more ash/dash/bash questions
Hello List, I've been writing shell scripts to be run as root lately; I'm not aware of any security-risk involving bash but I'd much more rather run those scripts with ash or dash instead of bash; some years ago I remember reading here that pointing /bin/sh to dash or ash would break a lot of important scripts in Debian; so, instead of pointing /bin/sh to ash or dash, I'm starting my scripts with ``#!/bin/ash''; I know that if I accidentaly remove ash my scripts will break but I decided to run the risk! but, in order to mimimize the risk, which interpreter should I use in my scripts, ``#!/bin/ash'' or ``#!/bin/dash'' ? []s & TIA! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
gateway to linux.debian.user
hi all, this list used to be gatewayed to linux.debian.user and now it seems that's not true anymore; is this correct? am I missing something? please reply to me directly as I'm NOT subscribed (I used to access it through linux.debian.user) to the list; thanks! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
integrity of .deb files
hi all, this is a repost of a message that AFAIK didn't reach the newsgroup :-( I'm downloading some debian packages that I wish to keep but don't intend to install right now; what's the best way to check their integrity? I see that dselect has a routine for doing this but wish there was a more simple way of doing it; like: "dpkg --test-integrity package.deb" any clues? I'm using 1.3 bo TIA! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
checking package integrity
hi all, is there a way to check package intgrity in a given .deb file? I'm using debian 1.3 (bo); beers & TIA! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
simple smail questions
hi all, I'm using Debian 1.3 bo with a 2.0.29 kernel; my smail installation is preety standard; I have a permanent connection to the internet; smail is working fine for most everything here; ..but I still have some questions.. see if you can help me, please: - I saw a message here stating (not in these words) that "smail is dead; use exim instead"; is it true? I can see that smail is still (according to Debian) the "reccomended MTA for Debian"! note that I'm not spam-asking which one you use! :-) - on the subject of moving files from /var/spool/smail/error to /var/spoll/smail: I did just that yesterday and the files are still there.. I'm afraid my crond is *not* calling runq. how do I make sure runq is beeing called? there is a /etc/smail/crontab but I don't see how crond would find it there.. - how do I set the retry fequency for smail; after reading the man page I just created /etc/smail/retry with a single line like '* 55m/3d' hoping that it would instruct smail to keep trying each 55 min during 3 days; it looks like it didn't work: I'm still getting error messages after 1 hour; what else do I have to do? 'killall' something? TIA! -- Otavio Exel /<\oo/>\Nada tema, com Smith nao ha problema. [EMAIL PROTECTED](Dr Zacary Smith 'in' Perdidos no Espaco) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: re-thinking partitions
> > > > - are symlinks really fast? > > >Quite, except on NFS. > > could you be more specific? could you describe a situation where > > symlinks would slow-down things and a possible solution eliminating > > thesymlink? > Normally, no NFS data is cached on the client, which means that for > a small file, accessed via a symlink, the time to get the file will > be twice, because the server will be contacted once to resolve the > symlink, and a second time to access the file. This can get quite > slow if you've got bad latency. > Solution: don't use symlink forests on NFS exported volumes. why not have symlink resolving done by the server? AFAIK there's no need for the client to know about symlinks! > Phil. -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: re-thinking partitions
Dale Martin wrote: > > Daniel Mashao <[EMAIL PROTECTED]> writes: > > why not put /var and /tmp in the same root partition? > > When you run out of space on your root partition while compiling, it's a > real pain. I've had it happen. I've never run out of space on /var, or on > / if I don't have /tmp on it. So, I have /tmp soft linked to the partition > with the most free space on my system. (This is a peculiarity of my > environment, I know, but I still think that 16M is a tad small for /tmp, > which is what I was originally responding to.) yes, all this partitioning stuff becomes not so important if you use symlinks; AFAIK the only true reason to avoid symlinking /var or /tmp into another partition is that /var and /tmp are more vulnerable to crashes on (say) power faillures; thanks! > > Later, > Dale > -- > + finger for pgp public key -+ > | Dale E. Martin | University of Cincinnati Savant Research Laboratory | > | [EMAIL PROTECTED]| http://www.ececs.uc.edu/~dmartin | > +--+ -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: re-thinking partitions
Daniel Mashao wrote: > > On 14 Nov 1997, Dale Martin wrote: > > > Philippe Troin <[EMAIL PROTECTED]> writes: > > > > > This is a lot of space wasted unless you have large spool directories > > > (news/mail). For a reasonable single-user station, 64MB should be > > > largely enough on /var. /tmp is left to your choice (16 is a good > > > number). > > > > I would find 16M for /tmp WAY too small. I do development, and the > > compiler wants to put it's temporary files /tmp - some of them can be > > why not put /var and /tmp in the same root partition? basically because I was afraid symlinks could slow it down; but I'l do just that; create /dev/hda2/tmp and /dev/hda2/var and symlynk /tmp and /var there; thanks! > /---/ > Daniel J. Mashao > Electrical Engineering[EMAIL PROTECTED] > University of Cape Town http://www.ee.uct.ac.za/~daniel > Rondebosch, 7700, S. Africa(w) 27+21+6502816 (h) 27+21+6863662 > /---/ -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
re-thinking partitions
hi everybody, I'm re-thinking my HD partitioning and would apreciate some advice! - are symlinks really fast? - I read somewhere that "500mb for /var and /tmp" is fine; what exactly does that mean? two partitions summing up 500mb? one 500mb partition and symlinks from /var and /tmp? - why is /usr/tmp symlynked to ../var/tmp instead of /var/tmp? same applies to /usr/spool; is it important? - what will happen with hard lynks if I "cp -d" from one part to another? TIA! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
backing-up dselect state files
hi everybody, I see all dselect stuff live in /var/lib/dpkg; I'd say /var is not a safe place to keep such important stuff; any opinions? now some questions: - what files should I backup from /var/lib/dpkg to be able to resume using dselect in case of crash in /var? - is there a program that would scan the filesystem and re-build dselect's state files? TIA! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
mutt for bo & libc5
hi everybody, I'm using Debian 1.3 bo libc5 only; for some reason (high degree of newbieness :-) I wouldn't like to mess with libc6 for now; last month I received a message from debian-changes@lists.debian.org stating that there was a mutt package for libc5; like that: > Description: > mutt - Text-based mailreader supporting MIME, PGP and threading. > Changes: > mutt (0.85-0) stable; urgency=low > . > * Fixed the "sendmail=no" problem reported by Damir J. Naden. > * This is essentially 0.85-3 built for libc5 systems. > * The "0.85-0" prevents dpkg from viewing it as an upgrade to > any libc6 release of 0.85. > * Built without preserved timestamps. I'm looking forward to have mutt installed but I just can't find this package in ftp.debian.org; could anyone please tell me where I can find it? this is the 3rd time I post this question!! (help me! I'm tired of downloading ls-lR.gz every day..) TIA! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
help needed installing hamm
hi everybody, I've got a newbie index of 2 in a 0..10 range :-) I've tried to update from a fresh install of bo to hamm but couldn't; as I'd like very mutch to experiment with hamm & 2.0.30 now I decided to try a fresh install of hamm; how do I do that? the set of disks I can see in ftp.debian.org/debian/hamm/../disks-i386 are bo based! btw: if there's no way to fresh-install hamm would you please help me with upgrading from bo? my problem is: I tried to follow the libc5-libc6-HOWTO but failed on the second step: 'dpkg -i ldso_???' worked fine but 'dpkg -i libc6_???' failed due to a conflict with libc5; am I missing something obvious? is there an other HOWTO on this subject? TIA! Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
bo-updates & other updates
hi everybody, I really don't understand ftp.debian.org directory structure; is there a document explaining every bit of it? it fisrt happened with samba-1.9.17p2; I read in the newsgroups that it was available (and it was important!) but couldn't install it with dselect: dselect still showed 1.9.16p11 as the current release; so I: - uninstalled my current 1.9.16p11 with dselect; - downloaded 1.9.17p2 using ftp; - installed 1.9.17p2 with 'dpkg -i ' is (was) there a more well-behaved way to do it? now I want to install mutt-0.58.0; I've received a message stating that mutt-0.58.0-i386.deb is available but I can't find it (let alone install it (let alone install it with dselect))!! I also would like to keep my installation up-to-date with whatever is in 'bo-updates'; is there a way to do it with dselect? btw: dselect is GREAT!!! congratulations, Debian guys! TIA! Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
just add IP_FORWARDING/GATEWAYING
hi everybody, just gave-up my first try at kernel compilation :-( the reasons why it failed are unimportant now, I probably overlooked one or more needed modules; the fact is that I was just trying to add IP_FORWARDING GATEWAYING capabilities to my Debian 1.3 in order to gain access to the internet through a PPP connection; now I'd like to do my second atempt configuring it *exactly* like Debian does except for the IP_FORWARDING GATEWAYING capabilities; how do I do this? is Debian's /usr/src/linux/.config available somewhere? BTW: is there a *good* reason why Debian 1.3 comes without IP_FORWARDING GATEWAYING included? TIA! Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
installing hamm via ftp
hi everybody, what are the parameters for dselect via ftp in order to install the latest hamm packages? I'm using a local up-to-date mirror and debian/.message says: > --- Unstable, or hamm. Access this release through dists/unstable. > > The current development snapshot is named hamm. Active development is > continuing here. Note that hamm contains many symbolic links into bo. ..but I don't know where to type "dists/ustable" :-( sorry for the dumb question! TIA! Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Debian & Adaptec AHA 2840VL/2842VL
hi everybody, I'm trying to install Debian 1.3.1 with an Adaptec AHA 2840VL/2842VL SCSI card; while the RESC1440 disk is beeing loaded I get a lot of error messages; some of them are: - encountered spurious interrupt - failed to initialize module ? - In swapper task - not syncing ..and it hangs! is there a way I can save/print the output of the loading process? I'd like to take a closer look at it or even post it here.. I've read in the SCSI-HOWTO that 90% of SCSI problems are related to cables; I can boot MSDOS w/o a problem; does this mean that cables are ok? is Linux more sensitive to cable problems than MSDOS? I booted with my RedHat 4.12 boot floppy and it looks like RedHat would have no problem with this SCSI card; is this kind of hardware support distribuition-specific? TIA! hlp: I'm using Win95 to write this!!!! -- Otavio Exel /<\oo/>\ [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .