Linux-Misc Digest #253, Volume #26 Mon, 6 Nov 00 23:13:01 EST
Contents:
Re: ext2 undelete ([EMAIL PROTECTED])
Re: Why, ext2 don't need defrag ([EMAIL PROTECTED])
Re: changing shells ([EMAIL PROTECTED])
Re: Looking for web-based e-mail checker for linux (Rubin Bennett)
Re: Looking for web-based e-mail checker for linux (nuk)
Re: KDE vs GNOME: specific issues ([EMAIL PROTECTED])
Re: Upgrading the Linux Kernel (Dances With Crows)
Re: Linux/UNIX=Windows (Dances With Crows)
Re: Netware-like file system rights on Linux (Dustin Puryear)
Re: OpenLDAP and MS Outlook (Dustin Puryear)
Re: Linux/UNIX=Windows
permission for ntfs... (tue)
Re: Yamaha CRW8824 and Tekram 390F (Frank Pittel)
Re: Whats good about debian and slackware? (John Hasler)
NFS export problem with netgroup (David YEUNG)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Reply-to: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.system
Subject: Re: ext2 undelete
Date: Tue, 07 Nov 2000 02:17:10 GMT
[EMAIL PROTECTED] writes:
> On Sat, 04 Nov 2000 11:26:38 +0000 MENON Jean-Francois
> <[EMAIL PROTECTED]> wrote:
>
> | I'm working on a kind of undelete for ext2 fs. I would like to
> | know if there is other people working on the same problem.
>
> I don't know of a problem.
Indeed :-)
> | I would like to know how I can override the "unlink" function with my
> | own function?
>
> Do you want to do this across all filesystem types? Or just for
> ext2 only? If your intent is to change the behaviour of unlink() to
> move a file to a garbage area and later empty the garbage, you need
> to consider a lot of things about how this should work.
>
> 1. How will you empty the garbage?
>
> 2. What do you want unlink() to do is it is a file with multiple links?
A not-too-outrageous methodology might be for unlink() to do two
things:
I. Create, (if it's not already there), a _hard_ link to the inode,
named [mount-point]/.deleted/[inode-number]
II. Drop an entry naming the filename, inode, and time of deletion
into [mount-point]/.deleted/deletion.log
Resolution of Question #1 would be that a daemon could go look at
[mount-point]/.deleted/deletion.log and purge the "eldest links" once
in a while, likely "when things get pretty full."
Resolution of #2 heads in two directions:
a) If the link was a soft link, then the link can outright go away
immediately; deleting a soft link doesn't normally make the file
go away.
b) Suppose file FOO had three hard links.
The first "deletion of hard link" results in a hard link being
created in /.deleted/
The second and third deletions don't make another hard link,
because the hard link is _already_ there in /.deleted/
This actually sounds surprisingly usable, as a scheme...
--
(concatenate 'string "cbbrowne" "@acm.org") <http://www.ntlug.org/~cbbrowne/>
Rules of the Evil Overlord #18. "I will not have a son. Although his
laughably under-planned attempt to usurp power would easily fail, it
would provide a fatal distraction at a crucial point in time."
<http://www.eviloverlord.com/>
------------------------------
From: [EMAIL PROTECTED]
Reply-to: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.system
Subject: Re: Why, ext2 don't need defrag
Date: Tue, 07 Nov 2000 02:17:37 GMT
Robert Heller <[EMAIL PROTECTED]> writes:
> [EMAIL PROTECTED] (Paul Lew),
> In a message on Sun, 05 Nov 2000 17:36:15 GMT, wrote :
> PL> ??? SuSE 6.3 has the following msg:
> PL> EXT2-fs warning: checktime reached, running e2fsck is recommended
> PL>
> PL> Doesn't e2fsck also "defrag" a bit as well as checking the file-system?
>
> No. It just checks the file system. The "EXT2-fs warning: checktime
> reached, running e2fsck is recommended" is not specific to SuSE. It is
> a normal part of the EXT2-fs code.
That's not 100% fair; supposing e2fsck finds problems, it may wind up
shifting the broken files around.
Mind you, I'm not sure that I'd count the ability to reorganize
_broken_ files to be a major benefit :-). (Thus, it's only less than
100% fair by a Very Slim Margin...)
--
(concatenate 'string "cbbrowne" "@acm.org") <http://www.ntlug.org/~cbbrowne/>
Rules of the Evil Overlord #61. "If my advisors ask "Why are you
risking everything on such a mad scheme?", I will not proceed until I
have a response that satisfies them." <http://www.eviloverlord.com/>
------------------------------
From: [EMAIL PROTECTED]
Reply-to: [EMAIL PROTECTED]
Subject: Re: changing shells
Date: Tue, 07 Nov 2000 02:17:41 GMT
Robert Schweikert <[EMAIL PROTECTED]> writes:
> I know this is simple, I just can't seem to find the information.
>
> How do I change my shell from Bash to TCshell as the default?
The command you're after is "chsh" which is short for "change shell."
The relevant configuration is in /etc/passwd, as well as /etc/shells
[which controls what choices you can make when you use chsh].
I used to use tcsh, many moons ago; I have found zsh to be
considerably more powerful, particularly in its interactive "globbing"
capabilities. <http://www.zsh.org/>
--
(concatenate 'string "cbbrowne" "@hex.net") <http://www.ntlug.org/~cbbrowne/>
"I've discovered that P=NP, but the proof is too long to fit within
the confines of this signature..."
------------------------------
From: Rubin Bennett <[EMAIL PROTECTED]>
Subject: Re: Looking for web-based e-mail checker for linux
Date: Tue, 07 Nov 2000 02:45:48 GMT
Bob Hauck wrote:
>
> On Fri, 3 Nov 2000 10:50:16 -0500, J.R. Farrar <[EMAIL PROTECTED]> wrote:
>
> >Anyone know where I can find a down and dirty e-mail interface using a web
> >server on my Redhat box?
>
> I've used Twig with good results. It may be more than you need, but
> what the heck. http://www.screwdriver.net/
>
> --
> -| Bob Hauck
> -| To Whom You Are Speaking
> -| http://www.haucks.org/
I've used both twig and IMP (http://horde.org/imp) and MUCH prefer IMP.
It's interface is a lot more intuitive and it's code is a lot cleaner
(IMHO)
Rubin
------------------------------
From: nuk <[EMAIL PROTECTED]>
Subject: Re: Looking for web-based e-mail checker for linux
Date: Mon, 06 Nov 2000 19:55:23 -0700
Rubin Bennett wrote:
> Bob Hauck wrote:
> >
> > On Fri, 3 Nov 2000 10:50:16 -0500, J.R. Farrar <[EMAIL PROTECTED]> wrote:
> >
> > >Anyone know where I can find a down and dirty e-mail interface using a web
> > >server on my Redhat box?
> >
> > I've used Twig with good results. It may be more than you need, but
> > what the heck. http://www.screwdriver.net/
> >
> > --
> > -| Bob Hauck
> > -| To Whom You Are Speaking
> > -| http://www.haucks.org/
> I've used both twig and IMP (http://horde.org/imp) and MUCH prefer IMP.
> It's interface is a lot more intuitive and it's code is a lot cleaner
> (IMHO)
>
> Rubin
I haven't used this personally, but I've heard good things about it:
NeoMail
Find it on FreshMeat.net
Monte
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.advocacy,linux.redhat,alt.os.linux
Subject: Re: KDE vs GNOME: specific issues
Date: Tue, 07 Nov 2000 02:57:24 GMT
[EMAIL PROTECTED] (Donovan Rebbechi) writes:
> On Tue, 07 Nov 2000 00:31:04 GMT, Jeff Jeffries wrote:
> >In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED]
> >(Donovan Rebbechi) wrote:
>
> >Why not just get Codewarrior for Linux, which does all of this?
>
> Does Codewarrior support automake/autoconf based builds ?
>
> Does it have built-in support for GTK/GNOME and/or QT/KDE ?
>
> Glade has support for XML based GNOME GUIs which can be loaded at *run time*.
> I don't believe anything else has this.
The other well-known system that provided this capability was
NeXTStep, albeit with the consideration that it's "NIB" system vastly
predates XML.
libGlade is, really and truly, a _very_ valuable thing about
Gnome/GTK...
--
(concatenate 'string "cbbrowne" "@hex.net") <http://www.ntlug.org/~cbbrowne/>
A man without religion is like a fish without a bicycle.
------------------------------
From: [EMAIL PROTECTED] (Dances With Crows)
Subject: Re: Upgrading the Linux Kernel
Date: 7 Nov 2000 03:12:25 GMT
Reply-To: [EMAIL PROTECTED]
On Mon, 06 Nov 2000 20:56:01 GMT, mike wrote:
> I am presently using Redhat 6.1. It comes with kernel version
>2.2.12-20.
>I was wondering if upgrading this kernel by patching it could,
>in any way damage the system causing loss or improper functioning.
>Assuming one keeps the major and minor version of the kernel the
>same, ie. version 2.2, is it a good idea to patch it to the latest
>version?
Patching one of RedHat's custom kernels is unlikely to succeed. They've
made a number of alterations/improvements/sillinesses to the standard
kernel code, so the patch-2.2.XX.gz files you can download from
ftp://ftp.us.kernel.org/ will fail when you try to run them through
patch.
Your best bet is to grab the official 2.2.17 tarball if you can, then
unpack it and recompile. You should *always* keep a known good kernel
image in /boot and accessible through /etc/lilo.conf so that when/if
your new kernel can't boot, you have something to fall back on.
There's a guide to compiling a kernel at
http://linuxdoc.org/HOWTO/Kernel-HOWTO.html and I suggest you read it
before trying it out. It's not hard, but it will take a while on slower
systems. Going through the process will at the very least teach you
something about the workings of Linux. Good luck,
--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin / Workin' in a code mine, hittin' Ctrl-Alt
http://www.brainbench.com / Workin' in a code mine, whoops!
=============================/ I hit a seg fault....
------------------------------
From: [EMAIL PROTECTED] (Dances With Crows)
Subject: Re: Linux/UNIX=Windows
Date: 7 Nov 2000 03:12:27 GMT
Reply-To: [EMAIL PROTECTED]
On Mon, 06 Nov 2000 23:31:27 GMT, Peter wrote:
>On Mon, 06 Nov 2000 21:37:56 GMT, "Mats Pettersson" wrote:
>>If you get to something like compiling and installing sendmail, mysql,
>>apache... you preferable have to be a C/C++ programmer, like to read
?? All right, I *am* a programmer (not a particularly good one, but one
nontheless) and I never had problems compiling/installing those programs
that required me to go in and patch code. I did have to read the
directions before doing those things. This is a hurdle that some people
cannot overcome, I guess.
If you want to see how much worse it can be, try installing Oracle8i for
Linux. That was worse than upgrading glibc from source.
>>Don't get me wrong, i admire the people doing all those free
>>applications and i understand they don't have all the time in the
>>world to write big comprehencive manuals, especially for free, but
>>nevertheless it's a problem.
Yep. Most of the people working on the apps/utilities would rather code
than write docs, and good doc-writers are in short supply.
>[I can burn a CD that will install NT on]
>100s of PCs with no problems. To do the same with Linux, you have to
>start with far more knowledge of Linux and people with that knowledge
>are in short supply.
Not as short as you think. In the USA, anyway, there are a lot of LUGs,
and a fair number of people at these LUGs who know a great deal. BTW,
there's this thing called "kickstart" that RedHat does which is very
similar to the process you describe.
> Hard disks remain unchanged after 30 years.
???! OK, so vast improvements in capacity, size, price, and transfer
speed are enhancements rather than radical changes. OK, so things like
SMART and GMR heads aren't that visible to the end-user. OK, so maybe
you don't like the idea of hot-pluggable FireWire 30-giggers. *I* think
they've changed a lot.
>admitting disk brakes are also cheaper to make. The closest equivalent
>in the computer industry is the serial port. USB's predecessor worked
>on HP calculators 20 years ago but no one took up the idea because it
>was too new.
...which is one of the reasons many consumers are still stuck with IDE
and x86 when better stuff exists. However, backwards combatibility was
one of the design goals of the first IBM PC, and it will be with us for
a very very long time. (As the old joke goes: SGI invents a box that
fits in your shirt pocket, has 1G RAM, 2T disk, a 4096x4096 true-color
display projected right on your eyeballs, built-in wireless Ethernet,
and costs $300. The first thing Marketing asks is, "Will it run
DOS?")
>The mass market, the user of the future, works at about the level of
>PHP/Visual Basic/Excel macros so will not be able to resort to reading
>C code.
<semi-rant>
"User of the future" my left hind buttock. Anyone in the business of
education will tell you that if you treat kids (users) like they're
stupid, most of them will act stupid. If you treat them like they're
smart and capable of learning things, most of them will do so. A
computer is a very complex device, and most people (myself included) are
incapable of using it to its true potential--but the least that can be
done is to show people that the potential is *there*.
</semi-rant>
Anyway, it is possible to write hideous kludges in *ANY* computer
language. Well-written C++ code is far easier to understand than
poorly-written Visual Basic.
>They need the GUI apps to warn them disk space is running out
>and show them where the space went. That was where I started this
>discussion. What is the Linux equivalent of Diskeeper or Raxco?
OK, I'll bite. What in the seven hecks is Diskeeper or Raxco supposed
to do? du and df do a great job of showing me which directories and
filesystems are filling up, and tar/bzip2, rm, and rpm -e are very nice
for saving space or getting rid of stuff. There are graphical
interfaces for these tools--I'm just old-school enough to like the
command line better.
>I do not need the GUI apps but find them many times faster so will
>stick with NT Workstation until the right set of apps are Gnomed.
Whatever floats your boat. If you want to speed up the process of
getting the right set of apps going, check Sourceforge and see if
there's a worthy project or 2 that could use help. Documentation,
artwork, and ordinary user testing are just as important as coding, so
don't feel you have nothing to contribute.
--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin / Workin' in a code mine, hittin' Ctrl-Alt
http://www.brainbench.com / Workin' in a code mine, whoops!
=============================/ I hit a seg fault....
------------------------------
From: [EMAIL PROTECTED] (Dustin Puryear)
Subject: Re: Netware-like file system rights on Linux
Reply-To: [EMAIL PROTECTED]
Date: Tue, 07 Nov 2000 03:14:11 GMT
On Sun, 05 Nov 2000 01:40:55 GMT, Zip <[EMAIL PROTECTED]> wrote:
>I come from the Netware camp and am interested in moving over to SAMBA on
>Linux. In Netware, when rights are granted on a directory, those rights flow
>down the file system until they are masked by a lower level directory, but
>only the rights that are masked are restricted below that level, and any other
>rights continue to flow down the file structure.
>
>My question is, can this functionality be emulated in SAMBA? Currently we have
>about 5 drive letters mapped when we log into Netware. Will we have to map
>more drive letters with SAMBA, or can we convert to use \\system\share?
You will want to assign group rights to subdirectories as appropriate and
let the Linux permissions, and not Samba, allow or disallow access. Personally,
I find this much cleaner than having several drive mappings, which can get
confusing fast.
--
Dustin Puryear <$email = "dpuryear"."@usa.net";>
Integrate Linux Solutions into Your Windows Network
- http://www.prima-tech.com/integrate-linux
------------------------------
From: [EMAIL PROTECTED] (Dustin Puryear)
Crossposted-To: comp.os.linux.networking
Subject: Re: OpenLDAP and MS Outlook
Reply-To: [EMAIL PROTECTED]
Date: Tue, 07 Nov 2000 03:17:20 GMT
On Sun, 5 Nov 2000 19:16:48 +0100, Lars Grenzendoerfer
<[EMAIL PROTECTED]> wrote:
>i'm using OpenLDAP on my Linux server. How do I have to configure OpenLDAP
>to work with MS Outlook? What attributes and objectclasses do I have to add?
>Are there any other changes to be made?
All you really need to worry about are the mail and cn attributes. FYI, there
are several references to be found at openldap.org. Just so you are warned,
the filter used by Outlook is rather weak so don't expect too much.
--
Dustin Puryear <$email = "dpuryear"."@usa.net";>
Integrate Linux Solutions into Your Windows Network
- http://www.prima-tech.com/integrate-linux
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: Linux/UNIX=Windows
Date: Tue, 07 Nov 2000 03:27:48 GMT
On 7 Nov 2000 03:12:27 GMT, Dances With Crows <[EMAIL PROTECTED]> wrote:
>On Mon, 06 Nov 2000 23:31:27 GMT, Peter wrote:
>>On Mon, 06 Nov 2000 21:37:56 GMT, "Mats Pettersson" wrote:
>>>If you get to something like compiling and installing sendmail, mysql,
>>>apache... you preferable have to be a C/C++ programmer, like to read
>
>?? All right, I *am* a programmer (not a particularly good one, but one
>nontheless) and I never had problems compiling/installing those programs
>that required me to go in and patch code. I did have to read the
>directions before doing those things. This is a hurdle that some people
>cannot overcome, I guess.
>
>If you want to see how much worse it can be, try installing Oracle8i for
>Linux. That was worse than upgrading glibc from source.
That's just because the installer in the main directory is defective. If you
search deja for oracle8'n'linux, it is pretty easy to find the correct
installer.
------------------------------
From: tue <[EMAIL PROTECTED]>
Subject: permission for ntfs...
Date: Tue, 07 Nov 2000 03:30:06 -0000
Hi ...
I'm using the new linux-mandrake 7.2 and able to mount the win2000 ntfs
partition. It seems that I can only read the contents from the partition
and could not write on it although I log in as a root. I have tried to
change the permission .... but it will say that it is a read-only devices.
To overcome this problem, I have to create another small partition in
fat32. I would like to use the ntfs still ... is there any work around for
this problem ?
thanks
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: Frank Pittel <[EMAIL PROTECTED]>
Subject: Re: Yamaha CRW8824 and Tekram 390F
Date: Tue, 07 Nov 2000 03:37:07 -0000
I had the same type of problem with a cdrom drive. the 390F needs an active
terminator and the drive I had didn't do active termination.
Aris Basic <[EMAIL PROTECTED]> wrote:
: Hi
: i have Tekram 390F controler adn it was working with very old CDROM
: and with external ORB Drive so i suppose its ok
: but i cannt bring it to work properly with Yamah8824 Drive i allways get
: some scsi errors
: (linux 2.2.16 sym875 kernal driver)
: on cdrecord -scanbus i usaly get some kernel SCSI errors and cdrecord tells
: something like
: Drive returned more data ten it should
: drive gets recognized on booting (i triedn turning on 512 block feature but
: that made more troubles)
: can anyone have an idea what could be wrong ?
: (Firmware upadate of controler or drive or both) ?
: or anything else
: thx
: Aris
--
Keep working millions on welfare depend on you
===================
[EMAIL PROTECTED]
------------------------------
From: John Hasler <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux
Subject: Re: Whats good about debian and slackware?
Date: Tue, 7 Nov 2000 02:21:48 GMT
H. Gibson writes:
> One minor moan... lots of software is distributed in rpm's now.
There is very little that is not also packaged for Debian.
--
John Hasler
[EMAIL PROTECTED]
Dancing Horse Hill
Elmwood, Wisconsin
------------------------------
From: David YEUNG <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,comp.os.linux.networking
Subject: NFS export problem with netgroup
Date: Tue, 07 Nov 2000 11:30:39 +0800
I have a number of Linux machines and try to control the NFS export
using NIS netgroup, but it seems to have problem to do so. Here
is my configuration with a server 'hqlx10' and client 'rcz006':
In the server machine, I have setup the NIS and NFS server:
hqlx10 # ypcat -k netgroup
linux (rcz006,,)
hqlx10 # cat /etc/exports
/usr/local @linux(ro)
hqlx10 # exportfs -a
hqlx10 # exportfs
/usr/local
@linux
In the client machine:
rcz006 # ypcat -k netgroup
linux (rcz006,,)
rcz006 # mount hqlx10:/usr/local /usr/local
mount: hqlx10:/usr/local failed, reason given by server: Permission denied
I always get the Permission denied message when I mount it. However, if I change
the server /etc/exports to hostname, e.g:
hqlx10 # cat /etc/exports
/usr/local rcz006(ro)
Then the mount is susccessful.
Does anyone have idea what is wrong, or how to solve this problem.
I am using Redhat 6.2 with the latest kernel (2.2.16) on a PII-500 machine
Thanks
David Yeung
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
