | | Thankfully, I forget the details[0]. From experience, no, it won't | | be exactly hosed: you'll end up with a .nfs004950384672385721380937 | | file that will grow and eventually fill up the partition... nothing | | an rm -rf / won't fix. And then there's negative cookies and stale | | mounts that require a reboot on most unices I've seen... | | Ok. It sounds like it would still result in data loss :-(. | |Hmm, I'm not an NFS expert but I'll play one on the mailing-list for | you ;-) Please, if there are experts out there, correct me if I'm wrong. grin Everyone's an expert on the internet :-). |AFAIU, NFS has its own mechanism to recover from lost packets. So it | won't be a problem if a packet is lost. Similarly I believe NFS RPCs That's nice to know. | cannot span UDP packets, so there is no chance that a lost packet would | change the meaning of an RPC. The RPC will be lost, pure and simple, and | NFS will have to reissue it or something similar. So I don't think | packet loss is an issue. Good. |What NFS is 'lacking' is congestion control, as in the TCP slow star | and exponential back-off. This means NFS will blast UDP packets as fast | as it cans with no regard for other trafic. This is not really an issue | on a lan and actually had a performance at a time (I think). But if you | go over multiple links, then you may saturate a slower link, causing the | router that is just before it to start dropping packets. Especially if | multiple streams converge there. And once you start dropping packets | performance degrades very significantly. I believe that's why NFS is bad | if there are multiple hops (I get it from a very reliable source that | this is also why it's very bad if the traffic will go over ATM, you need | buffering/traffic shaping). Ok, yeah. I'm not worried about congestion on my home LAN, I was just curios about (theoretical) reliability knowing it used UDP. Thanks for that info! -D
Re: VPN for linux?
On Mon, Jul 16, 2001 at 03:43:56PM -0400, Walter Tautz wrote: Dunno about VPN. | Basically the problem is to able to setup a lab where people with | laptops can login into the VPN with the appropriate authentication | and be able to do `work'. A wireless capability would be bonus but | not absolute necessary. What kind of server will the people be logging in to? What kind of connection will they have? With *nix systems I think ssh and X forwarding over ssh should be all you really need. (ssh uses encryption so you don't need to have a private connection) -D
Re: sl anyone? was [Re: ./ in PATH, always bad?]
On Mon, Jul 16, 2001 at 08:36:09PM -0600, Jimmy Richards wrote: | Hi, | | I found that I type 'la' by mistake more often than 'sl', so I made | symlink for sl to la. :-) I Love that train coming down the track! Ok, Aliases are better than symlinks because it doesn't muck up your filesystem. (see man bash) 'la' is more commonly aliased to 'ls -a', but personally I don't use that. -D
Re: man command made easy?
On Mon, Jul 16, 2001 at 03:41:01PM -0700, Paul Mackinney wrote: | D-Man uttered: | | [I haven't been following most of this thread, but] | | On Mon, Jul 16, 2001 at 10:59:23AM -0700, Paul Mackinney wrote: | | Night before last I ran 'find /usr/doc -name index.html foo.txt' and | | then spent 2 hours adding links from foo.txt to my local home/start | | 2 hours!? Wow. Learn vi(m) or some other advanced editor. If you | have that foo.txt with each filename on its own line the following | will take very little time to create a web page of links out of it. | | Vim is indeed my editor of choice. The 2 hours was mostly because I got | distracted reading checking out the stuff I found. But my edits Oh, ok. That's better. Two hours reading docs, not two hours creating HTML links. (BTW, thanks guys for those other generation techniques) | weren't nearly as slick as what you suggest. | | Thanks to a conversation at a BAD (Bay Area Debian) meeting last week, I | was clued in to visual mode. Here's what I did: | 1. At the first line, type 'v' to enter visual mode. | 2. Scroll to the last line, type ':' | 3. Vim starts the command text for you. Append | s/^/a href=file:\/\//g enter | | 4. Repeat the command, this time appending | s/$/\ DESCR \/a/g | | This turned a line like | /usr/share/doc/apache/manual/mod/index.html | | into | a href=file:///usr/share/doc/apache/manual/mod/index.html DESCR /a Yes, visual mode is great when you want to operate on a subset of the buffer. The '%' at the beginning of my command means use the whole buffer. | But your technique for capturing the original line using it for the | link description is very cool will save me much effort for my updates. Thanks. Actually, I'm not 100% sure that vim supports backreferences like that (the \1 in the replacement text). I know that Python's and Perl's regex engines do. I think I've heard/read that vim does do backreferences. BTW, I probably spent more time making my sample commands a little shorter than you spent with your slightly simpler version. Also, the simpler the command is the more likely it is correct ;-). -D
Re: sound card not working properly
On Tue, Jul 17, 2001 at 12:28:21AM +, Robin Gerard wrote: | okay, windoz give me : sound card: sb | irq :05 | dma :01 | io : 220 | linux give me :sound card sb | irq: 07 (and irq conflict) | dma: 01 | io: 220 | | and modprobe sb irq=05 dma=01 io=220 ^^^ I think you missed my mention of this in the last post. modprobe is probably reading this as decimal (base 10) but it should be hexadecimal (base 16). Change it to io=0x220 and see if it helps at all. IIRC, but I'm not completely sure, 'cat /proc/pci' will give info about PCI cards. It may help some. | gives no change for irq | what is the good way to change realy my irq ? Not sure. I thought the modprobe command would let you set it. -D
Re: Installing Debian Linux
On Sun, Jul 15, 2001 at 04:58:23PM -0400, [EMAIL PROTECTED] wrote: | Hi! | | My name is Jim. Windows is a pain in the neck so i decided to use linux ;o) Great! Welcome to the community. | But i have a problem with the install. I downloaded the three mirror files | (.iso) and made myself the debian cds. | | When i start to install Linux, everything is fine until i get to this spot: | | Install the operating System kernel and the Device Drivers | | I choose to install from the 1st cd, but instead of asking me what kind of | interface type i want to use (master, slave, ...), the computer checks the cd | and then suddenly it wants to install debian from the rescue disk. But even | that is not working, because after ca 10 sec of loading i get the message, | that it was not possible to load the rescue disk (i downloaded the mirror and | made one). Hmm. If you are installing from CDs then you shouldn't need a rescue disk. Depending on which packages you want (and wether you want -dev and older stuff too) you may only need 1 cd, or all 3. I assume by rescue disk you are referring to the floppy image (ie rescue.bin that is about 1.4 MB). | Did i do something wrong by creating the cds or the disk? When you put the CDs in the drive, can Windows read them and tell you what files are on them? If so the cds are fine. How about with the rescue disk? If Windows can read the rescue disk, then it is messed up. The image, rescue.bin, is an exact binary duplicate of the bits that should be on the floppy disk. The floppy needs to be created with a program, such as 'dd', that will dump the raw bits to the disk. You will then have a compressed ext2 filesystem on the disk, which windows can't read. HTH, -D
Re: The format (or lack thereof) of the list...
On Sat, Jul 14, 2001 at 03:56:17PM +0200, Carel Fellinger wrote:
| On Fri, Jul 13, 2001 at 09:22:44PM -0500, will trillich wrote:
| On Thu, Jul 12, 2001 at 06:01:39PM -0400, D-Man wrote:
| ...
| folder-hook . set sort=date ; set index_format=%4C %Z%{%b%d}
%-15.15F(%4l) %s
|
| I think this should read:
|
| folder-hook . 'set sort=date ; set index_format=%4C %Z%{%b%d}
%-15.15F(%4l) \%s'
|
| folder hook expects one command, you need a compound command, so group
| them with single quotes.
I should try that. Maybe that line isn't really setting index_format
at all so I don't see the errors Will is seeing. It could be that my
later folder-hooks are masking that.
| when i try setting index_format in ~/.muttrc i get
| %Z: unknown variable
| so i tried %S instead of %Z:
| %S: unknown variable
| i ditched %Z and %S altogether, and got
| %{%b: unknown variable
| aaugh!
|
| And what is it that you are trying? Will, you should know to get good
| response here, you need to include as much info as you can, the error
| codes, but also exactly what you are doing:]
Yeah, I found that some parts of the string were a bit picky wrt
spelling (like, extra spaces and stuff). Maybe your line is just
slightly different than mine. I think we are using the same version.
Anyways, the manual is available online at www.mutt.org, and can be
downloaded (IIRC) for off-line use too.
-D
Re: Testing upgrade: Potentially harmful(?) problem
On Fri, Jul 13, 2001 at 09:56:03PM -0400, Sebastian Canagaratna wrote: | | Hi (Jim and D): | | Some further notes. The dist-upgrade seems to get stuck trying | to install libc6-dev2.2.3-5 ( it seems to think that | this depends on libc62.2.1 ) but this is not available in testing: I did have some trouble with libc6 and libstdc++. I would try removing libc6-dev (since you can easily install it later) and trying again. If you get things like can't upgrade libc6 because you already have libc6 (some similarly ridiculous message, it is ridiculous because it is a computer and not a person) add --force to your apt command. Once libc6 and libstdc++ are upgraded, kill apt (^C) and start again without the --force. It worked for me ;-). -D
Re: D-Link DFE-530TX+
On Sat, Jul 14, 2001 at 08:39:27AM -0500, Donald R. Spoon wrote: | Evan Flynn wrote: | | Has anyone had any luck doing a network install of debian with a D-Link | DFE-530TX+? What module can I use and if it's not included with the driver | set where can I download it? | | I have about 5 of them here, all working just fine in various flavors of | Debian (Debian Potato, Progeny, Corel Linux, and Stormix). Use the | rtl8138.0 module...the stock one that comes with all 2.2.XX kernels ^ should be a 9 | has worked great for me. my friend has this card and it works great for him (Potato and stock 2.2.x kernel) | Dunno about 2.4 series kernels. Note: That is '8139too' I don't know why they changed it or why they put too in the name, but that's what it is. -D
Re: XFree86 4.0.3 configuration (and autoviewing HTML mail in mutt)
On Sat, Jul 14, 2001 at 04:59:06AM +, florentin ionescu wrote: | | Sorry for the inconvenience but I am used Hotmail web-mail because can't use Yeah, it is no fun when your preferred MUA can't be used. See below for a solution someone gave me so that HTML mail isn't too bothersome anymore (except when quoting is totally hosed and I can't tell who said what). | pine, have no X server running! I attached the original question please Actually, pine is a console MUA (last time I checked) so you don't need X to use it (FYI). -- Bek: I see that you are using mutt. Someone else gave me this clue, and it works great! ~/.mailcap text/plain; less %s; nametemplate=%s.text; copiousoutput text/html; lynx -dump %s; nametemplate=%s.html; copiousoutput ~/.mime.types text/plain text txt text/html html htm ~/.muttrc # some people insist on using HTML mail # this will automatically convert HTML to Plain Test (via settings in # ~/.mailcap and ~/.mime.types) auto_view text/html # This tells mutt which MIME type to preferr to show, if both are in a # message alternative_order text/plain text/html I am running this on my school's Solaris 8 system where my mail is received. They only have 'lynx', not 'links', but the latter will probably give better looking output (especially if tables, etc, are used because lynx doesn't support tables). -- Florentin: I recommend backing up your XF86Config file from 3.3.6 first (hopefully it's not too late for you). Then use your favorite text editor to tweak the version 4 config file to include the options you had with the old version. This worked quite well for me (diff. card though). HTH, -D
Re: Installing Java
On Sat, Jul 14, 2001 at 09:04:45AM -0700, Paul Mackinney wrote: | D-Man uttered: | On Fri, Jul 13, 2001 at 04:41:40PM -0300, Juan wrote: | | HI, | | | | Which packages will I have to install to run compile Java? And run Java | | Server Pages? | | kaffe might be enough. There are different versions and vendors of | Java interpreters so it depends on which one you want. Most aren't | DFSG free so they aren't really in Debian though you can download and | use them just fine from, eg Blackdown or IBM. | | -D | | It's java, why not get it from Sun? http://java.sun.com/j2se/ It's not Free. | Not quite sure what you mean by run Java Server Pages, but I've He's talking about Java Server Pages (aka JSP) wink : http://java.sun.com/products/jsp/ It's a server-side technology as opposed to a client-side (applets). -D
Re: D-Link DFE-530TX+
On Sun, Jul 15, 2001 at 12:04:24AM +0100, Paul Tansom wrote: | On Sat, 14 Jul 2001, Rick Commo wrote: | The module is the via-rhine and it's been included in | kernel since 2.2 IIRC. | | Watch out here. I believe that the via-rhine driver works with an earlier | of version of this card. The correct driver for the DF350-TX+ is rtl8139.o. | The card uses the Realtek chip and in fact on my Win2K machine reported | itself as the Realtek card. The rtl8139 driver was in the r2.2 release CDs | that I bought some time ago. I have a few of them going here with no | problem. | | Correct, the DFE530TX uses the via-rhine driver, but the DFE530TX+ uses the | rtl8139 driver. That plus makes all the difference, and the naming being so | close, but the cards/chipsets being so different causes a lot of confusion! Yeah. Not to mention the DE-530TX -- IIRC it is a tulip card. | for various reasons and break compatibility (notable the Linksys Etherfast | (tulip) and Netgear FA311 (nat-semi), which I have known to have issues. I have a LinkSys LNE100TX (rev 2, its a couple years old) that I have never had any trouble with. I use the tulip driver (since version 0.89 I think, 2.0.36 kernel). I has a Lite-On tulip clone. -D
Re: The format (or lack thereof) of the list...
On Fri, Jul 13, 2001 at 09:16:50AM -0500, Dave Sherohman wrote: | On Thu, Jul 12, 2001 at 06:01:39PM -0400, D-Man wrote: | On Thu, Jul 12, 2001 at 03:42:33PM -0500, Dave Sherohman wrote: | | Or, better, just tell it that debian-user is a list. Maybe I'm just | | doing something wrong, but whenever I add a subscribe list to my | | .muttrc, the to/from column shows To listname for all list | | messages. If I use lists list, it does the Right Thing and shows | | the sender's name. | | | | Of course, if anyone could tell me what I've done wrong... | | What you did wrong was leave mutt with the default settings ;-). I | suppose it's useful if you don't have procmail to tell you already | which list the message came from. Here's what I do (lots of | irrelevant options snipped): | | You missed a few. Unless mutt is far stranger than I think, handling | of reply-to on lists, forward formats, and sort orders have nothing | at all to do with whether sender or recipient addresses are | displayed. The index format seems to be the only thing actually | relevant. You're right, the index format was the (one) thing you were looking for, but I included the rest of those options because I find them useful and, somewhat, related (and someone may not have been aware of those capabilities). | Thanks, though, for pointing me at the right part of TFM to R! You're welcome. Actually, now that I think of it, someone posted an index format a while back that shows either the length of the message _or_ the number of messages in the thread if the thread is collapsed. That would use less screen real estate than my (current) format that shows the length of the first message _and_ the thread length if it is collapsed. I've been meaning to copy it so I can have longer subject lines, but haven't gotten around to it yet. -D
Re: RAM size.
On Fri, Jul 13, 2001 at 09:32:54AM -0500, Dave Sherohman wrote: | On Fri, Jul 13, 2001 at 12:13:20PM +0100, J.A.Serralheiro wrote: | On Fri, 13 Jul 2001, Alexey wrote: | You know, while running DOS or Windows, the CPU is hot (I can touch it), | even if I do nothing. It becomes cool under Linux!!! | | strange, never heard of that. | | Linux (and NT, incidentally) sends HLT (HaLT) instructions to the CPU | telling it to shut itself down (until the next interrupt) when there's | nothing for it to do. So if your linux system tells you you're at | 30% CPU utilization, the CPU is essentially turned off 70% of the | time. This would also (theoretically) lead to less power consumption and a lower electric bill. Pretty nice! Say, does that HLT instruction work on a i486 or only on newer CPUs? I also seem to recall, back when I was learning m68k assembly, that the halt instruction on there shouldn't be used if you want to ever do any procesing again (without a reboot). -D
Re: fetchmail, mutt, mail
On Fri, Jul 13, 2001 at 04:58:59PM +0200, Schoppitsch Dieter wrote: | Hi, | what is going wrong? | | Dialing in to my ISP (with pppd) and fetching mails (with fetchmail) | works well. But then I get the message SMPT connect to localhost | failed Do you have an MDA (or MTA) installed on your system? fetchmail needs to hand off the messages to something else, such as procmail, sendmail, exim, postfix, etc. | respectively (using fetchmail -S hostname) This message is | MIME format. Since your mail reader does not understand. Where does this come from? mutt can handle MIME just fine. | * Where can I see/set my local smtphostname (or should I set my domainname)? Probably in the config of your MTA. | * Sending myself a mail (with mutt) doesn't work - which address | should I use (root or [EMAIL PROTECTED] doesn't work)? Mutt needs an MTA to send (transfer) the message somewhere. Either setup an MTA/MDA on your system if you want to deliver locally, or set up ssmtp to accept the message and hand it off to a real SMTP server. HTH, -D
Re: RAM size. [OT]
On Fri, Jul 13, 2001 at 11:52:35AM -0500, Dimitri Maziuk wrote: | * Lamer ([EMAIL PROTECTED]) spake thusly: | This would also (theoretically) lead to less power consumption and a | lower electric bill. Pretty nice! Say, does that HLT instruction | work on a i486 or only on newer CPUs? I also seem to recall, back | when I was learning m68k assembly, that the halt instruction on there | shouldn't be used if you want to ever do any procesing again (without | a reboot). | | however, i hate 68k assembly so much because i don't know how can i use a | complete text-mode on an apple.. | | ... which is precisely why decent schools use amigas to teach m68k assembly. Maybe I don't go to a decent school wink but we had custom m68k boards. They weren't apples and they weren't amigas. They were hooked up to win9x boxes through the serial port and we used hyper terminal to send the cross-assembled code to the board. They had a tutor prompt which was some sort of program that handled the transfer of code to the board and also debugging stuff like stepping through instructions and displaying the register contents. BTW, most of those boards died before the end of the quarter. On a few occaisions I saw the PC (that's program counter, not personal computer) skip a byte or two which would result in a address trap. -D
Re: Installing Java
On Fri, Jul 13, 2001 at 04:41:40PM -0300, Juan wrote: | HI, | | Which packages will I have to install to run compile Java? And run Java | Server Pages? kaffe might be enough. There are different versions and vendors of Java interpreters so it depends on which one you want. Most aren't DFSG free so they aren't really in Debian though you can download and use them just fine from, eg Blackdown or IBM. -D
Re: SID out of Date ??
On Fri, Jul 13, 2001 at 03:53:45PM -0400, Case, Benjamin wrote: | I am currently d/l all the SID Cd images from | ftp://ftp.uni-bremen.de/pub/mirrors/debian_cdimages/debian-unofficial/sid/ | . | I noticed that they are all dated 3/10/2001. How far behind will these | discs be ? They'll be as out-of-date as the date on them. Sid changes constantly as developers upload new packages. Any CD is, therefore, out-of-date almost immediately after its creation. If you want CDs that aren't out-of-date get Potato discs. Of course, much of the software in Potato is out-of-date because development has continued. :-) -D
Re: SID out of Date ??
On Fri, Jul 13, 2001 at 04:46:58PM -0400, Case, Benjamin wrote: | That makes complete sence. Is 3/10 ancient times as far as SID is March is ancient in OSS terms. | concerned ? I just want to work with the 2.4 kernel, some Xfree 4 | stuff, and I know that there are SID level dependancies that I will Nope, I have woody on my box and X4 is really nice. I downloaded (via apt-get) kernel 2.4 but I haven't installed it yet since I know it will either work or fsck my modutils (which would make 2.2 unusable). | need to meet for those. This is just temporary until my DSL | comesso I am d/l all of SID to have for the next week or so. | Does that sound like a good apporach, or am I goin overboard. I just | dont want to find myself in search of Packages without any internet | to get them with. I'd recommend getting Potato CDs to install from. Then download the packages you want, and any dependencies from woody to put on removeable media (CD-R works fine, or zip disk, or move a hard drive, whatever). packages.debian.org will list all the packages and the dependencies each package has. If you have a dial-up connection that you don't pay per-[minute|byte|whatever] you can use it to download stuff overnight. (I only have dial-up at the moment so that's what I do). Point sources.lst to woody and apt-get update apt-get --download-only dist-upgrade YMMV on the dist-upgrade though. Several weeks ago I had no show-stopper problems. Getting every package is going overboard since, presumably, you don't use every package. Just get the ones you need. -D
Re: ./ in PATH, always bad?
On Fri, Jul 13, 2001 at 04:57:14PM -0400, Daniel Patrick Berdine wrote: | The Redhat machine I use at work seems to include ./ in the PATH | variable, I can always run executables from my current directory | without using ./ like on my home debian system. This has always seemed | more convenient to me and I wondered why Debian doesn't do this until I It's up to you wether you want it or not. | read that it is considered a security flaw. Is this always so? Is | there a way to enable this without compromising security? Sure it's a flaw : suppose someone creates an executable trojan in the current directory named 'cd'. If '.' is the first thing in the path you will execute the trojan rather than the usual /bin/cd. A similar thing happened at school one time, but it wasn't a security issue. The professors had a lab which included a program called 'test'. We (the students) were supposed to run it with certain arguments and observe the results (or time it or something like that). It turned out that csh people had to problems. bash users, however, kept getting weird results and error messages that had nothing to do with the lab. Eventually someone realized that 'test' is a builtin in bash so when people ran test they were not running the sample program the staff had provided. Using csh or, after the cause was known, using './test' in bash would solve the problem. If you want to have '.' in the PATH, go for it. Just be aware of what is in '.' that is executable. -D
Re: [OT] job title
On Sat, Jul 14, 2001 at 12:19:01AM +0200, Martin F. Krafft wrote: | also sprach Joost Kooij (on Fri, 13 Jul 2001 11:51:14PM +0200): | Rechnerbeschwoerer So what's the literal translation? | cool beans! i'll write that. and if everything fails, you'll hire me, | right? Sure wink. -D
Re: Testing upgrade: Potentially harmful(?) problem
On Fri, Jul 13, 2001 at 06:49:43PM -0400, Sebastian Canagaratna wrote: | Hi: | | I am trying to upgrade testing on my second machine (apt-get | update; apt-get dist-upgrade) but it aborts. So I first tried to | install debconf ( apt-get install debconf ) and I get: | | Extra Packages to be installed: | libhtml-parser-perl libhtml-target-perl libmime-base64-perl | libperl5.6 perl perl-5.004 perl-5.004-suid perl5.6 | perl-base perl-modules. | | Removed: perl-5.004-base | perl-5.005 | perl-5.005-base | | New: libhtml-tagset-perl | libperl5.6 | perl | perl-modules | perl-suid | | Essential package to be removed: perl-5.004-base. You are about to do | something potentially harmful. Don't do this unless you know what you | are doing. | | At this stage I aborted. Anybody have any ideas whether its Ok | to say yes? The perl version changed. However, perl is a critical component of the Debian core (dpkg or apt or debconf or something like that). If you do a dist-upgrade then everything will get upgraded together and there won't be a problem because you replace one perl with another, and all the packages that depend on it are replaced with packages that depend on the new perl instead. I did a dist-upgrade a few weeks ago without any trouble. Maybe if you try again, as opposed to upgrading stuff one-at-a-time it will work better? Why did it abort in the first place? Did dpkg just die? If so, run it again (it happened to me a couple times). At the moment I have : $ dpkg -l perl\*base Desired=Unknown/Install/Remove/Purge/Hold | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: |uppercase=bad) ||/ Name VersionDescription +++-==-==- pn perl-5.004-bas none (no description available) ii perl-5.005-bas 6.2Transitional package. un perl-5.6-base none (no description available) ii perl-base 5.6.1-5The Pathologically Eclectic Rubbish Lister. un perl5-base none (no description available) -D
Re: sysadmin won't allow linux - PLEASE HELP
On Sat, Jul 14, 2001 at 12:34:59AM +0100, Stig Brautaset wrote: | Kent West [EMAIL PROTECTED] writes: | ray p wrote: | | Or better yet get putty. It supports SSH 2 with public key | authentication. And is small enough that you can put a key the | client and the scp and sftp (FTP tunnled through SSH) client on ~2 | floppies It can be run from the floppy and can connect to any SSH | server has a lot of very cool features. Use the devel snapshot to do | SSH 2 with public keys. | | I *think* you can just put it on a share somewhere, and then map a | drive to that share and run putty off of it. I reckon it depends on if | you have access to the floppy, to mapping drives, either, or neither. | | You can, but you are not (to my experience) able to save any of the | setup-changes. Yeah, I believe that profiles are stored in the registry. Blast the registry! If you can actually write to the registry, you can save the options but it would only be for that particular terminal. -D
Re: [OT] job title
On Sat, Jul 14, 2001 at 01:21:39AM +0200, Martin F. Krafft wrote: | also sprach D-Man (on Fri, 13 Jul 2001 07:06:05PM -0400): | | Rechnerbeschwoerer | | So what's the literal translation? | literally, this is the incantation or conjuration of computers... Haha! That's good (although I am not familiar with german culture to appreciate the connotations it may have). It makes it sound like you are a powerful magician or something ;-). -D
Re: linksys nc100 problem. modversions.h not found.
On Fri, Jul 13, 2001 at 04:27:11PM -0700, crombie wrote: | hi, | | attempting to install linksys nc100 ethernet card on I haven't heard of the nc100 yet, but ... | debian potato. when compiling using that long gcc | command, | | gcc -DMODULE -D__KERNEL__ -I/usr/src/linux/net/inet | -Wall -Wstrict -prototypes -O6 -c tulip.c ^ I am familiar with the tulip driver. For my LinkSys LNE100TX (rev 2) I simply put alias eth0 tulip in /etc/modutils/local_options and ran 'update-modules' and the card works fine. Maybe the tulip driver that ships with the kernel will work with your card too? -D
installing gcj, is this ok? (gcc-3.0)
I want to install gcj on my system. I am running woody, and when I try and apt-get install gcj it says it needs gcj-2.95. Ok, so I add that to the apt line. Now it says it will install cpp-3.0 gcc-3.0 gcc-3.0-base and libgcc300. Will there be any problem with letting apt install gcc-3.0 next to gcc-2.95? If not is there any reason not to install gcj-3.0 instead of 2.95? (I really want gcc 3.0, but I'm not ready to deal with converting my system yet. Also I need to build a kernel first.) -D
Re: OT Ram upgrade options
On Thu, Jul 12, 2001 at 08:52:35PM +0930, David Purton wrote: | questions: | | 1. I gather you can get PC66, PC100 and PC133 MHz ram. Does my board | limit my options to one of these speeds? If so, how can I tell which | speeds are supported? Pretty much all I know is that PC133 memory is backwards compatible with PC100 memory. That is, you can put PC133 memory into a board that wants PC100 and the board won't know the difference. I actually tried that on my old Compaq (bought off-shelf in '98, who knows when it was assembled) when I accidentally fried the memory that came with it and CompUSA only had the PC133 available (it was a sale -- buy 64MB for $30 and get a $30 rebate). -D
Re: [users] Re: sysadmin won't allow linux - PLEASE HELP
On Thu, Jul 12, 2001 at 09:15:23AM -0400, Hall Stevenson wrote: |They can either let you run linux, or pay for your |MCSE courses. | | That's reasonable ... as long as they've paid for | everyone else's MCSE course who uses Windows | on the campus. | | speaking from experience (oops), the mcse doesn't | buy you jack. okay, i didn't take courses, but the | certifications are totally stupid, so i don't figure you | learn a whole lot more in the training sessions. | | a person with an mcse knows how to click pretty | buttons and how to interpret pretty (?) dialogs, but | know jack about computers or networks... | | In all honesty, there's a difference between taking | MCSE-related classes and taking Introduction to Windows | 9x/NT/2000. Cost is a *huge* one. Yeah, the whole point of the request for the course is based on $$ -- it would be cheaper for the admins to let [you] (the OP) run linux than pay for an MCSE course. | An MCSE here at work had the task of fdisking my old Linux box | ;-( Another admin had to walk her through booting off a | floppy disk and starting fdisk. She was lost once it started | though. No buttons to click, no wizards to walk her through | the steps, no Next buttons... Hehe. -D
Re: what is up the list
On Thu, Jul 12, 2001 at 10:52:25AM -0400, Hall Stevenson wrote: | Is it useful to subscribe again to this mailing list? | | Since you're message made it to the list and was | posted, I believe you're subscribed. | | Posting to the Debian mailing lists is open to anybody: | it's not limited to subscribers. | | Thanks, I didn't know that... You just won't get any replies | unless you specify for people to cc or reply directly | though, huh ?? Right. Or you could use the Mail-Followup-To header (that mutt and gnus obey) to request a copy. This is what I like about open lists, if I don't want to subscribe and get a large volume of messages to sort through. Of course people using clients like OE, Netscape, etc, won't automatically reply to you because they (the mua) ignore the Mail-Followup-To header. -D
Re: GRUB Rieserfs
On Thu, Jul 12, 2001 at 11:52:03AM -0400, Chuck Stickelman wrote: | D-Man wrote: | What does the single argument to the kernel mean? (check Linux | docs, not grub). I am not familiar with it and that may be causing | the problem. | | the `single` argument on the kernel command line stops the boot process | before it goes into multi-user mode (you get the prompt to enter the root | password or type Ctrl-D to continue booting). Ok. | Just to start over, and since I didn't pay much attention to the | beginning of the thread, where does the booting fail and what message | do you get? Does grub manage to load the menu.lst file? Did you | include Resierfs support in the kernel, not as a module? | | | Ok. There's been a lot of suggestions and I've tried many things. | I can get the GRUB menu to come up - no problem. | I can get the system to boot my 2.4.5 kernel - no problem. | I can get the 2.4.5 kernel to mount ext2 FS as root - no problem. | Once I have an ext2 root I can mount a reiserfs partition - no problem. | I've double checked that reiserfs support is built into the kernel and not | as a module. | I can get the kernel to boot and mount the reiserfs as root but it generates | a kernel panic as it can't find `init` - BIG PROBLEM. | | Now, I know that there is an /sbin/init on the reiserfs. Here's how I've | checked: | I've mounted the FS and physically looked. | I've been in the GRUB command line and run `find /sbin/init` and it's returned | every drive that I'd expect it to return. (With some exceptions I'll go into | in a minute...) | | Questions: | Does GRUB support reiserfs? The answer seems to be Yes! | Does the kernel have reiserfs support built-in? Appears to also be yes. | Is GRUB installed correctly? I NOW NOT SO SURE. I do get a GRUB | menu. But... when I drop out to the GRUB command line I don't think | it's reading my device.map file... though the partitions I'm working with | are all on /dev/hda (which GRUB correctly sees as (hd0)). | Does GRUB load the kernel? Yes. I think, then, that grub is fine -- it has no trouble loading the menu and the kernel from disk. Once the kernel is loaded the machine is no longer under the control of the boot loader. | Does the kernel load the reiserfs FS as root? yes. | Does the kernel find `init` and continue booting? NO! I think there must be a kernel problem somewhere because everything up to the kernel loading is fine. How about adding init=/bin/sh to the kernel line in your grub config? This instructs the kernel to use /bin/sh (provides a root shell) instead of the usual SysV init. If this still fails, then try to get an ext2 disk somewhere (floppy, whatever) and put sh on it. Then try again using that ext2 disk for the init= argument, and maybe the root partition also. These are just to see that the kernel can, indeed, start init. Say, can you boot with a different disk, then mount that reiser disk and run /sbin/init on it? Are the perms ok on init (should that even matter during bootup?)? Maybe init is broken or something? -D
Re: harsher kill than kill -9 ?
On Thu, Jul 12, 2001 at 12:09:11PM -0400, Daniel Patrick Berdine wrote: | On Thu, 12 Jul 2001, Andrei Ivanov wrote: | | Well, if its a zombie process, it'll go away by itself after a while. | Can you send an output of ps aux for that process (or top), please? | Andrei | | not zombie... | | 19355 rothaar9 0 90588 88M 476 D 0.0 17.6 0:01 plot.out ^^^ IIRC (not at a linux box to check) that column is % of CPU being used. Since it isn't using any CPU it is only a problem for you as far as memory hogging goes. As Joost was indicating, the problem is the process is waiting (blocked) in the kernel; in some sort of system call (maybe sbrk or fork or something else that requires kernel services). The difficulty is that if the kernel just killed it, who would clean up after the kernel? Instead what current implementations do is hold any signals for the process until it returns to userland from inside the kernel. Then it is safe for the kernel to kill it because it isn't in the middle of the kernel anymore. So aside from rebooting, praying, and coming up with a new breakthrough in OS theory, you don't really have anything you can do about it. -D
Re: ssh strangeness
On Thu, Jul 12, 2001 at 01:57:48PM -0400, Andrew Dixon wrote: | Hi All, | I've been having some weird problems with ssh. I can connect to our | server (RH 7.1) but when I try to connect to another Debian box on the | network (PPC) I get the following message: | | [EMAIL PROTECTED]:~$ ssh -l andrew.dixon stiq | ssh_exchange_identification: Connection closed by remote host | | I get the same error when I try on a different PPC Deb box as well. I got this problem when I upgraded one of my boxes to woody -- the potato box and the win boxes (with putty) couldn't connect. I also couldn't ssh to localhost! I then started looking at how to generate keys, since with debug messages turned on it seemed that the new sshd wanted a key somewhere. After using ssh-keygen to create RSA and DSA and ssh1 keys in my home directory everything seems cool. I didn't even need to copy the keys over to the client boxes. (Now I need to do some research, but for now it works is good enough) | I CAN log into the PPC Deb box from the redhat server and I also CAN log | into my Debian box from the server. I didn't have a RH box to try with, but none of the boxes I have could ssh into my woody box for a while (see above). -D
Re: The format (or lack thereof) of the list...
On Thu, Jul 12, 2001 at 02:30:18PM -0400, User zos wrote: [ suggestion to put list name in subject ] I'm on some lists that do this and some that don't. I've found that when my filters work I like it better when the list doesn't because it keeps the subject shorter so it may actually fit on my screen (a lot of subjects are truncated do to limited terminal width). When a message wanders on over to my inbox instead, I find that knowing which list the message belongs to is nice, but that can be found in the X-Mailing-List: or the To: or Cc: headers so it's really not a big deal to me. This is just my opinion and experience with this, not the start of a flame war (I hope :-)). -D
Re: MUAs that compare with Outlook (your chance to show how much better Linux is than MS!!)
On Thu, Jul 12, 2001 at 01:35:42PM -0400, Brian Nelson wrote: | Why on earth would you want a single application to do all of that? Do This was my reaction also. Just for the record I used to be a full-time windows user, before I was introduced to Unix and Linux. I have tried a lot of different MUAs, both on Windows and Unix. I have found that I like mutt the best. It doesn't have some of the features that were listed as requirements, but it does a lot and does it well and does it efficiently (both CPU wise and UI wise). It sounds like the most important features are : o notification when important messages arrive o easy way to update schedule from messages relating to schedule type things The other requirements were more like GUI specifications to match Outlook. (if you disagree, read those requirements again and count how many _don't_ fall into the first item above) For the first, there are several different mail notification apps/applets that can be checked out. I have one in gnome (I think it is called MailCheck Applet) and I like it. It will run fetchmail periodically and it plays an animation and a sound when mail arrives. Of course, fetchmail directs the message through procmail so the notification only happens for one mailbox, my inbox, and list mail just goes to its respective folder. When I click on it it fires up whatever MUA I want it to (it used to be Balsa, then Mahogany, now mutt in a gnome-terminal). For the second feature, I think hand-updating gnome-cal (or whatever the calender in the gnome-pim package is called) when you read an interesting message that discusses your schedule. I sorta tried out evolution once, quite a while ago, and I didn't really like it even taking its at-that-time instability due to heavy development into account. What I didn't like about it was it tried to do everything, though it wouldn't make dinner or clean my room. I think quoting mutt's creator is well suited for this discussion: All mail clients suck. This one just sucks less. (but s/this one/your favorite/) -D
Re: Summary: MUA clients similar to Outlook
On Thu, Jul 12, 2001 at 12:14:56PM -0700, Kurt Lieber wrote: | Thanks to all who posted constructive feedback regarding my question | about MUAs. Here's a synopsis of the suggestions, along with URLs for | the various products: | | KDE2.2/Kmailer -- http://www.kde.org/ | Screenshot: http://www.kde.org/screenshots/large/kde2b3_2.png | | Looks promising. However, as I mentioned in my previous message, I'm | brand-new to Linux so I'm going to stick with Gnome for now until I get | more comfortable with general system administration. You can use Qt/KDE apps while using GNOME for the desktop -- you just need to install the proper libs (for example libqt and libkde). | GNUS -- http://www.gnus.org and http://my.gnus.org | Screenshot: http://my.gnus.org/screenshot.php?id=25 | | I believe this can run in a console (i.e. no X) which is definitely | appealing. yep. | Requires emacs, which I haven't learned yet (still working | on vi) so it may be a while before I give this one a shot. FYI mutt also runs in a console and allows you to pick your editor. You can use gedit if you want (a GNOME notepad-lookalike) or vim or emacs. | Someone else mentioned Balsa in passing. I used Balsa a couple of | months back and found the IMAP support to be rudimentary at best. I was | also annoyed at the number of different windows you had to open just to | open a single message. It's also designed (IIRC) purely as an email | client. Give it a few months to mature and I think it will be a solid | candidate for pure email. It is a great candidate for _pure_ email, except it doesn't have list support like mutt does (it also can't run without X). I seem to recall that I didn't like its behavior with IMAP, but I learned to setup fetchmail and that problem disappeared immediately (HINT: check out fetchmail :-)). Also I would view messages with a window like this : +-+ | | | | |-| | | | | | | | | | +-+ The left-hand column showed the folder list. The top box showed the message headers for the current folder and the lower box showed the message contents. I didn't have multiple windows. I think there is a config option for it. The other problem I had with it was I couldn't (easily) use vim as my editor. Has anyone mentioned Mahogany yet? You may want to look at it, http://mahogany.sourceforge.net. It even runs on windows too (BTW, gnus and mutt run on windows as well; mutt requires cygwin though, I think (that's how I'm using it anyways)). AFAIK mutt and gnus are the only mailers that do a good job of dealing with mailing lists (anyone have experience to the contrary?) with the Mail-Followup-To: header. They also have really good threading views (well, so I've heard about gnus, haven't tried it myself). Some people have mentioned that a few other MUAs have a threading view also. -D
Re: The format (or lack thereof) of the list...
On Thu, Jul 12, 2001 at 03:16:30PM -0400, Hall Stevenson wrote: | Mutt also has a G or g function for groups but I don't | use it. FYI, g stands for group-reply, which is called Reply To All in some other MUAs. G is used to retrieve mail from a POP3 server. -D
Re: The format (or lack thereof) of the list...
[ I missed this one the first time. ] | User zos wrote: | | Also since everyone here finds it in good taste to keep all replies | directed to the list (so we can all benefit) why not add a simple | reply-to: line? | | reply-to: debian-user@lists.debian.org Don't make this mistake. Instead read http://www.unicom.com/pw/reply-to-harmful.html and then read up on how to configure your mail system to operate nicely in the presence of mailing lists. (Please ask on the list if you have any questions or difficulty configuring your system. This is what we are here for!) -D
Re: installing gcj, is this ok? (gcc-3.0)
On Wed, Jul 11, 2001 at 10:13:04PM -0700, Eric G. Miller wrote: | On Wed, Jul 11, 2001 at 11:44:30PM -0400, D-Man wrote: | | I want to install gcj on my system. I am running woody, and when I | try and apt-get install gcj it says it needs gcj-2.95. Ok, so I add | that to the apt line. Now it says it will install cpp-3.0 gcc-3.0 | gcc-3.0-base and libgcc300. Will there be any problem with letting | apt install gcc-3.0 next to gcc-2.95? If not is there any reason not | to install gcj-3.0 instead of 2.95? (I really want gcc 3.0, but I'm | not ready to deal with converting my system yet. Also I need to build | a kernel first.) | | Try it and let us know ;) LOL! This isn't the answer I was looking for wink. | Presumably, the set-up still has gcc-2.95 as gcc and you'd have to | specify CC=gcc230 or some such to use the 3.0 version. I think I should suggest to the developers that the pages on packages.debian.org should all have a link to the upstream web site and display a list of all the files that are in the package. I would like to see, before I spend hours downloading the package, what executables it will create so I can look for any conflicts I may see. For example, I want to have kaffe, gcj, jdk1.1.8 and jdk1.2.2 on my system so I want to be sure there aren't any conflicts with more than one of them providing, for example, /usr/bin/java. It's probably best to use a different root for each package, and check this sort of stuff first. | Guess this pain will have to be dealt with pretty soon anyway... Yep. I read through the discussion on the -dev archives. No resolution yet. -D
Re: The format (or lack thereof) of the list...
On Thu, Jul 12, 2001 at 03:42:33PM -0500, Dave Sherohman wrote:
| On Thu, Jul 12, 2001 at 03:16:30PM -0400, Hall Stevenson wrote:
| If you use mutt to read your mail, tell it that you subscribe
| to debian-user mailing list.
|
| Or, better, just tell it that debian-user is a list. Maybe I'm just
| doing something wrong, but whenever I add a subscribe list to my
| .muttrc, the to/from column shows To listname for all list
| messages. If I use lists list, it does the Right Thing and shows
| the sender's name.
|
| Of course, if anyone could tell me what I've done wrong...
What you did wrong was leave mutt with the default settings ;-). I
suppose it's useful if you don't have procmail to tell you already
which list the message came from. Here's what I do (lots of
irrelevant options snipped):
--
# work-around broken mailing list software
set ignore_list_reply_to = yes
set forward_format = FW: %s
# default action (index_format is default, except it displays From instead)
folder-hook . set sort=date ; set index_format=%4C %Z%{%b%d} %-15.15F(%4l) %s
# sorting for inbox
folder-hook ! set sort=date-received
# display To, not From, in Sent folder
#folder-hook Sent set index_format=%4C %Z%{%b%d} %-15.15F(%4l) %s
# sorting for lists
# display the author's real name, not the list (or my name!) in index view
folder-hook lists.* set sort=threads ; set sort_aux=date ; \
set index_format=%4C %Z %{%b%d} %-15.15n(%4l,%3M) %s%|
--
The other difference bewteen lists and subscribe is that with
subscribe your address is left out of the Mail-Followup-To: header so
you don't get a double-copy. With lists mutt thinks you are not
subscribed to the list so it adds your address you get a double
copy, which is really the only copy you get because you aren't
subscribed.
| Great feature, isn't it?
Definitely :-).
-D
Re: sysadmin won't allow linux - PLEASE HELP
On Fri, Jul 13, 2001 at 01:22:39AM +0200, Martin F. Krafft wrote: | also sprach Michael A. Miller (on Thu, 12 Jul 2001 06:16:15PM -0500): | Can anyone suggest a way to scp from a machine where I'm not | allowed to install scp? (For example, win98 and win2000 machines | in our libraries) | | can't you install a cygwin version into your homedirectory??? I doubt there is a home directory, or even multiple accounts, on the machines he is referring to. Perhaps they have zip drives or cd drives. If so then you might be able to put the 'doze client on that and run it straight from the removable media. -D
Re: The format (or lack thereof) of the list...
On Fri, Jul 13, 2001 at 01:00:25AM +0200, Martin F. Krafft wrote:
| also sprach Hall Stevenson (on Thu, 12 Jul 2001 03:16:30PM -0400):
| If you use mutt to read your mail, tell it that you subscribe
| to debian-user mailing list. Then, when you want to reply to
| a message and have it go *only* to the list, hit the letter
| L (or is it lower-case l ??) instead of R or r.
|
| but that makes all messages appear as To debian user in mutt's
| index. they are already procmail'ed all into a single folder, so i
| want to see author names rather than destinations...
This is only with the default format string. See this snippet from my
.muttrc (and the earlier post in this thread with it) :
---
# Folder hooks :
# default action (index_format is default, except it displays From instead)
folder-hook . set sort=date ; set index_format=%4C %Z%{%b%d} %-15.15F(%4l) %s
# sorting for inbox
folder-hook ! set sort=date-received
# display To, not From, in Sent folder
#folder-hook Sent set index_format=%4C %Z%{%b%d} %-15.15F(%4l) %s
# sorting for lists
# display the author's real name, not the list (or my name!) in index view
folder-hook lists.* set sort=threads ; set sort_aux=date ; \
set index_format=%4C %Z %{%b%d} %-15.15n(%4l,%3M) %s%|
#
---
| Mutt also has a G or g function for groups but I don't
| use it.
|
| i'll look into it.
g is group-reply (aka Reply To All), G is download POP
messages
-D
Re: DocBook User Documentation
On Thu, Jul 12, 2001 at 10:07:36PM -0400, David H. Silber wrote: | Greetings, | | I would like to use DocBook to create the documentation for a project | I'm working on (see signature) but have been unable to find instructions | for generating various formats of output. See the newbiedoc project, they did it. newbiedoc.sourceforge.net. -D
Re: GRUB Rieserfs
Disclaimer : I have no experience with ReiserFS, but I do use grub. On Wed, Jul 11, 2001 at 11:54:49AM -0400, Chuck Stickelman wrote: | San Segkhoonthod wrote: | | GRUB do *support* ReiserFS. My debian boxes have | ReiserFS root file system and I boot them with GRUB. | | Then I've done something wrong! | I'll try providing more specifics. | kernel 2.4.5 | hda1~120MBext2/boot | hda2~128MBswap | hda3~27GBreiserfs/ | I've tried: | root (hd0,0) | kernel /boot/vmlinuz-2.4.5 root=/dev/hda3 single There's the problem. There is no directory named boot on (hd0,0). Grub doesn't understand the OS's mount tables because it isn't your OS. Every partition is called / by grub, or (hdn,m)/ if you want to specify which disk/partition. Instead use root (hd0,0) kernel /vmlinuz-2.4.5 root=/dev/hda3 single and it should work fine. HTH, -D
Re: Scroll Wheel
On Mon, Jul 09, 2001 at 10:03:16PM -0700, Erik Steffl wrote: | Jesper Holmberg wrote: | | What is the recommended way of adding support for the scroll wheel | (Logitech mouse) under X4.0.3 and Woody? | | here's what I have, it's for the mouseman wheel (with the sidebutton): snip I tried this with my MouseMan Wheel (the USB/PS2 kind with the USB-PS2 adapter on) and the wheel didn't work. Are you using gpm? I am, and I tried setting its protocol to 'mman'. That could be my source of trouble. However, using IMPS/2 in X and imps2 in gpm works (with 4/5 as wheel) but the side button and the wheel button are both the middle button. -D
samba mounting problem
I did some more experiementing with samba mounts last night, and read the mount, fstab, and smbmount man pages, but couldn't find an explanation or a solution. I would like to have fstab entries that reference the samba shares on the win* boxes on my LAN and provide a global (ie under /mnt) location where they can be mounted and accessed by users. In my experimentation it seems that smbmount will only mount the share if the user mounting it owns the mount point. This means that only root and 1 user can mount any given samba share. According to the mount (or fstab) man page the option user will allow any user to mount the device, but mount/smbmount doesn't seem to be obeying that option. Is there any solution other than (a) make a wrapper that will chown the mount point first or (b) make each user have their own mount point (such as under $HOME)? Thanks for any input, -D
