Linux-Misc Digest #805, Volume #27 Mon, 7 May 01 20:13:02 EDT
Contents:
Re: database ([EMAIL PROTECTED])
Re: Shell-Script question, please give me a hand ? ("Harlan Grove")
Re: recycling XF86Config (wroot)
Re: modifying filesystems to accomodate files >2GB ("Jonathan Epstein")
Re: apt-get on Redhat (wroot)
Re: How to do a compressed dump backup? (Juergen Heinzl)
Re: Quick question: how to copy files *and* directories? (Jim Cochrane)
PPP dialing program? (Ken Arromdee)
Re: Quick question: how to copy files *and* directories? (Jason Green)
Making space for upgrade to RH7.1 (Leonard Evens)
Do I need /usr/local, /opt, /tmp and /var partitions? (wroot)
Re: database ("Steve Wolfe")
Re: GNOME RPMs ("Garry Knight")
Configuring CDRW & RH 7.1 ("ThanhVu Nguyen")
RedHat 6.1 Initrd? (MoosE)
Re: news/mail clients (Chiefy)
Re: PPP dialing program? (Bill Unruh)
Re: Configuring CDRW & RH 7.1 (Bill Unruh)
Re: Do I need /usr/local, /opt, /tmp and /var partitions? ("Solar Eclipse")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Subject: Re: database
Date: Mon, 07 May 2001 20:12:07 GMT
Stefano Ghirlanda <[EMAIL PROTECTED]> writes:
> Ying Hu <[EMAIL PROTECTED]> writes:
>> I need one database for our bioinformatics server on linux. Which
>> free one is the best?
> Warning: I have very limited db experience.
> Fastest: mysql.
> More features: postgresql.
> On www.mysql.org they have some comparison among databases if I
> rememeber well...
You can expect that www.mysql.org would have some degree of bias; at
minimum, that should be augmented by some equal-but-opposite bias:
<http://openacs.org/why-not-mysql.html>
There is now a way of making MySQL provide transactions, although this
requires using a different backend which is slower, likely eliminating
would-be performance advantages.
And the performance differences are controversial, in any case: they
seem only to apply to the very simplest of queries, and if you're
building web apps atop this, it's likely to be the Internet link that
will be the performance bottleneck, irrespective of which database is
being used.
--
(reverse (concatenate 'string "gro.gultn@" "enworbbc"))
http://vip.hyperusa.com/~cbbrowne/resume.html
"For be a man's intellectual superiority what it will, it can never
assume the practical, available supremacy over other men, without the
aid of some sort of external arts and entrenchments, always, in
themselves, more or less paltry and base. This it is, that forever
keeps God's true princes of the Empire from the world's hustings; and
leaves the highest honors that this air can give, to those men who
become famous more through their infinite inferiority to the choice
hidden handful of the Divine Inert, than through their undoubted
superiority over the dead level of the mass." --Moby Dick, Ch 33
------------------------------
From: "Harlan Grove" <[EMAIL PROTECTED]>
Subject: Re: Shell-Script question, please give me a hand ?
Date: Mon, 07 May 2001 20:16:38 GMT
Eric Chow <[EMAIL PROTECTED]> wrote...
>Would you please to teach me how can I split a file into different files in
>shell script?
>
>For example,
>
>index.dat
>========
>001 12
>002 25
>003 08
>004 12
>005 25
>006 08
>007 02
>
>content.dat
>=========
>001 aaaaaa AAAAA .....
>001 bbbbb BBBBBB .....
>001 cccccc CCCCCCC ....... some other contents
>002 .... another content for 002 ....
>002 ... 002 datas ...
>003 ....... 003 .....
>003 .. This is another 003 data ...
>004 ..... 004 ....
>005 ... 005 ...
>006 .... 006 ....
>007 ... 007 1 ...
>007 ... 007 2 ...
>
>result-12.dat
>==========
>aaaaaa AAAAA .....
>bbbbb BBBBBB .....
>cccccc CCCCCCC ....... some other contents
>..... 004 ....
>
>result-02.dat
>==========
>... 007 1 ...
>... 007 2 ...
...
Since this'd be immensely easier in a scripting language with associative
arrays than in any shell, I'll show an awk approach (you did install gawk
didn't you?).
awk 'BEGIN {
while ((getline < "index.dat") > 0) {
fn = "result-" (id[$1] = $2) ".dat"
printf "" > fn
close(fn)
}
FS = "[ \t]+"
}
/^[ \t]+/ { sub(/^[ \t]+/, "") }
{ fn = "result-" id[$1] ".dat"
match($0, FS)
print substr($0, RSTART + RLENGTH) >> fn
close(fn)
}' content.dat
The BEGIN action reads index.dat into an array flexibly (with default FS)
while 'resetting' the output files, then makes FS somewhat stricter. The
next pattern-action strips leading whitespace, if any, from lines in
content.dat. The final pattern-action sets the filename, locates the
beginning of text after the index key, appends the text without the index
key BUT WITH original whitespace to the given output file, then closes the
output file. The close statements ensure you don't have too many open files.
Alternatives include using join to produce lines with the 2nd field from
index.dat with all but the 1st field from content.dat, then sort the result,
and pass it through a simpler awk script to produce each file. (I prefer
fewer processes, though there's a lot of implicit open and explicit close
operations in my single awk script.)
------------------------------
From: wroot <[EMAIL PROTECTED]>
Subject: Re: recycling XF86Config
Date: Mon, 7 May 2001 16:18:18 -0400
Karsten Jensen wrote:
> On RH7.1 is
> the configuration file probably differnt because of the new X.
Possibly. The strangest thing is that my monitor (Panasonic PanaSync E50)
was never listed in Mandrake or Redhat options, so I used "E15" instead
(don't ask why). When it was finally there (RH7.1), X refused to work! I
tried everything: E15, E50, Custom. Very frustrating!
Wroot
------------------------------
From: "Jonathan Epstein" <[EMAIL PROTECTED]>
Subject: Re: modifying filesystems to accomodate files >2GB
Date: Mon, 7 May 2001 16:38:25 -0400
Sorry, never mind. bugzilla.redhat.com is the answer for the bug report,
and there are gzip patches described at:
http://www.gzip.org/#faq10
which is apparently part of the "gzip" rpm. Now, it's unclear to me why the
patches aren't incorporated into the code baseline ...
Also, I don't know for sure whether the patches would fix this particular
problem.
-Jonathan
> So now I have a workaround for my problem, but would still like to issue a
> bug report and/or build and test a fix on my own system. Any pointers for
> doing this? I have all the RedHat source code on CDs, but it's in a bunch
> of RPM's and I don't know which RPM corresponds to zcat's source.
------------------------------
From: wroot <[EMAIL PROTECTED]>
Subject: Re: apt-get on Redhat
Date: Mon, 7 May 2001 16:32:46 -0400
Fordman wrote:
> 1. Go to site and list all files
How can one do this in a script? Besides wget, I only know interactive
programs that will allow one to "go to site and list all files" (lynx, ftp,
netscape)
> 2. Pipe the output of that listing to a text file (rpm.old for ex.)
<snipped>
wget has mirroring capabilities, so I probably won't even have to write a
complicated script:
wget -m --level=1 ftp://$UPDATES
wget -m --level=1 ftp://$POWERTOOLS
rpm -Fvh ${DIR}/${UPDATES}*.i386.rpm ${DIR}/${POWERTOOLS}*.i386.rpm 2&>
`echo -n errors.; date +"%Y-%m-%d"`
Thanks
Wroot
P.S. On the downside, I need to download all updates, even the ones that
are irrelevant to my system.
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: How to do a compressed dump backup?
Date: Mon, 07 May 2001 20:55:35 GMT
In article <[EMAIL PROTECTED]>, james montoya wrote:
>Does anyone know the device name that provides a compressed dump backup
>on Redhat linux?
>I am currently using dump -0auf /dev/nst0 which isn't compressed.
>On a SUN Solaris system it is /dev/rmt/0cn for example.
[-]
Given you've got the right version you can enable compression via mt.
Ta',
Juergen
--
\ Real name : Juergen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (Jim Cochrane)
Subject: Re: Quick question: how to copy files *and* directories?
Date: 7 May 2001 15:07:36 -0600
In article <[EMAIL PROTECTED]>,
Jason Green <[EMAIL PROTECTED]> wrote:
>Floyd Davidson <[EMAIL PROTECTED]> wrote:
>> [EMAIL PROTECTED] (3FE) wrote:
>> >On Mon, 16 Apr 2001 11:19:52 +0200, Erik Veenstra <[EMAIL PROTECTED]> insisted:
>> >> [EMAIL PROTECTED] wrote:
>
>> >> > I want--in a single command--to copy a directory ***and all files and
>> >> > subdirectories within that directory*** to another location.
>
>> >> cp -a
>> >
>> >cp -R
>
>> ...
>How about copying two directories into one, while preserving links
>(including hardlinks), permissions, ownership, and timestamps?
How about:
mv -f foobar backupdir
mkdir foobar
find foo bar -print | cpio -pdmv foobar
>
>For example, say we have the following structure:
>
>foo/aa
>foo/sub/xx
>bar/.hidden
>bar/sub/yy
>
>We could do:
>
>$ rm -rf foobar
>$ cp -a foo foobar
>$ cp -a bar foobar
>
>But that doesn't work, because it copies the 2nd directory _into_ the
>destination directory.
>
>This is better:
>
>$ rm -rf foobar
>$ cp -a foo foobar
>$ cp -a bar/* foobar/
>
>But it misses the hidden file. Also, I think the 2nd copy won't preserve
>hardlinks (could be wrong though - don't have immediate access to a 'nix
>box to test this right now).
>
>I think a regular exp is needed to match the contents of the 2nd
>directory without matching . and .., but how?
>
>> Read the man page for cp...
>
>BTDT, and still scratching my head over this one.
>
>Hmm, just had a thought - maybe I could do this with tar...
>
>Anyone care to suggest a more elegant solution?
--
Jim Cochrane
[EMAIL PROTECTED]
------------------------------
From: Ken Arromdee <[EMAIL PROTECTED]>
Subject: PPP dialing program?
Date: 7 May 2001 21:15:31 GMT
I installed Mandrake 8 and couldn't find a graphical PPP dialing program
included. I'm not running Gnome/KDE. (I tried kppp anyway but I got some
odd problems and it ultimately didn't connect.) Has anyone got any
recommendations for good programs? Typing 'ifup ppp0' in a root shell every
time is getting annoying.
--
Ken Arromdee / [EMAIL PROTECTED] / http://www.rahul.net/arromdee
"You know, just once I'd like to meet an alien menace that wasn't immune to
bullets." --The Brigadier, Dr. Who
------------------------------
From: Jason Green <[EMAIL PROTECTED]>
Subject: Re: Quick question: how to copy files *and* directories?
Date: Mon, 07 May 2001 22:53:03 +0100
Floyd Davidson <[EMAIL PROTECTED]> wrote:
> [EMAIL PROTECTED] () wrote:
> >
> >If I understand right, you could do it like this:
> > mkdir foobar
> > for i in foo bar; do cd $i; cp -pr . ../foobar; cd ..; done
> >
> >Or...you could do, based on your attempt above:
> > rm -rf foobar
> > cp -a foo foobar
> > cp -a bar/. foobar
> >
> >Note I'm not using a wildcard - instead I use '.' to say the directory.
> >
> >Chris...
>
> Thanks for posting that!
Yes, Thankyou very much.
Magic, yes that's how it works - it's magic!
BTW, the for loop could be simplified:
for d in foo bar; do cp -a $d/. foobar; done
------------------------------
From: Leonard Evens <[EMAIL PROTECTED]>
Crossposted-To: redhat.general
Subject: Making space for upgrade to RH7.1
Date: Mon, 07 May 2001 16:39:22 -0500
I am starting to upgrade several computers to RedHat 7.1. But
I have a lot of junk on them that is probably not needed. In
my first try, I ran into problems since the installer thought
it needed more than 1.5 GB and was 125 MB short.
I've been trying to remove things from these computers, but it
is hard to be sure exactly what I need to keep. So here are some
questions.
In a typical installation of a gnome workstation, one still gets
a certain amount of KDE stuff. Is any of this really necessary
or could one remove all of it if one didn't intend to use KDE?
Generaly development libraries are needed if one intends to
develop something. But some of them are needed just for
other programs to work. Assuming that all I want to do at
present is some C and C++ programming, some writing of
Perl programs, and similar tasks, just how much of the development
libraries are safe to remove?
Any other suggestions of things to remove which I might not have
thought of? Any legacies I might still have left from RH5.* and
RH6.* that it would be safe to remove?
--
Leonard Evens [EMAIL PROTECTED] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
------------------------------
From: wroot <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,linux.debian.user,linux.redhat
Subject: Do I need /usr/local, /opt, /tmp and /var partitions?
Date: Mon, 7 May 2001 18:17:40 -0400
Hi,
Does it really make sense to create /usr/local or /opt or /tmp or /var?
If I run Debian Potato or Redhat6.2 and install
KDE2 or other software that is ahead of its distribution in /opt or
/usr/local (separate partition) instead of installing it with all other
software, will I be
a) more happy
b) less happy?
I'm asking this because it's not trivial to me how Debian, for example,
will update Pototo if I have KDE2 on it.
Thanks
Wroot
------------------------------
From: "Steve Wolfe" <[EMAIL PROTECTED]>
Subject: Re: database
Date: Mon, 7 May 2001 16:26:53 -0600
> > I need one database for our bioinformatics server on linux. Which free
> > one is the best?
> > Thanks for your any suggestions
>
> Warning: I have very limited db experience.
> Fastest: mysql.
> More features: postgresql.
Actually, in relistic tests, PostgreSQL is faster than MySQL. Of the
four tested (MySQL, Oracle, MS SQL, and PostgreSQL), only PostgreSQL was
able to scale fairly linearly. When MySQL was highly tuned, it was *just
barely* able to eek out a faster score than Postgres for exactly three (if
I recall) simultaneous connections, but at any other number, Postgres was
faster - and faster by a very significant margin once there were more
simultaneous connections.
> On www.mysql.org they have some comparison among databases if I
> rememeber well...
And those documents are skewed to the point of bordering on falsity. : )
steve
------------------------------
From: "Garry Knight" <[EMAIL PROTECTED]>
Subject: Re: GNOME RPMs
Date: Mon, 07 May 2001 20:55:35 +0100
Reply-To: [EMAIL PROTECTED]
In article <yonJ6.3984$[EMAIL PROTECTED]>, "bidalah"
<[EMAIL PROTECTED]> wrote:
> What are the packages that need to be installed in order to use GNOME?
If you look on Freshmeat you'll find the Gnome Installation Guide.
<URL: http://freshmeat.net>
--
Garry Knight
[EMAIL PROTECTED]
------------------------------
From: "ThanhVu Nguyen" <[EMAIL PROTECTED]>
Subject: Configuring CDRW & RH 7.1
Date: Mon, 07 May 2001 18:34:11 -0400
Hi,
I read that I need to recompile the kernel and add scsi support in order
to configure cdrw - but then someone told me RH 7.x should already recognize
CDRW, all I need is the cd record software ... But I need to use lilo to
boot to linux (which I don't, I boot to linux using a floppy )
Has anyone faced the same situation like I do and have everything worked
out ?
------------------------------
Subject: RedHat 6.1 Initrd?
From: MoosE <[EMAIL PROTECTED]>
Date: Mon, 07 May 2001 22:45:11 GMT
Hey all!
Can someone help me; I'm trying to remove the initrd startup from a
RedHat 6.1 install, but I can't find the bloody file. I deleted the
image from /boot, removed the entry from lilo.conf, re-ran lilo-- and
searched the whole harddrive for initrd* files, coming up with
absolutely nothing.
Any ideas? (It STILL loads at startup!)
Thanks a lot!
------------------------------
From: [EMAIL PROTECTED] (Chiefy)
Subject: Re: news/mail clients
Date: 7 May 2001 23:08:00 GMT
06 May 2001 18:46 UTC, andi did say to the dudes:
> I'm in the process of a slow migration to linux - but I need
> recommendations regarding mail and news clients.
SLRN is a very cool console news client. It's absolutely stuffed with
features. As for a console mail client, UW-Pine is very hard to beat.
We don't use X here, but I'm told that both programs look pretty good in
an Xterm, with SLRN making good use of colour.
--
Chiefy. <[EMAIL PROTECTED]>
Remember: LINUX spelled backwards is XUNIL.
------------------------------
From: [EMAIL PROTECTED] (Bill Unruh)
Subject: Re: PPP dialing program?
Date: 7 May 2001 23:19:12 GMT
In <9d739j$gih$[EMAIL PROTECTED]> Ken Arromdee <[EMAIL PROTECTED]> writes:
]I installed Mandrake 8 and couldn't find a graphical PPP dialing program
]included. I'm not running Gnome/KDE. (I tried kppp anyway but I got some
]odd problems and it ultimately didn't connect.) Has anyone got any
]recommendations for good programs? Typing 'ifup ppp0' in a root shell every
]time is getting annoying.
chmod a+rs /usr/sbin/{pppd,chat}
chmod a+s /usr/sbin/pppd
chmod a+rw /dev/ttyS?
Then type it in a user shell.
------------------------------
From: [EMAIL PROTECTED] (Bill Unruh)
Subject: Re: Configuring CDRW & RH 7.1
Date: 7 May 2001 23:21:23 GMT
In <[EMAIL PROTECTED]> "ThanhVu Nguyen" <[EMAIL PROTECTED]>
writes:
>I read that I need to recompile the kernel and add scsi support in order
No.
>to configure cdrw - but then someone told me RH 7.x should already recognize
Yes.
>CDRW, all I need is the cd record software ... But I need to use lilo to
>boot to linux (which I don't, I boot to linux using a floppy )
You do. That floppy also has lilo on it. But is is far far faster to not
use the boot floppy, but rather to put lilo on the floppy and tell it to
boot from the hard drive
lilo -b /dev/fd0
>Has anyone faced the same situation like I do and have everything worked
>out ?
------------------------------
From: "Solar Eclipse" <[EMAIL PROTECTED]>
Subject: Re: Do I need /usr/local, /opt, /tmp and /var partitions?
Date: Mon, 7 May 2001 16:48:52 -0700
Advantage to making such partitions:
If you must reinstall the OS, you won't lose the data in those partitions.
Disadvantage:
The size of the partitions is inflexible and if you end up finding that you
put too much into /var and /opt and not enough into /usr/local and /tmp, you
might find yourself very frustrated.
My first time installing Linux, I made all these partitions as recommended
and ended up regretting it for the above reason. I had all kinds of space
in partitions I didn't even know about and in no time ran out of space for
my MP3s!
I suggest you keep it down to one or two partitions plus a partition for
swap space.
Back up your data often (a good idea in general...though I myself don't
follow it) and it won't matter so much if you lose the information that
would have been protected by being in their own partitions.
And yet, I am but an egg when it comes to Linux, and I encourage you to get
a second opinion from somebody who knows more.
However, that has been my experience.
-A
"wroot" <[EMAIL PROTECTED]> wrote in message
news:9d777o$4if$[EMAIL PROTECTED]...
> Hi,
>
>
> Does it really make sense to create /usr/local or /opt or /tmp or /var?
> If I run Debian Potato or Redhat6.2 and install
> KDE2 or other software that is ahead of its distribution in /opt or
> /usr/local (separate partition) instead of installing it with all other
> software, will I be
> a) more happy
> b) less happy?
>
> I'm asking this because it's not trivial to me how Debian, for example,
> will update Pototo if I have KDE2 on it.
>
> Thanks
>
> Wroot
------------------------------
** 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 by posting to comp.os.linux.misc.
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
******************************
