March 2009 (#160):

• Mailbag
• Talkback
• 2-Cent Tips
• News Bytes, by Deividson Luiz Okopnik and Howard Dyckoff
• Away Mission - 2008 in Review - part 2, by Howard Dyckoff
• The Unbearable Lightness of Desktops: IceWM and idesk, by Ben Okopnik
• Joey's Notes: Bash shell basics, by Joey Prestia
• SCaLE 7 Speed-through, by Kat Tanaka Okopnik
• Development Builds Layered on Top of a Stable System by Means of Unionfs, by Dirk Wallenstein
• XKCD, by Randall Munroe
• The Linux Launderette

Mailbag

This month's answers created by:

[ Ben Okopnik, Kapil Hari Paranjape, René Pfeiffer, Neil Youngman, Rick Moen, Suramya Tomar, Thomas Adam ]
...and you, our readers!

---

Our Mailbag

---

Can you give your suggestion on this?

kathik kumar [karthikgiri1984 at gmail.com]

Thu, 26 Feb 2009 18:26:18 +0530

Hi,

I have installed my ipsec tools in my linux-server. I am observing some issues while testing.

These are the steps, i done:

============================
 
   https://ipsec-tools.sourceforge.net https://ipsec-tools.sourceforge.net for
installation (tar file location)
 
   a. Copy the file *.tar.gz to your linux machine
   b. Untar the file tar xvzf *.tar.gz
   c. Then issue ./bootstrap
   d. Issue ./Configure
   e. Go to src directory. Setkey is installed
   f. Setkey is available in the /sbin directory. so i changed the name and
Copy the setkey directory to /sbin.
   g. Then created set-key file.
 
[root@linux scripts]$ cat set-key
#!/sbin/setkey -f
flush;
 
spdflush;
 
add 10.0.0.1 11.0.0.2 ah 500 -A hmac-md5 0xabcd1234abcd1234abcd1234abcd1234;
add 11.0.0.2 10.0.0.1 ah 600 -A hmac-md5 0xabcd1234abcd1234abcd1234abcd1234;
 
#add 10.0.0.1 11.0.0.2 ah 500 -A hmac-sha1
0xabcd1234abcd1234abcd1234abcd1234abcd1234;
#add 11.0.0.2 10.0.0.1 ah 600 -A hmac-sha1
0xabcd1234abcd1234abcd1234abcd1234abcd1234;
 
add 10.0.0.1 11.0.0.2 esp 500 -E 3des-cbc
0x1234abcd1234abcd1234abcd1234abcd1234abcd1234abcd -A hmac-md5
0x1234abcd1234abcd1234abcd1234abcd;
add 11.0.0.2 10.0.0.1 esp 600 -E 3des-cbc
0x1234abcd1234abcd1234abcd1234abcd1234abcd1234abcd -A hmac-md5
0x1234abcd1234abcd1234abcd1234abcd;
 
spdadd 10.0.0.1/32[any] 11.0.0.2/32[any] any -P out ipsec
ah/transport//require esp/transport//require;
spdadd 11.0.0.2/32[any] 10.0.0.1/32[any] any -P in ipsec
ah/transport//require esp/transport//require;
 
spdadd 10.0.0.1/32[any] 11.0.0.2/32[any] any -P out ipsec
ah/tunnel/10.0.0.1-11.0.0.2/require;
spdadd 11.0.0.2/32[any] 10.0.0.1/32[any] any -P in ipsec
ah/tunnel/11.0.0.2-10.0.0.1/require;
 
  h. The issue is, The file has full permission, but it says permission
denied.
 
-rwxrwxrwx  1 root     root 1054 Feb 26 04:44 set-key
[root@linux scripts]$ ./set-key
./set-key: Permission denied.  -------------->>>>>>>>>>ISSUE
[root@linux scripts]$

Could you please let me know about this issue?

Thanks -Karthik

[ Thread continues here (15 messages/15.52kB) ]

---

[Apertium-stuff] English->Esperanto language pair released

Jimmy O'Regan [joregan at gmail.com]

Wed, 4 Feb 2009 18:12:14 +0000

---------- Forwarded message ----------

From: Jacob Nordfalk <jacob.nordfalk@gmail.com>
Date: 2009/2/4
Subject: [Apertium-stuff] English->Esperanto language pair released
To: Apertium-stuff <apertium-stuff@lists.sourceforge.net>

(English below / pli sube en la angla)

Ni ĵus publikigis novan lingvoparon, de la angla al Esperanto, apertium-eo-en. Temas pri la 27a publikigita aŭtomata tradukilo evoluiigita surbaze de la liber-programara motoro kaj ilaro por maŝina tradukado Apertium, la 4a en la angla kaj la 3a en Esperanto.

Oni povas jam nun provi la paron ĉe: https://www.apertium.org/index.php?id=translatetext&lang=eo

La komencaj celoj por la eldono de tiu versio estis:

* Por nescianto de la angla, esti kapabla kompreni la enhavon de
novaĵo pri ĝenerala temo.
* Esti kapabla kompreni, ĉu la koncerna novaĵo estas sufiĉe interesa
por esti konvene tradukita.
* Mallongaj frazoj devus esti akcepteble tradukitaj.

Ni kredas, ke ni atingis tiujn celojn sufiĉe akcepteble kaj ke la rezulto povas esti utila.

Entute, temas pri 'komunume' disvolvita pakaĵo, t.e. evoluigita per volontula laboro.

==Estonto==

La sekvaj paŝoj estos en diversajn direktojn:

'' - Plibonigi la aktualan version. - Labori pri la tradukilo de Esperanto en la anglan. - Evoluigi novajn ilojn surbaze de la tradukilo kaj la leksikonoj. ''

==Agnoskoj==

Koran dankon al ĉiuj evoluigintoj de la multaj malferma-fontaj iloj kaj liberaj lingvo-rimedoj, kiuj permesis starigi ĉi tiun novan tradukilon. Precipe dankon al ABC Enciklopedioj S.L., IULA of the Universitat Pompeu Fabra kaj Prompsit Language Engineering, S.L por la lingvoparoj Catalan->Esperanto kaj Spanish->Esperanto kiu donis la elirpunkto por tiu ĉi paro.

Koran dankon al Mike Leon (traduku.net), kiu liveris grandegan leksikonon kun 114 082 paroj da vortoj kaj idiomaĵoj.

Koran dankon al Francis Tyer, Prompsit Language Engineering, kaj Jimmy O'Reagan pro la konsiloj kaj pro la dizajno de la kerno de la sistemo de transdon-reguloj.

==Statistikoj==

;Leksiko:

'' Esperanto-vortaro: 22 462 lemoj English-vortaro: 13 733 lemoj Dulingva vortaro: 30 773 duopoj ''

;Reguloj:

'' 1a ŝtupo (chunk): 44 2a ŝtupo (inter-chunk): 8 3a ŝtupo (post-chunk): 6 ''

==Teknikaj notoj==

Vidu https://wiki.apertium.org/wiki/English_and_Esperanto

We've just released a new language pair, for English to Esperanto, apertium-eo-en. It's the 27th released automatic translator developped with the open-source Apertium machine translator engine, the 4th from English and the 3rd to Esperanto.

The pair is already available at https://www.apertium.org/index.php?id=translatetext&lang=eo

The stated release goals for this version were:

[ ... ]

[ Thread continues here (1 message/6.11kB) ]

---

can we assume that 1 block = 1 kibibyte when implementing Linux quota?

Mulyadi Santosa [mulyadi.santosa at gmail.com]

Fri, 27 Feb 2009 15:13:23 +0700

Hi all

As the subject says, is that assumption true? I asked that because I am confused whether Linux quota refers to filesystem block or it is another kind of block?

thanks in advance...

regards,

Mulyadi.

[ Thread continues here (2 messages/0.91kB) ]

---

Financial Post Story

Rick Moen [rick at linuxmafia.com]

Sat, 28 Feb 2009 14:35:33 -0800

Quoting Serge G. (galabs2000@gmail.com):

>    IT IS NOT A SPAM, but if you received that message second and plus
>    time JUST CLICK DELETE button and have a nice day.

Uh-huh.

[Stuff about cloning, from a possibly mentally ill individual snipped, about which see: https://jeffvrabel.com/2008/11/24/scarlet[...]ly-troubled-e-mail-i-have-just-received/]

In the headers, one notes:

 Received: from localhost ([127.0.0.1]:48981 helo=linuxmafia.com)
          by linuxmafia.com with esmtp   (Exim 4.61 #1 (EximConfig 2.0))
          id 1LdUqI-00032b-NR   ; Sat, 28 Feb 2009 11:27:27 -0800
 Received: from sendmail.canada.com ([199.71.40.31]:32455)
         by linuxmafia.com with esmtp   (Exim 4.61 #1 (EximConfig 2.0))
         id 1LdUpa-00032O-1z
         for <tag@lists.linuxgazette.net>; Sat, 28 Feb 2009 11:27:08 -0800
 Received: from gmail.com ([10.146.8.73]) by Sendmail.Canada.Com with Microsoft
         SMTPSVC(6.0.3790.3959); Sat, 28 Feb 2009 14:27:51 -0500
 From: "Serge G." <galabs2000@gmail.com>

So, the headers claim that it came from a typical throwaway GMail account (though that might be a lie), but in any event the mail somehow ended up at Canada.com, which is a commerical outfit billed as "Canadian news, entertainment, television, newspapers, free email and more". So, in addition to blacklisting the (alleged) GMail sender, I attempted to contact "abuse@canada.com" and "postmaster@canada.com", both of which are RFC-required mail drops for any domain that does SMTP mail.

Why? Because I really don't like having to blacklist entire large mail concerns, just because one bad apple sneaked abusive mail through. Those mail drops are the canonical places to get a provider's attention to exactly such problems. So, what happened next?

  This message was created automatically by mail delivery software. 
 
  A message that you sent could not be delivered to one or more of its 
  recipients. This is a permanent error. The following address(es) failed:
 
  postmaster@canada.com                                                         
     SMTP error from remote mail server after RCPT TOpostmaster@canada.com>:
    host mx1.canada.vsi.net [64.22.133.10]: 550 Mailbox disabled for this recipient - <postmaster@canada.com>                                            
   abuse@canada.com
   SMTP error from remote mail server after RCPT TOabuse@canada.com>:        
    host mx1.canada.vsi.net [64.22.133.10]: 550 Account temporarily suspended   due to inactivity. User must login to reactivate mailbox.                      

So, wow! They've decided to disable postmaster, and to ignore the abuse mailbox to the point where the software automatically disabled delivery. Well, if they want to completely disregard key RFC requirements on all SMTP senders -- requirements so key that you need to go out of your way to disable them in commodity SMTP software -- then my server doesn't need to hear from them again.

[ ... ]

[ Thread continues here (2 messages/5.38kB) ]

---

How to make wget exclude a particular link when mirroring

Suramya Tomar [security at suramya.com]

Wed, 04 Feb 2009 18:36:50 +0530

Hey Everyone,

I am trying to mirror an Invision Powerboard forum locally on my system (With permission from the admin) using wget and I am having issues.

When I start downloading wget visits each and every link and makes a local copy (like its supposed to) but in this process it also visits the "Log out" link which logs me out from the site and then I am unable to download the remaining links.

So I need to figure out how to exclude the Logout link from the process. The logout link looks like: www.website.com/index.php?act=Login&CODE=03 So I tried the following:

wget -X "*CODE*" --mirror --load-cookies=/var/www/cookiefile.txt 
https://www.website.com

but it didn't work.

I can't exclude the index.php itself because all the links are based off the index.php with parameters.

I have tried searching the web but didn't find anything relevant.

Any ideas on how to do it?

Thanks,

Suramya

[ Thread continues here (30 messages/51.32kB) ]

---

Linux Groups?

Rex Backus [rabackus at sonic.net]

Sat, 28 Feb 2009 12:45:32 -0800

Hi Gang,

In SUSE Linux Unleashed (pub 2005) on p20, there is a listing for Groups of Linux Users Everywhere. The URL referenced is broken. Does this list still exist, and if so where?

[https://glue.linuxgazette.com/]

Thanks,

Rex Backus

[ Thread continues here (3 messages/2.56kB) ]

---

email post on LG

Ben Okopnik [ben at linuxgazette.net]

Mon, 23 Feb 2009 13:06:52 -0500

----- Forwarded message from Danny Holstein <dgholstein@embarqmail.com> -----

Date: Sun, 22 Feb 2009 12:12:56 -0500 (EST)
From: Danny Holstein <dgholstein@embarqmail.com>
To: ben@linuxgazette.net
Subject: email post on LG

Ben;

I read a post of yours on mass emailing on Linux Gazette -- I found the post since I was looking for information on mall emailing in Linux. Since we have a small gourmet foods shop in New Jersey and have compiled a customer list of nearly 1000 emails, we figured we'd use it to announce a sale and specials; nothing nefarious about our intentions. It turns out kmail and evolution aren't well-suited to the task, a significant proportion of emails have been changed (earthlink/mindspring -> embarqmail), mistyped into the database, or have simply been abandoned; we got about a 15% bounce rate.

I imagine a good mass mailer would have hooks for database import (MySQL on a LAMP server) and have a means for "fixing" the database based on bounces and "UNSUBSCRIBE" requests.

Regards;

Danny Holstein

----- End forwarded message -----

-- 
* Ben Okopnik * Editor-in-Chief, Linux Gazette * https://LinuxGazette.NET *

[ Thread continues here (5 messages/6.63kB) ]

---

Hey James, Looking for a MySql DBA

Rick Moen [rick at linuxmafia.com]

Thu, 26 Feb 2009 18:08:36 -0800

[[[ Original message elided from the thread. -- Kat ]]]

Quoting NeoHire (Jobs) (jobs@neohire.com):

> Hey James 
> 
> I am helping a company in Santa Barbara find a MySql DBA for their company.
> They are privately held and doing very well. 
> 
> Take a look @ the role, and let me know your thoughts. 

"tag@lists.linuxgazette.net" has never been the address of Jim Dennis. It's always been the address of The Answer Gang, a public mailing list associated with _Linux Gazette_ magazine. Jim Dennis was for a very long time a member of that mailing list, but hasn't been subscribed for some years.

You might have gotten a clue about it being a mailing list from the word "list" as the hostname, by the way.

Please make sure you excise "tag@lists.linuxgazette.net" from all of your records as a means of reaching Jim: I simply never was that, and isn't now.

What you just did is the Internet equivalent of walking into a crowded cinema during a movie and yelling out "Hi, I heard Jim Dennis might be here. I want to talk to him about a job opening."

You really don't want to be known for that sort of thing. You really don't.

---

Disk Errors causing machine to freeze

Ben Whyte [ben at whyte-systems.co.uk]

Sat, 07 Feb 2009 11:55:03 +0000

Hi

When I am writing to my disk every now and again, the machine stops responding for a period. When I look in syslog I see messages like this

Feb  7 11:18:02 thor kernel: [ 1515.415879] ata1.00: exception Emask 0x0
SAct 0x0 SErr 0x0 action 0x6 frozen
Feb  7 11:18:02 thor kernel: [ 1515.415896] ata1.00: cmd
35/00:e0:57:5c:8e/00:03:08:00:00/e0 tag 0 dma 507904 out
Feb  7 11:18:02 thor kernel: [ 1515.415899]          res
40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Feb  7 11:18:02 thor kernel: [ 1515.415905] ata1.00: status: { DRDY }
Feb  7 11:18:02 thor kernel: [ 1515.415923] ata1: soft resetting link
Feb  7 11:18:03 thor kernel: [ 1515.964911] ata1.00: configured for UDMA/133
Feb  7 11:18:03 thor kernel: [ 1515.964911] ata1: EH complete
Feb  7 11:18:03 thor kernel: [ 1515.964911] sd 0:0:0:0: [sda] 1953523055
512-byte hardware sectors (1000204 MB)
Feb  7 11:18:03 thor kernel: [ 1515.964911] sd 0:0:0:0: [sda] Write
Protect is off
Feb  7 11:18:03 thor kernel: [ 1515.964911] sd 0:0:0:0: [sda] Mode
Sense: 00 3a 00 00
Feb  7 11:18:03 thor kernel: [ 1515.964911] sd 0:0:0:0: [sda] Write
cache: enabled, read cache: enabled, doesn't support DPO or FUA
Feb  7 11:21:07 thor kernel: [ 1709.741765] ata1.00: exception Emask 0x0
SAct 0x0 SErr 0x0 action 0x6 frozen
Feb  7 11:21:07 thor kernel: [ 1709.741780] ata1.00: cmd
35/00:e8:4f:76:ad/00:01:0a:00:00/e0 tag 0 dma 249856 out
Feb  7 11:21:07 thor kernel: [ 1709.741782]          res
40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Feb  7 11:21:07 thor kernel: [ 1709.741785] ata1.00: status: { DRDY }
Feb  7 11:21:07 thor kernel: [ 1709.741799] ata1: soft resetting link
Feb  7 11:21:08 thor kernel: [ 1710.444091] ata1.00: configured for UDMA/133
Feb  7 11:21:08 thor kernel: [ 1710.444091] ata1: EH complete
Feb  7 11:21:08 thor kernel: [ 1710.444091] sd 0:0:0:0: [sda] 1953523055
512-byte hardware sectors (1000204 MB)
Feb  7 11:21:08 thor kernel: [ 1710.444091] sd 0:0:0:0: [sda] Write
Protect is off
Feb  7 11:21:08 thor kernel: [ 1710.444091] sd 0:0:0:0: [sda] Mode
Sense: 00 3a 00 00
Feb  7 11:21:08 thor kernel: [ 1710.459700] sd 0:0:0:0: [sda] Write
cache: enabled, read cache: enabled, doesn't support DPO or FUA

So far this appears to cause the hard drive to collect errors and eventually for filesystem corruption to occur.

Has anyone seen this sort of thing before and are you able to shed any light on what might be happening.

This has been on going for a while now, and I have tried different drives and different cables.

Although it may not be immediately obvious its using SATA

Thanks

Ben

[ Thread continues here (6 messages/6.64kB) ]

---

GUI for idesk ??

J.Bakshi [j.bakshi at icmail.net]

Fri, 27 Feb 2009 22:32:22 +0530

Hello All,

Hope you all are well.

Is anyone of you using idesk presently ?

I have no idea about its present status. Is there any GUI available to configure the .lnk file and .ideskrc ? In my debian lenny apt shows only the idesk itself.

I would by happy if there is any GUI ( preferably .deb ) which can configure idesk.

Thanks Kindly CC to me

[ Thread continues here (4 messages/3.83kB) ]

---

free patent resource

[james at ryley.com]

Thu, 26 Feb 2009 22:42:56 -0500

Hi,

I saw that you had some trademark info at https://www.lg.ps.pl/issue97/TWDT.html. I don't know if you also have any interest in patents, but if so I wanted to suggest that you check out https://www.freepatentsonline.com and https://www.sumobrain.com

They are by far the best free patent searches on the web, and trademark data will shortly be added to them as well.

If you have a spot, a link would be great.

Sincerely, James

[ Thread continues here (2 messages/2.18kB) ]

---

Problems booting into Ubuntu 8.10

Deividson Okopnik [deivid.okop at gmail.com]

Thu, 12 Feb 2009 19:21:20 -0300

Hello everyone!

A student here brough to me a problem i couldnt solve, so i'm forwarding it to tag.

After an hibernation, his laptop froze and he had to do a hard-shutdown on it, and after that ubuntu stopped booting, giving him this kernel panic message:

[ 2.222518] Kernel panic - not sycing: VFS: Unable to mount root fs on
unknown-block(0,0)

After a few trys, he noticed that he can boot into other kernels he's got on his machine, but they don't work properly (weird beeps, auto-mount doesnt work, compiz - that kind of stuff)

Heres some messages:

Mounting something without su:
DBus error org.freedesktop.DBus.Error.NoReply: Did not receive a reply.
Possible causes include: the remote application did not send a reply,
the message bus security policy blocked the reply,
the reply timeout expired, or the network connection was broken.
 
Starting compiz:
root@ricardo-laptop:~/a# compiz
Checking for Xgl: not present.
Detected PCI ID for VGA:
Checking for texture_from_pixmap: not present.
Trying again with indirect rendering:
Checking for texture_from_pixmap: present.
Checking for non power of two support: present.
Checking for Composite extension: present.
Comparing resolution (1280x800) to maximum 3D texture size (2048): Passed.
Checking for Software Rasterizer: present.
Software rasterizer detected, abortingaborting and using fallback:
/usr/bin/metacity

--- Then several of the above errors ---

I checked his menu.lst, and heres the entry for the problematic kernel:

title		Ubuntu 8.10, kernel 2.6.27-11-generic
uuid		1ac63e0f-dff7-48f8-9506-ce783a5dd383
kernel		/boot/vmlinuz-2.6.27-11-generic
root=UUID=1ac63e0f-dff7-48f8-9506-ce783a5dd383 ro locale=pt_BR quiet
splash
initrd		/boot/initrd.img-2.6.27-11-generic
quiet

I noticed that weird root=UUID= on kernel, and changed it to root=/dev/sda2 UUID=**** but it didnt work

Well, this is all for now - thanks for reading Deividson

[ Thread continues here (12 messages/21.16kB) ]

---

Need for a 'viral' marketing campaign to introduce Linux (especially Ubuntu) but it doesn't matter which to Windows sufferers

Charles Norrie [cbnorrie at hotmail.com]

Mon, 2 Feb 2009 07:54:56 +0000

I am concerned that despite the Conficker affair, the message about using Ubuntu isn't getting through especially to industry commentators. So I wrote to Mark Shuttleworth.

Dear Mark Shuttleworth,

Marketing Ubuntu

I am still delighted with Ubuntu, and believe me after using computers for 40 years, it's quite the best OS/package I've ever come across.

But the message isn't getting out there. I became aware how bad it was when I did a single word Google News search for 'Conficker'.

You and I know that if only we could get people to put Linux on their pcs their malware problems would vanish. But that message is not getting through.

Here are the results:

[ ... ]

[ Thread continues here (3 messages/16.93kB) ]

---

Device driver program

Maneesh M P [maneesh.poyil at gmail.com]

Wed, 11 Feb 2009 00:55:35 +0530

[[[ How surprising that this was sent in .html, eh? -- Kat, rolling her eyes in irony. ]]]

Hello,

I have read the article of yours from https://linuxgazette.net/136/anonymous.html . Sounds interesting and it was very informative too. Thanks for sharing it.

I have a requirement, i need a device driver program which reads ctrl +k, this key sequence will send a SIGUSR1 signal to another application program which will just do a multiplication or sort of few numbers once its recieve the SIGUSR1 signal.

We have written the application program, which is pretty easy though. Now in need of a device driver program which reads ctrl + m and send SIGUSR1 signal.

I will be really thankful if you could help me in this.

- Regards, Maneesh(=E0=B4=AE=E0=B4=A8=E0=B5=80=E0=B4=B7=E0=B5=8D )

[ Thread continues here (9 messages/10.48kB) ]

---

Talkback: Discuss this article with The Answer Gang

Talkback

Talkback:155/moen.html

[ In reference to "Book Review: Understanding Open Source and Free Software Licensing" in LG#155 ]

Jimmy O'Regan [joregan at gmail.com]

Fri, 27 Feb 2009 00:03:59 +0000

2008/9/22 Rick Moen <rick@linuxmafia.com>:

> the licensed work.)  Is it possible to donate a work of original
> ownership directly to the public domain, despite the lack of any legal
> mechanism for doing so?  Is it desirable to have a choice-of-law

Probably not, but you can now dedicate a work to 'the Commons': https://creativecommons.org/licenses/zero/1.0/

'The person who associated a work with this document has dedicated
this work to the Commons by waiving all of his or her rights to the
work under copyright law and all related or neighboring legal rights
he or she had in the work, to the extent allowable by law.'

[ Thread continues here (2 messages/3.20kB) ]

---

Talkback:158/vogel.html

[ In reference to "Using Hyperestraier to search your stuff" in LG#158 ]

Karl Vogel [vogelke+unix at pobox.com]

Fri, 20 Feb 2009 20:23:02 -0500 (EST)

Greetings:

I ran into some problems while searching for portions of words in one of my indexes. If the word "workstation" was present, I wanted to be able to search for (say) "orksta" without getting 0 hits. I can use substrings like leading or trailing asterisks in a command-line search by adding the "-sf" option:

  me% estcmd search -sf -vu -max 40 $db/srch "$pattern"

I'd rather avoid the wildcard stuff when doing a browser search using estseek.cgi, and I found that adding a synonym list works really well. To build a wordlist for synonyms, run "estcmd words" on your search index:

  me% estcmd words srch | awk '{print $1}' | sort -u > /path/to/synonyms

Here are the changes to estseek.conf:

   1  # phraseform: specifies the phrase form.  "1" is usual form, "2"
   2  # is simplified form, "3" is rough form, "4" is union form, "5"
   3  # is intersection form.
   4  phraseform: 2
   5  
   6  # candetail: specifies whether to enable detail display of a
   7  # document.  "true" or "false".
   8  candetail: true
   9  
  10  # candir: specifies whether to enable directory display of a
  11  # document.  "true" or "false".
  12  candir: true
  13  
  14  # If you want query expansion, enable an outer command by editing
  15  # qxpndcmd in estseek.conf.  It specifies the absolute path of
  16  # a command which outputs synonyms of a word specified by the
  17  # environment variable "ESTWORD".
  18  qxpndcmd: /usr/local/share/hyperestraier/filter/myxpnd

Here's the expansion script "myxpnd":

   1  #!/bin/ksh
   2  # myxpand: list synonyms
   3  
   4  # set variables
   5  LANG=C ; export LANG
   6  LC_ALL=C ; export LC_ALL
   7  PATH="/usr/local/bin:/bin:/usr/bin" ; export PATH
   8  
   9  # show help message
  10  case "$1" in
  11      --help) echo 'List synonyms of a word'; exit 0 ;;
  12      *)      ;;
  13  esac
  14  
  15  # list synonyms
  16  exec fgrep "$ESTWORD" /path/to/synonyms
  17  exit 0

Here's part of a search form that works:

  <form method="get" action="estseek.cgi">
    <input type="text" name="phrase" value="" size="20" />
    <input type="hidden" name="phraseform" value="2" />
    <input type="hidden" name="perpage" value="10" />
    <input type="hidden" name="qxpnd" value="1" />
    <input type="hidden" name="navi" value="0" />
    <input type="submit" value="Search" />
  </form>

-- 
Karl Vogel                      I don't speak for the USAF or my company
 
WEDDING DRESS FOR SALE:  Worn once by mistake.  Call Stephanie.
                                                    --seen in the want-ads

---

Talkback:159/okopnik.html

[ In reference to "Hyperestraier Redux - A User-friendly Approach" in LG#159 ]

Ben Okopnik [ben at linuxgazette.net]

Tue, 3 Feb 2009 21:59:16 -0500

[ Karl, I hope you don't mind me copying this exchange to The Answer Gang; I'd like for the error that you pointed out to be noted in our next issue, and this is the best and easiest way to do it. If you have any further replies, please CC them to 'tag@lists.linuxgazette.net'. ]

On Tue, Feb 03, 2009 at 02:59:57PM -0500, Karl Vogel wrote:

> Very cool follow-up article!

Thanks, Karl; I appreciate that. That's a very, very fun program - again, thanks for introducing me to it!

[ ... ]

[ Thread continues here (4 messages/16.55kB) ]

---

Talkback:144/lg_mail.html

[ In reference to "Mailbag" in LG#144 ]

Kapil Hari Paranjape [kapil at imsc.res.in]

Sat, 21 Feb 2009 11:31:46 +0530

Dear TAG-ers,

I am enclosing a qeury received regarding #144.

Regards,

Kapil.

P.S. (to aditya) please do not mail TAG members directly. Use the mailing list address as above instead.

----- Forwarded message from Aditya Bhiday <aditya.bhiday@gmail.com> -----

Date: Sat, 21 Feb 2009 11:18:15 +0530
Subject: Regarding Proxy Tunneling (TLDP)
From: Aditya Bhiday <aditya.bhiday@gmail.com>
To: kapil@imsc.res.in

Hi,

I came across a post at https://tldp.org/LDP/LGNET/144/misc/lg/qu[...]om_being_used_as_a_socks_proxy.htmlwhich said that

"AllowTcpForwarding Specifies whether TCP forwarding is permitted. The default is "yes". Note that disabling TCP forwarding does not improve security unless users are also denied shell access, as they can always install their own forwarders."

I was just experimenting around with tunneling and as to how to block it. Please could explain to me how one can install their own forwarders if ssh tunneling is blocked, or the name of such a forwarding software?

Thanks,

Regards, Aditya Bhiday

----- End forwarded message -----

[ Thread continues here (6 messages/4.93kB) ]

---

Talkback:159/dokopnik.html

[ In reference to "Installing VMWare Server 2 on Ubuntu Server 8.10 (Intrepid Ibex)" in LG#159 ]

Stack, David [Stack at sbmc-law.com]

Fri, 13 Feb 2009 23:51:46 -0800

Hello

Thanks for the article. I'm attempting to make this work with the 64 bit version of Ubuntu server 8.10 and the 64bit version of vmware server 2.0. The install went fine but I cannot access the vmware console on the IP of my ubuntu server on port 8222? I know it's alive on my network as it accessed packages from the internet and I have a windows server file share mounted. Any Idea's what to check? Thanks

Dave Stack

-- 
   StackAble IT Solutions LLC
   For SBMC P.S.
   (509) 220-8517

[ Thread continues here (3 messages/2.74kB) ]

---

Talkback:115/orr.html

[ In reference to "WSGI Explorations in Python" in LG#115 ]

Mike Orr [sluggoster at gmail.com]

Tue, 10 Feb 2009 14:00:01 -0800

Here's a follow-up to my 2005 article "WSGI Explorations in Python" (https://linuxgazette.net/115/orr.html). Michael Will asked me what had happened since then, so I wrote this:

There is no state-of-the-Python-web overview that I know of, but a lot has happened since I wrote that article. Pretty much all new frameworks are written for WSGI, and the older ones have been retrofitted. (CherryPy can run as a WSGI server, Plone can run as an application, parts of Zope have been extracted to independent Repoze components, and Quixote has a WSGI gateway floating around somewhere.) Django works with WSGI sort of, and has been ported to Google App Engine via WSGI.

I'm involved with Pylons, a framework that's fully WSGI and modular to the core, built on top of Paste, which is a low-level WSGI library. TurboGears 2 is being built on top of Pylons. This means that different frameworks with different goals and target users can share the same technology, and essentially makes every TG developer a Pylons developer, doubling our developer base.

There's a group of WSGI framework developers including Pylons/TG/Repoze.BFG developers that is designing a new framework to potentially supercede all of them, with plug-in personalities to reflect their different application styles. This is still at the idea stage but may have some alpha code by the end of the year. If so it could point the way to the next generation of frameworks.

Another big issue is Python 3. Over the next year frameworks will either be ported to Python 3 or replaced by frameworks written for Python 3. (Though the Python 2 frameworks may continue in use for several years.) This has to be done on a dependency basis; e.g., Pylons can't upgrade until all the components it depends on have upgraded.

-- 
Mike Orr <sluggoster@gmail.com>

---

Talkback: Discuss this article with The Answer Gang

2-Cent Tips

2-cent Tip: Reading MHT files in Linux

Ben Okopnik [ben at linuxgazette.net]

Fri, 13 Feb 2009 20:08:48 -0500

Just ran across Yet Another Proprietary Format from Micr0s0ft: .mht files. Seems that Internet Explorer saves emails and HTML as an ugly mess that somewhat resembles an email; according to Wikipedia, there's no single standard, and the state of the state is best described as 'sauve qui peut' (which translates, at least in Redmond, as "all your ass are belong to me!") Bleh.

Searching the Web shows that there are a lot of people - the just-converted-to-Linux newbies, particularly - who have loads of these things and don't know what to do with them. Some people recommend Opera (I suppose a couple of hours of Kiri Te Kanawa is good for relieving all kind of stress...); some have had luck with various conversion utilities. I looked at it, and it looked something like a mangled email header, soooo...

I didn't go searching for more than just the one file that I had, but here's what worked fine for opening it:

# Convert line-ends to Unix format
flip -ub file.mht
# Prepend a standard 'From ' mail header to the file
sed -i '1i\'"$(echo From $USER $(date))" file.mht
# You should now be able to open it with your favorite MUA
mutt -f file.mht

It worked fine for me.

-- 
* Ben Okopnik * Editor-in-Chief, Linux Gazette * https://LinuxGazette.NET *

[ Thread continues here (13 messages/18.98kB) ]

---

2-cent Tip: Quick and Free reminders on your cellphone

Deividson Okopnik [deivid.okop at gmail.com]

Mon, 16 Feb 2009 14:07:49 -0300

Not very Linux related, but this is a nice tip anyway.

If you use a future mail page (like https://www.futureme.org/ ) in conjunct with a email to sms gateway (you can get a list here: https://www.mutube.com/projects/open-email-to-sms/gateway-list/ ), you can send reminders in a free, fast way, directly to your cellphone (and futureme doesnt even need you to signup for an account)

[ Thread continues here (2 messages/2.03kB) ]

---

2-cent Tip: Converting fonts to images

Ben Okopnik [ben at linuxgazette.net]

Sat, 14 Feb 2009 10:00:18 -0500

On Sat, Feb 14, 2009 at 12:40:15AM -0500, Matt Giwer wrote:

> 	I don't know this qualifies as one but I have used it quite a bit and I 
> have not seen it mentioned elsewhere. Nor have I found it in the docs. 
> This uses the convert function from ImageMagick.
>
> 	Why? Because many years ago I became addicted to collecting truetype 
> files. There are only a relative few significantly different ones. For 
> the majority they are the same files with a different names. Most of the 
> rest just don't look different enough to matter. I used this to find the 
> one or two gems per thousand.
>
> 	This works to produce a jpg, or any other image format, of the font. It 
> also works on otf files.

You may not be aware of this, but TrueType files can be viewed with the 'display' program - just as if it were another type of image.

> convert font_name.ttf font_name.ttf.jpg
>
> 	Generally a filed named ttf-convert
>
> #!/bin/bash
> while [ $# -gt 0 ];
> do
> convert $1 $1.jpg

This will, of course, fail if the file is called something like 'MS Verdana.ttf' (note the space.) It's always a good idea to double-quote variables used as arguments.

> shift;
> done

Um... wrong tool for the job, I'm afraid. A "while" loop is a state-tester (i.e., checks for true/false and loops); the "for" loop is an iterator, which walks ("iterates") over a list of arguments - which is the exact requirement here.

# Loop over the list of all command-line arguments
for f
do
	/usr/bin/convert "$f" "$f.jpg"
done

...but since it's really a one-liner, I would just tend to type things like

for f in *ttf; do convert "$f" "$f.jpg"; done

right on the command line. It's not much harder to remember than yet another script name.

-- 
* Ben Okopnik * Editor-in-Chief, Linux Gazette * https://LinuxGazette.NET *

---

2-cent Tip: Easy way to make security context of two files/directories exactly the same

Mulyadi Santosa [mulyadi.santosa at gmail.com]

Sun, 8 Feb 2009 00:23:52 +0700

If you're administering a Linux system with SELinux policy being enforced, you'll know that most of the times you need to set security context of certain files/directories so it can be properly accessed by daemons or users.

If the task is something like "I create new directory /var/apache/html to put HTML files for my new Apache virtual host", you should make security context of /var/apache/html exactly like /var/www/html (assuming it is the default DocumentRoot of Apache). This command does that for you:

# chcon --reference=/var/www/html /var/apache/html

For additional information, chmod and chown also employ the same option too. e.g:

# chmod --reference=/var/www/html /var/apache/html

makes the access permissions on both directories exactly the same.

regards,

Mulyadi.

---

2-cent Tip: What was that (masked) script's name?

Ben Okopnik [ben at linuxgazette.net]

Sat, 14 Feb 2009 10:56:06 -0500

Over the years, I've written hundreds of scripts to help me with various tasks. Many of them are still very useful - but whether my memory isn't what it used to be (hmm... I wonder if I can write a script to fix that...), or whether that's just too many things to keep in my head, I've often found myself searching and cursing when I couldn't find the script that I needed and knew I'd already written. Too many scripts!

To deal with this, I've started copying something that I already do elsewhere. When I teach, I give my students a group of scripts that I've made up for each class - and in each of these scripts, I insert an explanatory comment on line 3 (1 is the shebang, while 2 is a comment containing the author credit and a timestamp.) I try to come up with keywords that will be obvious even if I can't remember what I wrote originally. I also include a 'README' script which reports the third line of every file in the directory. There's even a method for ignoring files and adding your own descriptions (do try to avoid the latter, for obvious reasons.)

So, all you need to do is get in the habit of commenting your scripts in clear and obvious ways on the third line. Once you've done that, this little script will save you much time and frustration. Enjoy!

#!/bin/sh
# Created by Ben Okopnik on Sun Dec 10 08:05:39 EST 2005
# Run this script to see the descriptions of all files in this directory
 
reject='a.out regex.7 access.log'
printf "%-25s%s\n" "regex.7"    "# Henry Spencer's regular expressions"
printf "%-25s%s\n" "access.log" "# Typical Apache log file"
 
for n in *
do
    [ -f "$n" -a -z "`echo $reject|grep \"\<$n\>\"`" ] || continue
    printf "%-25s%s\n" "$n" "`/bin/sed -n '3p' \"$n\"`"
done

-- 
* Ben Okopnik * Editor-in-Chief, Linux Gazette * https://LinuxGazette.NET *

---

Talkback: Discuss this article with The Answer Gang

News Bytes

By Deividson Luiz Okopnik and Howard Dyckoff

Contents: News in General Conferences and Events Distro News Software and Product News

Selected and Edited by Deividson Okopnik

Please submit your News Bytes items in plain text; other formats may be rejected without reading. [You have been warned!] A one- or two-paragraph summary plus a URL has a much higher chance of being published than an entire press release. Submit items to bytes@linuxgazette.net.

---

Talkback: Discuss this article with The Answer Gang

---

Deividson was born in União da Vitória, PR, Brazil, on 14/04/1984. He became interested in computing when he was still a kid, and started to code when he was 12 years old. He is a graduate in Information Systems and is finishing his specialization in Networks and Web Development. He codes in several languages, including C/C++/C#, PHP, Visual Basic, Object Pascal and others.

Deividson works in Porto União's Town Hall as a Computer Technician, and specializes in Web and Desktop system development, and Database/Network Maintenance.

---

Howard Dyckoff is a long term IT professional with primary experience at Fortune 100 and 200 firms. Before his IT career, he worked for Aviation Week and Space Technology magazine and before that used to edit SkyCom, a newsletter for astronomers and rocketeers. He hails from the Republic of Brooklyn [and Polytechnic Institute] and now, after several trips to Himalayan mountain tops, resides in the SF Bay Area with a large book collection and several pet rocks.

Howard maintains the Technology-Events blog at blogspot.com from which he contributes the Events listing for Linux Gazette. Visit the blog to preview some of the next month's NewsBytes Events.

Away Mission - 2008 in Review - part 2

By Howard Dyckoff

March is conference madness month in the San Francisco Bay Area with week-to-week events that sometimes overlap. For the last few years, Open Source developers and entrepreneurs have had EclipseCon, Dr. Dobb's Software Development Conference (SD West), and the Open Source Business Conference to choose from. Starting in 2008, if that audience was interested in FOSS on mobile platforms, there was also the Emerging Communications Conference (eComm), which was reviewed in part 1 last month.

EclipseCon-08

This conference actually went up a little in value in 2008 even with a higher price because the tutorials were open to all conference attendees.

While EclipseCon is primarily about the Eclipse platform and the tools for its myriad projects, it is also about community and the larger Open Source ecosystem. There are sessions on creating and nurturing communities, on the future directions of Java, and on the new scripting languages that are supported on the Eclipse platform. There are presentations and panels on software deployment, modeling languages, and even SQL. In short, it's a really big tent. The tracks were well defined and taken all together were very comprehensive. Among the track topics were Eclipse as a Platform, OSGi, SOA, Modeling, Performance, Web Tools, and Emerging Technologies among others. However, some topics have sessions that are scattered across different rooms. The EclipseCon color codes help a bit with this to a degree, but with 8-10 simultaneous sessions, you could be in for a little exercise.

IBM held its own track on Tuesday, featuring its OpenUP Agile toolset and the Agile-oriented Jazz collaboration suite, both heavily based on Eclipse. In Jazz, and in Rational Team Concert, the team process is selected from process templates including Scrum, OpenUP, and the Eclipse Way (or EPF, the Eclipse Process Framework). You can also mix elements from these processes in custom configurations reflecting the actual work process at your own company or organization. Jazz uses Eclipse technology on both the client and the server, and Team Concert provides agile teams with an integrated tool set including source control, work item tracking and build management with continuous integration. The EclipseCon sessions are summarized here.

This is a very good set of collaboration tools, having been tested within Eclipse Projects for more than a year and also getting the benefit of IBM's own experience with the Rational RUP methodology, now available as OpenUP. Some of this effort is occurring via the Project Zero website at https://projectzero.org.

The presentations from the IBM Track are not in the EclipseCon archive for 2008, but here's a link to a representative Jazz web site at IBM Rational.

I also attended a great late session on Groovy, a dynamic scripting language that runs on JVMs. Groovy reduces the usual Java scaffolding for developing code and simplifies Java syntax. Semicolons are optional and it doesn't require getters and setters for each property; Groovy just does the right thing. Groovy also simplifes testing right out of box, working directly with JUnit. Groovy objects are called POGOs which are like Java POJOs on steroids.

POGOS write their own property accessors and provide a default constructor with named parameters. And like JRuby, Groovy compiles down to byte code. Check out the PPT-formatted presentation here: Introduction to Groovy.

EclipseCon offers lots of BOFs at night and several evening receptions with food and drinks. I especially recommend the annual Poster Session where new projects are presented by their developers with a pleasant dinner buffet. The Higgins credential framework, for example, was just such a poster presentation four years ago and is now widely used.

Each morning had Euro breakfasts: brioche, bagels, and pastry with coffee and tea. Sit-down lunches were the pattern again, with each table sporting signage for different Eclipse topics and projects (EMF,RCP, AJAX, etc.) There was plenty of choice here and lots of special meal categories were coordinated. And beside networking opportunities, the food was pretty good. The conference bag for 2008 was unfortunately just a canvas tote, but it did have a zipper closure. Previous years have had distinctive backpacks.

EclipseCon generally gives attendees USB thumb drives with about half of the conference presentations. The other presentations are available a day or two after their sessions and can be added at update stations. Some sessions show up weeks later. I personally think that is a problem for attendees planning their day with so many sessions overlapping. Moreover, many of the sessions on the USB drive get revised right before being presented. Consequently, the update sessions are very busy and are a bit of a bottle neck.

I also have to report a flaw in what is normally a very well organized mid-sized conference. The check-in process on the first tutorial day backed up badly and many attendees had to wait in line for 20-35 minutes. Clearly, they didn't have enough stations and personnel to check IDs and hand out badges and bags. But it was also asymmetric! They had 6 stations set up, each handling 4 letters of the alphabet, the last station also handling Y and Z. Last year, the A-D group was twice as long as any other and took the longer 35 minute interval. They should have gone through the list of registrants and set the letter groups up for roughly equal numbers of attendees. So plan on showing up at least 30 minutes before your tutorial if you want to attend EclipseCon this year.

Last year, EclipseCon was be held March 17-20, the week before the Open Source Business Conference (OSBC), but for 2009, EclipseCon is March 24-27, and overlaps OSBC.

EclipseCon 2008 materials can be found here: https://www.eclipsecon.org/2008/.

The Business of Open Source - OSBC

OSBC started a few years ago by discussing issues in OSS and legal ramifications. In prior years, a lot of energy went into software patents and the then-new GPLv3. But in 2008, with few dragons left to slay, it was mostly about successful adoption of OSS at the enterprise level. There were several presentations on best practices and several case studies. For example, Weather.com and Kaplan.com, the learning center company, revealed the works behind the curtains a bit.

There are four tracks that run consistently over the two days of the conference: one for CTOs, one for CFOs, one for lawyers or those interested in legal topics, and a product and service track which also doubles as a track for emerging trends. Each track has a dedicated room in close proximity to the other sessions. There is little exercise in changing tracks here.

The agenda page lists the four parallel sessions and also has links to the those presentations that are available online. Unfortunately, some of the better presentations are not included, such as the opening keynote.

A more concrete presentation on the merits of going Open Source showed how two developers moved the back end at the New York Times to FOSS. That led to indexing 11 million PDF files, over 1.5 TB of data, in 24 hours with Amazon Web Services and Hadoop for only $240. Now those files can be served with the existing infrastructure and no new hardware. That presentation is here: https://akamai.infoworld.com/event/osbc/08/docs/CIO-CTO-Gottfrid-Harris.pdf

OSBC takes place at the Palace Hotel on Market Street in San Francisco, near an exit of the Montgomery Street BART station and near several major bus lines. This is a great location and allows attendees to use public transportation easily. It is also a nice venue with a touch of Victorian ambiance. However, power for laptops is an issue. Some rooms have power strips near the front, but this is up to the organizers and was not true of all session rooms. There are no power strips in the keynote room. The moral here is to bring your own extension cord and maybe a multi-outlet adapter to encourage sharing. We can all get along, right?

OSBC has usually provided a full breakfast buffet in the mornings, which is taken away just before the keynotes start. So come fairly early to enjoy it and to get a good seat.

The Open Source Business Conference will be held March 24-25, 2009. Last year, OSBC followed EclipseCon, but in 2009, they will overlap.

For 2009, OSBC will have keynotes from Ron Hovsepian, CEO of Novell; Jim Whitehurst, CEO of Red Hat; Jonathan Schwartz, CEO of Sun Microsystems; and a mystery speaker from Microsoft (yes, Redmond has a presence at OSBC.) There will also be a keynote panel on the topic of "End Users and Linux: Do We Have a Participation Problem?" led by Jim Zemlin, Executive Director of the Linux Foundation.

SDWEST-08

Software Development West, or SD West, usually occurs in early March. Like EclipseCon, SD West is held at the Santa Clara Convention Center. But the perks and grub are more limited. The dates for 2009 are March 9-13. This year, SD West is co-located with the Mobile Connect conference.

The tracks here are expansive and cover most of the software development lifecycle. The SD conference organizers claim to provide a global view of the software industry and technological developments. There are separate tracks for Java, C++, .Net, Web services, Agile processes, and security, among others. New this year is the topical Cloud Computing and also a track on Parallelism in software.

Another nice feature is the range of vendor-sponsored free technical sessions. This year some of those sessions will be conducted by Coverity and Open Make.

Unfortunately for our readers, SD conference archives and presentations are only available to conference attendees. The parent company for SD and Dr. Dobbs - CMP - resells the content. So you have to go to SD West to get it.

I attended a lot of sessions on Agile process, modeling, and data architecture. Many of these sessions were led by Scott Ambler or other Agile leaders. I was generally satisfied with the sessions. I also attended sessions on Web services and XML servers and distributed Ruby applications which discussed SOAP4R, DRb and Rinda (which is a Ruby version of Linda, a primitive paradigm for parallel processing).

An informative session on the developments surrounding HTML 5 was led by Elliot Rusty Harold. Work around this emerging standard began accelerating in 2008, because Microsoft finally decided to support the new standards as the default behavior in IE-8. If they hadn't, it would have been a big stumbling block.

Another somewhat tongue-in-cheek session discussed "Anti-Patterns in Software Development".

The presenter, Rob Daigneau, was Director of Architecture at Monster.com and currently host of www.designpatternsfor.net (he will also appear again at this year's SD West, speaking on "Design Patterns for Domain Services"). Fortunately, he posts his presentations on that Web site and a 2007 version of the slides are available there as well.

Anti-patterns look like a good idea but backfire badly when applied. Here, the anti-patterns referred to are ones of human behavior in SW projects, based on roles and on personality types. Many of these behaviors impede the success of a SW project. Among the personalities discussed were Code Controllers, Perfectionists, Workaholics, and other behavior patterns for managers and leaders.

The underlying anti-patterns include a lack of trust in team mates, or optimizing too soon and too often, or - for Workaholics - too readily accepting a culture of burnout at the companies we work for.

Daigneau spoke about larger organizational factors: a Fight Club mentality that no one talks about with conflicts between groups or between tech leads and architects and even between top executives. Daigneau noted that "fish rot from the head down", meaning that real and sustained changes require acceptance and support from top executives, who ultimately have responsibility for lingering organizational problems.

What if the boss is a repeat offender? Daigneau says there are 3 choices: adapt and adopt your boss' point of view, getting your boss or bosses to change their practices, or to move on. That's the way the project crumbles!

SD West has a lot of traditions and one of those is its campy Developer Bowl, where teams from leading ISVs compete in a contest of technical and industry trivia. Over the years the questions and the answers have gotten pretty broad. Two years back, when Microsoft fielded a team, they appeared in costume: Darth Vader and two Imperial Storm Troopers.

There are two matched competitions and then the winners of each face off. The Developer Bowl in 2008 featured the return of Team Google, last year's winner. But first up was CodeGear vs. IBM, last year's second place winner. IBM passed them with ease to win the first set.

Second up was Google and Intel. Tim Matson of Intel introduced himself as a kayaking instructor after his more senior team mates had introduced themselves with variously 12 to 22 years of experience at Intel. The Googlers responded with a wry ploy of each claiming 42 years of work at Google, an indirect tribute to Douglas Adams of "Hitchhiker's Guide to the Universe" fame. Intel, initially behind, pulled ahead on the last question after a long tie.

In what was a close match with a tie score, IBM pulled ahead of Intel during overtime. Several of the last computer trivia questions stumped both teams, e.g.: "How are Dr. Dobbs and Kurt Cobain related?" (Answer, via Courtney Love who was a relation of Dr. Dobbs). IBM prevailed in the end after receiving the question about the number of "pockets" on an old punch card reader - easy trivia for old IBMers. (The answer was a non-metric '12', something typical of original IBM equipment. And for readers not old enough to remember, the size of an IBM punch card matched the size of a 1950's dollar since every card reader represented cash for IBM.)

[ Unfortunately, this is incorrect: the IBM card of the 1950s - actually designed in 1928 - was 7.375" by 3.25" in size, while the US dollar bill, the current size of which was set in 1929, is 6.14" by 2.61". The origin of this myth is relatively obvious: the original Hollerith card of 1887 was indeed the same size as the (large) US paper dollar of that time. Suggestions for the choice of size were that the inventor, H. Hollerith, felt that people would treat it with respect due to the similarity; that boxes for currency were cheaply available; and that equipment for handling media of that size was already being made for the US Treasury. However, there is no actual evidence or support for any of these, and they must remain urban legends. -- Ben ]

The link to find out more about SD West is https://www/sdexpo.com, which lists early-bird discounts and also has an embedded video player with highlights from 2008 at the bottom of the Web page. Join the March Madness for OSS if you dare!

Talkback: Discuss this article with The Answer Gang

---

Howard Dyckoff is a long term IT professional with primary experience at Fortune 100 and 200 firms. Before his IT career, he worked for Aviation Week and Space Technology magazine and before that used to edit SkyCom, a newsletter for astronomers and rocketeers. He hails from the Republic of Brooklyn [and Polytechnic Institute] and now, after several trips to Himalayan mountain tops, resides in the SF Bay Area with a large book collection and several pet rocks.

Howard maintains the Technology-Events blog at blogspot.com from which he contributes the Events listing for Linux Gazette. Visit the blog to preview some of the next month's NewsBytes Events.

The Unbearable Lightness of Desktops: IceWM and idesk

By Ben Okopnik

For the last couple of years, I've been running Ubuntu on my HP Pavillion zv6000 laptop (AMD Athlon™ 64 Processor, 374MB RAM, 40GB HD) - and have often found myself grousing at the grindingly slow performance, particularly when I wanted to run any heavy-duty applications. Until recently, I took it for granted that running anything like GIMP or even Firefox meant essentially constant swapping, huge latency, stuttering in my audio programs, skipped frames in video apps... in short, behavior that I would have expected out of some ancient decrepit piece of hardware rather than a 64-bit powerhouse that this laptop is supposed to be. Even though I got more or less used to it, there was always that bit of buried outrage that one of the friendliest Linux distros out there consigned me to a life of crawling while everyone else was sprinting. If a Windows user ever asked me "so, is Linux faster than Windows?", I'd have to point them to other people's machines - since my own was anything but an effective demonstrator. It was time, and past time, to put some effort into resolving the situation.

Sometimes, it takes a particular event to push us out of a rut; in my case, the trigger happened when I recently attended the Ubuntu Bug Jam at SCaLE (the Southern California Linux Expo.) The local Ubuntu team was comprised of a very nice group of helpful folks who helped me out with an odd question that I had regarding font viewers under Ubuntu, and who got me thinking of other problems that I wanted to address - including this one. My questions about speeding up my machine didn't produce any definitive answers, but the process did get me thinking in this direction - which was enough of a start. Thanks, guys!

Back to the Future, or Forward to the Past?

The more I thought about it, the more I recalled my old (previous to Ubuntu) setup: I was running Debian, with IceWM as the window manager and 'idesk' as an icon manager. The combination was highly flexible, very powerful, and gave me everything that I wanted in a GUI (interestingly enough, I found both GNOME and KDE lacking in several respects - even though they had much more functionality in other ways.)

Installing both of the above under Ubuntu was trivially simple:

sudo apt-get install icewm idesk

...which, of course, produced no visible result. Unsurprising, since GNOME was still running; fortunately, the GNOME display manager (GDM) provides us with a nice mechanism for changing WMs. All you have to do is restart X - whether by logging out or by hitting the magic 'Ctrl-Alt-X' key combo which kills the X server - and select 'Options/Select session'. This allows you to start X with a different window manager, and to make that change temporary or permanent. Initially, I used the "Just for this session" version, but once I had everything configured the way I wanted it, I made the change permanent.

Configuring IceWM

Like any properly-written application, IceWM will run just fine without any special configuration - but configuring it "kicks it up to the next level". From a very basic, boring desktop with minimal functionality, it can quickly become a sleek, beautiful, responsive environment for any kind of GUI operations that you can envision for a GUI manager. Now, I'll be honest: I've looked at the IceWM GUI configurator (icemc) and... was not impressed. No insult to the people who put it together, but I find that a plain text editor works better, at least for me. The syntax of the various toolbar and key configuration files is so simple and intuitive, especially if you copy and edit the sample files that come with IceWM, that there's no reason to add another layer of complexity between you and the results you want.

Some of the most basic things - e.g., choosing a theme or a focus method - are built right into the IceWM menus. As an example, in this image, I've replaced IceWM's basic greenish-gray background with a snappy-looking "Area51" theme just by choosing it from the menu (available either from the button in the lower left corner or right-clicking on the desktop):

IceWM: Area51 theme

I actually recommend choosing a theme for yourself before you start hacking the rest of the settings - this will create a ~/.icewm directory and place a 'theme' file in it containing the theme you've chosen. Since that's what we need to do anyway, this saves us a bit of work.

By default, IceWM will read the configuration files in /etc/X11/icewm. If you're going to tweak a lot of things, I'd suggest copying the files from there as well as from /usr/share/icewm and modifying them: the format is rather obvious in all cases, and it's a lot easier to change what already exists than to try to remember something new and unfamiliar. In practice, at least for me, this comes down to modifying only three files: 'preferences', 'toolbar', and 'keys'. The first one modifies the default preferences for how IceWM itself acts and runs, the second determines the mini-icons on the toolbar, and the third - in some ways the most useful - allows you to define key combinations which will execute the programs that you set for them. This last was, in fact, the functionality that I found either missing or badly restricted in other WMs. Again, since the default files are heavily commented, and the default state for each setting is shown, tweaking these is very easy. Let's start with ~/.icewm/preferences; please note that I've deleted all but the prefs that I want to be different from the default setting.

#  Support mouse wheel
UseMouseWheel=1 # 0/1

#  Clock Time format (strftime format string)
TimeFormat="%T"

#  Show APM/ACPI/Battery/Power status monitor on task bar
TaskBarShowAPMStatus=1 # 0/1

#  Beep when new mail arrives
TaskBarMailboxStatusBeepOnNewMail=1 # 0/1

#  Count messages in mailbox
TaskBarMailboxStatusCountMessages=1 # 0/1

#  Execute taskbar applet commands (like MailCommand, ClockCommand, ...) on single click
TaskBarLaunchOnSingleClick=0 # 0/1

#  Icon search path (colon separated)
IconPath="/home/ben/Pics/icons:/usr/share/icons:/usr/share/pixmaps"

#  Command to run on mailbox
MailCommand="/usr/local/bin/rmutt"

#  Command to shutdown the system
ShutdownCommand="/usr/bin/gksudo /sbin/halt"

#  Command to reboot the system
RebootCommand="/usr/bin/gksudo /sbin/reboot"

#  Network device to show status for
NetworkStatusDevice="ppp0 eth0 wlan0"

# Workspace names
WorkspaceNames=" 1 ", " 2 ", " 3 ", " 4 "

#  Show workspace switching buttons on task bar
TaskBarShowWorkspaces=0 # 0/1

Hopefully, all of the above is clear enough; the names of the settings are pretty self-explanatory, and the comments make the rest of it obvious.

One of the things I decided not to set here, although the capability exists, is the desktop background. Instead, as you will see later in this article, I've left that piece to be handled by idesk because I prefer the way that it handles backgrounds - and it's nice to have that kind of choices.

Next, I set up the toolbar file. By default, it shows only two mini-icons: xterm and Mozilla. Since the toolbar is always visible no matter what windows may cover the desktop, this is a good place to put your "must-have" apps - and that's what I did, reserving the desktop for things that are less immediate.

prog "BackFlip" "kmultiple_16x16.xpm" /usr/local/bin/bkgr
prog "Lock screen" "decrypted_16x16.xpm" /usr/bin/xlock -random
prog "Mix" "music_32ndnote_16x16.xpm" /usr/bin/tkmixer
prog "Firefox" "mozilla_16x16.xpm" /usr/bin/firefox-2
prog "JPilot" "pda_black_16x16.xpm" /usr/bin/jpilot
prog "Xcal" "spreadsheet_16x16.xpm" /usr/local/bin/Xcal
prog "Snap" "image_16x16.xpm" /usr/local/bin/snap
prog "Cell-up" "presence_online_16x16.xpm" /usr/local/bin/aircard-up

Again, we have a simple layout: the popup label for the icon, the icon image, and the program to run. Here, I made use of a number of scripts that I've written over the years as well as standard Linux programs: "bkgr" for changing the backgrounds at will, "Xcal" to display a convenient and simple year-long calendar that I got from a 2-cent tip here in LG, and so on. The icons that I used are located in my ~/Pics/icons directory - but IceWM knows about that because I set the "IconPath" variable in the 'preferences' file, above. No problem at all - except that I had to consult my wife, the resident graphics designer, about the choice of icons. Oh well - we all have our strengths...

Last, I set up the 'keys' file:

key "Alt+Ctrl+m" /usr/local/bin/rmutt
key "Alt+Ctrl+l" /usr/bin/xterm -e lynx
key "Alt+Ctrl+x" /usr/bin/xterm
key "Alt+Ctrl+g" /usr/local/bin/urlgrab
key "Alt+Ctrl+u" /usr/local/bin/urlsee
key "Alt+Ctrl+k" /usr/bin/xkill
key "Alt+Ctrl+a" /usr/local/bin/voladj -u
key "Alt+Ctrl+z" /usr/local/bin/voladj -d

Again, obvious syntax and lots of scripts - 'rmutt' to set up and fire off my Mutt email reader, 'urlgrab' to run "wget" on a currently-highlighted URL (via 'xclip'), 'urlsee' to browse a URL, and 'voladj' to adjust the audio volume on the system. Few other window managers make it this easy!

One of the really nice bits of running IceWM is the ability to test its configuration quickly. Most of the changes - as far as I can tell, everything except 'preferences' - happen immediately, as soon as you save the relevant file; testing the changes in 'preferences' simply requires selecting 'Window Managers/IceWM' on the menu, which restarts IceWM (a matter of two seconds or so.) Instant feedback is great - especially if you're used to having to restart GNOME or KDE, which takes a long time by comparison.

Configuring idesk

Now that IceWM is up and running, you may decide that you want desktop icons as well. Frankly, I see a number of arguments against doing so: given the combination of the menu button (as well as right-clicking anywhere on the desktop, which brings up the same menu), the toolbar icons, and the fact that you can run programs by clicking the mail icon, the network monitors, the CPU monitor, and the clock, and considering that the desktop itself will be covered 99% of the time, it seems like gilding the lily. However, if that's what you want to do, then I suggest sticking with an equally lightweight option: idesk. This is a simple but capable icon manager; its only weak point, in my opinion, is the lack of easy configuration for its icon files - so, as is often the case, I've made up the lack by writing a bit of code. This configurator is slightly clumsy with regard to setup, but works pretty well.

There are two parts to setting up idesk: you need to create a ~/.ideskrc file and the individual .lnk files in a directory called ~/.idesktop. Sample files for these can be copied from '/usr/share/idesk/default.lnk', but I'll make it even simpler than that (keep reading, and you'll see!) Everything that goes into those is well documented in /usr/share/idesk/README, but you don't actually need to go in there unless you run into problems.

To make all of the above easy, I used a combination of the 'thttpd' web server, a bit of CGI, and the Mozilla browser - along with some canned links. Here's all you need to do to make it work.

Setting up

First, make sure to install the 'thttpd' web server and the 'gnome-icon-theme' package. Next, copy the default .ideskrc file and create your .idesktop directory with a 'builder' directory under it:

sudo apt-get install thttpd gnome-icon-theme
cp /usr/share/idesk/dot.ideskrc ~/.ideskrc
mkdir -p ~/.idesktop/builder

Now, place the files from this tarball into the '~/.idesktop' directory. This will also create a symlink in '~/.idesktop/builder' called '24x24' that points to /usr/share/icons/gnome/24x24:

tar xvzf builder.tgz -C ~/.idesktop

At this point, you have the minimal setup necessary to run idesk. Start it from a terminal, watching for errors:

idesk

You'll probably see warnings when you do:

Idesk starting in :0.0
[idesk] Background's file not found.
[idesk] Background's source not found.

These are not critical, but they make a good reminder: you may want to add a background image to ~/.ideskrc at this point (all of the following should be added at the end of the 'table Config' section.) You could specify either a Background.File: /absolute/path/to/image/file.ext or a Background.Source: /directory/with/images for a randomly-chosen background that changes every so often (Background.Delay: 5 would set it to every 5 minutes; the default is 0 - i.e., never changes - which would just give you a random image once per session.) You might also consider setting Background.Mode: MODE, which will modify how the images are displayed; the valid modes are Stretch, Scale, Center, Fit, and Mirror.

Once idesk is up and running, double-click the 'Desktop' icon that appears on your desktop; this will start 'thttpd', bring up a Mozilla browser, and show you the following interface:

The Builder interface

Simply enter the caption (i.e., the label which will show under the icon), the name of the program - preferably including its path - and click on an icon. Builder will create the .lnk file based on your choices and show you the contents; you can then return to Builder to create another icon. When you're done, just close the browser; this will also kill the web server. To see your new icons, you'll need to restart idesk; this simply requires double-clicking any of its icons with both mouse buttons. To edit a link file manually, you need to double-right-click its icon; running the program, as always, takes a double-left-click.

The Wrap-up

In this article, I've tried to present a cohesive setup for anyone interested in converting their default Ubuntu setup to a lighter, faster, sleeker configuration. Obviously, I can't guarantee any specific gains if you do this, but on my machine, I've gone from a CPU that was always showing a load of 1 or even slightly more while completely idle to one that acts as a Linux machine should:

'top' output: load averages with IceWM

No longer is running Mozilla a concern, nor do I see much (if any) swapping when I run GIMP, etc. In fact, for the first time ever on this machine, I've been able to install and run 'openarena' (a video-heavy first-person shooter) and have it actually work - no lag! - instead of chugging slowly one frame at a time. Applications open with unbelievable speed, and even flipping open an xterm feels like neck-snapping acceleration. For this happy camper, IceWM and idesk provide a comfortable, competent, and above all fast graphical environment for Linux. Give it a try!

Talkback: Discuss this article with The Answer Gang

---

Ben is the Editor-in-Chief for Linux Gazette and a member of The Answer Gang.

Ben was born in Moscow, Russia in 1962. He became interested in electricity at the tender age of six, promptly demonstrated it by sticking a fork into a socket and starting a fire, and has been falling down technological mineshafts ever since. He has been working with computers since the Elder Days, when they had to be built by soldering parts onto printed circuit boards and programs had to fit into 4k of memory (the recurring nightmares have almost faded, actually.)

His subsequent experiences include creating software in more than two dozen languages, network and database maintenance during the approach of a hurricane, writing articles for publications ranging from sailing magazines to technological journals, and teaching on a variety of topics ranging from Soviet weaponry and IBM hardware repair to Solaris and Linux administration, engineering, and programming. He also has the distinction of setting up the first Linux-based public access network in St. Georges, Bermuda as well as one of the first large-scale Linux-based mail servers in St. Thomas, USVI.

After a seven-year Atlantic/Caribbean cruise under sail and passages up and down the East coast of the US, he is currently anchored in northern Florida. His consulting business presents him with a variety of challenges such as teaching professional advancement courses for Sun Microsystems and providing Open Source solutions for local companies.

His current set of hobbies includes flying, yoga, martial arts, motorcycles, writing, Roman history, and mangling playing with his Ubuntu-based home network, in which he is ably assisted by his wife and son; his Palm Pilot is crammed full of alarms, many of which contain exclamation points.

He has been working with Linux since 1997, and credits it with his complete loss of interest in waging nuclear warfare on parts of the Pacific Northwest.

Joey's Notes: Bash shell basics

By Joey Prestia

It is often said that the real power of Linux lies at the command line - and this is true. Unfortunately, many new Linux users start out trying Linux as a desktop operating system and get frustrated when they discover the frequent necessity of using the command line - an unfamiliar and thus uncomfortable environment. This usually happens when they need to install from a tarball (compileable source code in a tar.gz file) and discover that they have to type commands or navigate to a specific directory. To make matters worse, the directory-separator slash (/) is the wrong way if they're used to DOS.

Or perhaps they need to get video playback to work right - something that can still be quite challenging under Linux. These types of things can be quite difficult for new users because of numerous differences in the operating systems. My purpose here is to try to make the transition easier by explaining some basics that should help, whether you are trying to learn Linux formally or just starting out as a desktop user trying to break away from Microsoft.

In reality, Linux will become a whole lot easier if you get familiar with just a few key concepts. Among those things is the directory structure: for example, there's no more C:\ drive. Instead, Linux has the root of the file system located at '/', which is equivalent to C:\ in Windows. No matter how many drives there are in a Linux system, there will only be one filesystem tree - more drives just increase the available space. Linux does not use the concept of one directory tree for each drive like Windows does.

Learning the structure of this directory tree is one of the biggest hurdles for the new user. As an example, the part of the system containing your personal files is referred to as your home directory; its name is the same as your username, and it is located under the '/home' directory. There is a link to a great little tutorial at the end of this article that may take a couple of hours to go through, but it will definitely help you to get familiar with Linux.

Most commands in Linux are some short form of what the command does; e.g., the ls command stands for 'list', and cd being short for change directory. Just by learning a few simple commands you will be able to navigate your system effectively and perform many tasks, but be patient - it takes time. Linux has around 3000+ commands.

[ Actually, that statement is both true and misleading. Every program in Linux is a command, but you only need to know a tiny percentage of these to work on the command line effectively. In general, the standard Unix toolkit (e.g., this list of commands) is all that's necessary - and even experts are rarely completely familiar with all of these. In my opinion, knowing about 40 of these commands will suffice the average user 99.9% of the time. -- Ben ]

All of these have similarities in their syntax, and there are some basic concepts that are uniform across all commands. Perhaps the most important thing to remember is the man command; whenever you need to use some unfamiliar command, just type man command. This will pop up the manual page for that command, giving you information that you may need before you execute it.

Syntax

The syntax of a command refers to the grammar of the sequence that we enter on the command line. To verify the syntax for a specific command, a quick glimpse at the appropriate man page will usually clear things up; this information is found under the SYNOPSIS heading. For most commands, the syntax is 'command options argument'.

Command

The name of a program like ls or cat or grep.

Option

A letter or a word preceded by a dash for "short options" or double dashes for the "long form" of an option. -a is a short option and --all is a long option; both mean the same thing.

Argument

An argument may be a file name, a directory, a device specification like /dev/cdrom. It is usually tied to the option (e.g., '--file doc.txt') and is, in a sense, the object of the command.

Most people have problems deciding where to put a space in the command line. The best guideline is that there should be spaces after every command and between all options preceded by a minus. Arguments may or may not require preceding spaces. Again, the man page is often the best guide.

The Prompt

The first thing you'll see when working in the command shell is the shell prompt. Looking at it will normally tell you what user you are, your system hostname, and your location on the system.

[ This is not necessarily true on many other Unix systems, or even in some of the smaller Linux distros. Often, the prompt consists of nothing more than '$', '%', or '#' to show your user level and shell type; for more information, you may need to either configure the prompt or use some of the toolkit informational commands. -- Ben ]

Looking at the sample below, we can see that our username is 'tom'. This is followed by the @ symbol which separates the username from the system hostname; if you never named your system, it may have 'localhost' in this position. Following this, we have the '~' character which indicates that we are in this user's home directory. Finally, we have the '$' character; this tell us, in a general fashion, the level of permissions that we have as a user. This varies across different shells; for example, the C shells (csh, tcsh) will use '%' for a standard user, while the Bourne shell and its derivatives (sh, ksh, bash, zsh, and a number of others) will usually use '$'. The root user (the system's superuser) gets a special prompt of '#'; this lets the user know to be very, very cautious and precise in the commands they issue, since the root user can easily destroy or damage the system.

Regular User:

[tom@station1 ~]$            

Root User:

[root@station1 ~]#

Typing the command echo $SHELL tom finds out he is using Bash (a.k.a., the "Bourne Again Shell".) The 'echo' command in this example prints the value of the $SHELL variable to the screen. The Bash shell is normally the default shell on most modern Linux systems.

[tom@station1 ~]$ echo $SHELL
/bin/bash

If the current directory is not shown in the prompt, the first important command you should learn is pwd which is short for "print working directory". This will reveal the full path as we can see below.

 
[tom@station1 ~]$ pwd
/home/tom

Commands

Commands are either shell built-ins, compiled executable programs, or executable scripts located somewhere on your system. The shell is a command interpreter which reads certain environment variables (you can see these by typing 'env' at the command line) when you open a command shell. When you type a command, the command interpreter will search the directories listed in the PATH variable to find the program you've asked it to execute. If it is not a shell built-in, an alias to a command, a defined function, or a program in the defined path, the command interpreter returns the 'command not found' error.

[tom@station1 ~]$ print help me 
bash: print: command not found

There are many commands - maybe too many. To see how many you have available, just hit the 'tab' key twice in a terminal; this will show you a number and ask if you actually want to see all of them. Don't worry: you'll never need to know most of them (many of these are invoked by other commands or via GUI menus.) No one can remember them all - there wouldn't be any point to it in the first place - and learning just a handful will help you out greatly on any Linux system.

Commands are case sensitive and are normally lower case. Another important thing to remember is that many commands (e.g., copying, moving, attaching a device, etc.) work in a "what to where" fashion:

cp /my/source/file /destination/file/or/directory
mv /from/here /to/there
mount /dev/mydevice /directory/to/mount/in

Switches and Options

Switches and options are used to modify the action of the command. They are usually preceded immediately by either a single or double dash, and are in the form of letters or full words. Some commands will require options and some will not. Here are some samples:

Simple command:

[tom@station1 ~]$ ls

Short option example:

[tom@station1 ~]$ ls -a

Multiple short options:

[tom@station1 ~]$ ls  -a -h -i

Multiple short options can be grouped:

[tom@station1 ~]$ ls -ahi

Long option example:

[tom@station1 ~]$ ls --all 

Multiple long options:

[tom@station1 ~]$ ls --all --human-readable --inode

Normally, short options can be grouped together if more than one is required. In addition, some short options have equivalent long options which perform the same action.

Arguments

An argument is usually an object that the command operates on; whether a command takes an argument at all depends on the command. For example, the command ls can take an argument, but if none is supplied the command assumes the current working directory is the argument and lists the current directory. An example of a command that takes no arguments is 'true': it takes only options. Again, arguments vary depending on the command and a great many commands have a default argument.

Getting Help Using the Man Pages

When using the command line, it is advantageous to get in the habit of using Linux's built-in manual page system, also called the "man" pages. At first glance, the man pages can look intimidating and very cryptic. Be assured, though - they are very helpful and necessary. By using them you will greatly enhance the speed at which you are learning Linux. To see the man pages for a command, merely type 'man command' and a screen with information will pop up. To navigate a man page, you use the arrow keys and page-up/page-down keys to scroll, and the letter "q" to quit. There are a number of other navigation keys for searching, jumping to the top or the bottom, and so on; these depend on the pager program (usually 'more' or 'less', with the latter being most common.) You can read about these by hitting the 'h' key (help) while browsing the man page.

When you look at any man page, you will notice several section headings:

NAME:

The name of the command and often similar commands that may be of use.

SYNOPSIS:

This section will give you the detailed syntax for running the command correctly. You may have to review the OPTIONS section, below, to get the precise details. There are many commands that must have options supplied to work at all.

DESCRIPTION:

Usually a brief to detailed technical description of what the command does.

OPTIONS:

Under this heading, there will be a listing of the long and short options that the command accepts. You will almost always need to use some sort of option with a command, so this section is quite important. This section can be fairly long. Since most commands use different meanings for their options, be aware that whatever options you used on one command may mean something completely different when used with some other command.

EXAMPLES:

When using unfamiliar commands I almost always look for this section so I don't foul something up. This section, if it exists, may further clarify up the syntax of the command being researched.

There may be more categories for some commands and less for others. This depends on the commands.

Learning Linux can be a long process but you will most definitely find it fun and rewarding as you discover the awesome power that it contains.

Additional Help

The Linux system was originally designed for the command line user, and if you know how to find the help that you need, you will have access to far more power than the average desktop user can even imagine. Another helpful source of information, if it is installed on your machine, is the 'info' command; this is the Texinfo documentation system, which contains the official documentation of the GNU Project. Some systems have both 'man' and 'info' available. It is a good idea to check both and see if one offers more information that is to your liking. Generally info pages tend to be easier to understand then the man pages, although in some instances the information is identical. To use the info system, type info command or just info to get familiar with it.

Tab Completion

Another helpful feature of the Linux command line is tab command completion and path completion. If you are not sure of the correct complete spelling of a command or a file name you can just type a couple of the initial letters of that command and hit the 'tab' key to see what the Bash shell comes up with for assistance. If the command specified by what you've typed so far is unique, it will complete the task for you by filling in the spelling; if not, it will present you with a list of possible choices.

Assuming, for example, that you need a command that can help you configure your sound card but only remember that it starts with "syst". Just type

 
[tom@station1 ~]$ syst

Then hit tab (you may need to hit it twice depending on your configuration) which should reward you with the following output:

 
system-cdinstall-helper          system-config-rootpassword
system-config-authentication     system-config-samba
system-config-date               system-config-securitylevel
system-config-display            system-config-securitylevel-tui
system-config-kdump              system-config-selinux
system-config-keyboard           system-config-services
system-config-language           system-config-soundcard
system-config-lvm                system-config-time
system-config-network            system-config-users
system-config-network-cmd        system-control-network
system-config-packages           system-install-packages
system-config-printer            systool

From this, you can see the name of the command you need: system-config-soundcard. This also works with paths to files. To use this feature and get the feel of how it works, start by typing a directory path - but after typing a couple of characters hit the tab key and see if it auto-completes for you. If it does not, it's because there is not enough characters to differentiate it from another possible choice. You can try adding another character or couple of characters and hitting the tab key again; this will allow it to complete for you. This can actually save you a lot of typing: 'cat /usr/include/c++/4.0.4/i486-linux-gnu/bits/messages_members.h' becomes 'cat /u⇒i⇒c+⇒⇒0.⇒i⇒b⇒m⇒' - 22 characters instead of 66!

Command History

Using the arrow keys also can make things easier. In Bash, these are mapped to what is called the command history: using the up arrow will take you to the previous executed command, and hitting the up arrow twice will take you back two commands. Of course, the down arrow does the opposite and moves you forward in the command history until no more commands are left. To see everything in the command history type history on the command line; recalling a previous command anywhere in the list can be done with 'Ctrl-R' and typing a part of that command. A command can also be recalled by noting its number in the 'history' output and typing an exclamation point followed by that number. There are many more shortcuts available; you can read about these in the 'HISTORY' section of 'man bash'.

Some basic commands and common switches

Command	Description	Common Options	Description
pwd	Print working directory	--help	Display help information
man	Display manual page	[1-8]	Man page section number
		-a	All relevant information
		-k	Keyword
cd	Change directory	..	Up one level
		/	Root of filesystem
		~ or nothing	Current users home directory
ls	List files	-a --all	All even hidden files
		-l	Long listing
		-h --human-readable	Human readable sizes
cp	Copy files	-i --interactive	Prompt before overwrite
		-r --recursive	Include underlying directories
		-a --archive	Archive mode all files
mv	Move files	-i --interactive	prompt before overwrite
		-u --update	move if file is newer
		-f --force	Don't prompt before overwriting
rm	Remove files	-r --recursive	Include underlying directories
		-f --force	Don't prompt before removal
		-i --interactive	Prompt before removal
mkdir	Make directory	-p --parents	Make parent directories
		-m --mode=MODE	Set file mode
		-v --verbose	Print confirmation
rmdir	Remove directory	-p --parents	Remove directory and parents
		-v --verbose	Print confirmation
		--ignore-fail-on-non-empty	Ignore failure if files exist
cat	Concatenate files	-b --number-nonblank	Number the non blank lines
		-n --number	Number all lines
		-s --squeeze-blank	Only show one blank line
locate	Locate files	-c --count	Count matches instead
		-i --ignore-case	Be case insensitive
		-S --statistics	Locate database statistics
less	Show one page at a time	-N --LINE-NUMBERS	Print line numbers
		-s --squeeze-blank-lines	Display only one blank line
		-E --QUIT-AT-EOF	Make less stop at the last page
grep	Print lines matching a pattern	-i --ignore-case	Ignore case sensitivity
		-R -r --recursive	Read all files under directory
		-v --invert-match	Select non-matching lines
mount	Mount a device	-o	Specify additional options
		-t	Manually specify file system type
		-a	Mount all file systems in fstab
umount	Unmount a device	-t	Specify file system type
		-v	Verbose mode
		-r	On failure remount read olny

Resources

• Learning the Shell Tutorial
• The Filesystem Hierachy Standard

Talkback: Discuss this article with The Answer Gang

---

Joey was born in Phoenix and started programming at the age fourteen on a Timex Sinclair 1000. He was driven by hopes he might be able to do something with this early model computer. He soon became proficient in the BASIC and Assembly programming languages. Joey became a programmer in 1990 and added COBOL, Fortran, and Pascal to his repertoire of programming languages. Since then has become obsessed with just about every aspect of computer science. He became enlightened and discovered RedHat Linux in 2002 when someone gave him RedHat version six. This started off a new passion centered around Linux. Currently Joey is completing his degree in Linux Networking and working on campus for the college's RedHat Academy in Arizona. He is also on the staff of the Linux Gazette as the Mirror Coordinator.

SCaLE 7 Speed-through

By Kat Tanaka Okopnik

Now in its seventh year, it's hard to believe the Southern California Linux Expo (SCaLE) started off as a grouping of Southern California LUGfests. Regular readers of LG may remember that Ben and I first attended two years ago, mentioned in a series of writeups, in issue 136. Our now 18-month old son, Michael, attended for the second time as well, while his sibling-to-come upheld family tradition by attending his or her first Linux convention in utero. Some of the big names in Open Source attend this convention, and the organizers have really applied every lesson they've learned along the way to make this a smoothly-run success.

In contrast to the long lines reported at SCaLE 5, the organizers switched to self-check-in, which cut delays substantially. The check-in printout included a raffle ticket for the Saturday evening reception and a personalized Web address for speaker surveys. The convention schedule was well-designed and easy to read. Signs made what could have been "hard to find" rooms merely a long walk.

There were signs of economic slowdown everywhere. Freebies were barely in evidence, and most of those were inexpensive - distro CDs, pens, and stickers. Some of the very few memorable ones were the ubiquitous flashing FreeBSD felt devil horns on a headband, O'Reilly's selection of buttons, and the ZenOSS squishy zebra. Michael's favorite was the collapsible frisbee in a pouch from Bytware; his Mommy and Daddy really appreciated that, too, as one of the best child-safe and attention-holding gimmes ever! The convention itself gave away eco-friendly lightweight canvas totes instead of T-shirts - a smart move away from the usual plastic or paper bags.

The free WiFi available in many parts of the convention was maddeningly slow. (Another unfortunate victim of budgetary constraints - truly adequate bandwidth would have been prohibitively expensive, and efforts at more creative solutions were stymied by lack of line-of-sight to appropriate relay stations). The Westin itself is a T-mobile hotspot, so anyone inclined to pay was probably well-served.

Friday started the weekend off with a full schedule:

What was once the "medical track" at SCaLE has become its own one-day conference (Demonstrating Open-Source Healthcare Solutions - DOHCS). Ben and I didn't sit in on any seminars, but it looked like it was well-attended and well-regarded by the professionals it was designed for.

Fedora and ZenOSS had free-form tracks on Friday, respectively as "Activity Day" and "Community Day", while Subversion opted for a series of organized panels for their Community Day track. Sun had Solaris workstations set up in the room that became the "Try It Lab" for the rest of the weekend; this was a hands-on area with multiple workstations and a facilitator for trying out everything from Ruby to Moodle to OpenOffice - a very nice opportunity for anyone who wanted to get a little actual experience.

I stopped by one of the last panels in the Women in Open Source track on Friday, after ignoring the room with the mysterious label "WIOS". I wish they had opted for the complete expanded title instead. The OSSIE (Open Source in Education) track was also very well attended - they managed to fill an auditorium. This is the track that Ben regrets missing the most, since education has always been near and dear to his heart; perhaps we'll catch it in a future SCaLE.

Other multi-slot programming included BSD certification exams, the Ubuntu Bug Jam, and the League of Professional System Administrators (LOPSA) running four half-day classes for sysadmins as SCaLE University. The convention had its usual excellent spread of classes for all experience levels from utter beginner to pro, with some big name speakers at all levels. Ben and I didn't get to as many panels as we'd have liked, but we had a great time over all. We did a lot of tag-team babyminding, and some fellow attendees (and some staffers) entertained Michael as well. Many of the presenters shared their information at the SCaLE Web site, and I will definitely be following up on the ones I wish I'd seen. I wish I'd gotten to participate in "Growing up Free", and hope that SCaLE expands that concept along with the ubiquitous OLPC machines to a parent/child track (especially if they incorporated childcare or child-friendly panels).

The "Weakest Geek Theatre" (actually, a panel-style game show with contestants voting each other off for bad answers) was a lot of fun. Topics ranged from the obvious to the obscure, and Randal Schwartz was on hand to act as arbiter for the Perl questions. Contestants started to use "Al Gore!" (who, as we all know, didn't invent the Internet), as their favorite way of saying, "bleeped if I know the answer to that!", and I'm highly tempted take that on, myself. Midway through, someone announced the availability of FREE! beer in the back. Cue stampede for free beer.

As far as the venue itself, the Westin LAX's elevators were horribly slow, but that, of course, is out of SCaLE's control - and there were signs promising faster service (after renovation) coming soon. The front desk staff was extremely gracious and helpful.

We met some LG alums - sorry to say I don't remember your names, having been slightly distracted by the energetic explorations of Michael throughout the weekend. Please do remind me! We learned about some great open source software - a few games we had overlooked until now, and some projects we hope to showcase in the coming issues.

For anyone at any level of Linux expertise, SCaLE is a great "bang for your buck" gathering and one we look forward to attending again.

Talkback: Discuss this article with The Answer Gang

---

Kat likes to tell people she's one of the youngest people to have learned to program using punchcards on a mainframe (back in '83); but the truth is that since then, despite many hours in front of various computer screens, she's a computer user rather than a computer programmer.

Her transition away from other OSes started with the design of a massively multilingual wedding invitation.

When away from the keyboard, her hands have been found wielding of knitting needles, various pens, henna, red-hot welding tools, upholsterer's shears, and a pneumatic scaler. More often these days, she's occupied with managing her latest project.

Copyright © 2009, Kat Tanaka Okopnik. Released under the Open Publication License unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 160 of Linux Gazette, March 2009

        /usr                    # the mountpoint for your unionfs
        /usr_stable             # your distro's stable branch (formerly known as /usr)
        /mnt/devel/usr          # your development branch

            $ mv /mnt/hda1/usr /mnt/hda1/usr_stable

            $ mkdir /mnt/hda1/usr
            $ cd /mnt/hda1/usr
            $ ln -sn -t . ../usr_stable/*
            $ touch YOU_SHOULD_NEVER_SEE_THIS_FILE

            $ echo "/usr_stable  /usr none  bind" >> /mnt/hda1/etc/fstab

            $ grep -Hw -e mount -e fstab /etc/init.d/*

    --- /etc/init.d/localmount.orgy 2009-02-15 10:26:22.000000000 +0100
    +++ /etc/init.d/localmount      2009-02-15 10:33:02.000000000 +0100
    @@ -23,6 +23,9 @@ start()
            mount -at "${types}"
            eend $? "Some local filesystem failed to mount"

    +       # conditionally mount development branch through unionfs
    +       /etc/init.d/unionmount-usr
    +
            # Always return 0 - some local mounts may not be critical for boot
            return 0
     }

    #!/bin/bash
    # check the kernel command line (/proc/cmdline) for the id-String and if given release the bind mount on /usr.
    # and put the union mount in its place.
    UNION_WANTED_FLAG="UniteWithDevel"      # this is what would be given on the kernel cmdline
    MOUNTPOINT="/usr"
    STABLE_BRANCH="/usr_stable"
    DEVEL_BRANCH="/mnt/devel/usr"
    ERROR_LOG_FILE="/var/log/develUnionFailed.log"

    errorOut() {
        MSG="ERROR ${0} $(date) : ${@}"
        echo "${MSG}" >&2
        echo "${MSG}" >> "${ERROR_LOG_FILE}" 
        exit 88
    }

    wantUnionMount() {
        grep "${UNION_WANTED_FLAG}" /proc/cmdline &>/dev/null
    }

    wantUnionMount && {
        echo "trying to unite..."
        umount "${MOUNTPOINT}" || errorOut "umounting failed"
        mount -t unionfs -o "dirs=${DEVEL_BRANCH}=rw:${STABLE_BRANCH}=rw" none ${MOUNTPOINT} || errorOut "mounting unionfs failed"
    }

        kernel /boot/unionkern-2.6.27.10/unionkern-2.6.27.10 root=/dev/hdc9 vga=0x31B UniteWithDevel 

        $ ebuild /usr/portage/x11-base/xorg-server/xorg-server-1.5.3-r2.ebuild compile
            #... now modify the sources like so:
            $ cd /var/tmp/portage/x11-base/xorg-server-1.5.3-r2/work/xorg-server-1.5.3
            $ sed -i "s/\"(II)\"/\"(I am Bob and I almost completely rewrote the xserver)\"/g" os/log.c
            #... and restart the build process
            $ make
        $ ebuild /usr/portage/x11-base/xorg-server/xorg-server-1.5.3-r2.ebuild install
        $ cp -a /var/tmp/portage/x11-base/xorg-server-1.5.3-r2/image/usr/* /mnt/devel/usr/

    --- kaffeine/src/player-parts/xine-part/xine_part.cpp.orgy      2008-11-10 19:24:18.000000000 +0100
    +++ kaffeine/src/player-parts/xine-part/xine_part.cpp   2008-11-10 19:24:23.000000000 +0100
    @@ -511,7 +511,7 @@ void XinePart::slotPosPlusSmall()

     void XinePart::slotPosMinusSmall()
     {
    -       slotJumpIncrement( -20 );
    +       slotJumpIncrement( -5 );
     }

     void XinePart::slotPosPlusMedium()

    DEVEL_BRANCH=$(readlink -f "${DEVEL_BRANCH}") 

From: Jubouri Omar <solveig.skold2@comhem.se>
Date: Wed, Feb 4, 2009 at 6:43 AM

To: undisclosed-recipients

-- 
Mike Orr <sluggoster@gmail.com>

-- 
Kat Tanaka Okopnik
Linux Gazette Mailbag Editor
kat@linuxgazette.net