Into.the.Void.

Following a post found on Digg about downloading the Enterprise version of Acrobat Reader I decided to install that on my Gentoo box.

First step was to download the tar.gz Linux version of Acrobat Reader Enterprise edition. Then extract it someplace. You will also need to extract the 2 new tars inside that directory as well. Now login as root (su -, probably) and make a backup of your old /opt/Acrobat7. Now go to the directory where you had the previous tars extracted and find a directory called Reader. Replace all files inside /opt/Acrobat7/Reader with the new ones. Do the same with the /opt/Acrobat7/Resource dir as well.

Now, if you don’t have openldap support in your box, go to /opt/Acrobat7/Reader/intellinux/plug_ins dir and do a chmod -x PPKLite.api. That will save you from an error message when acroread starts.

That’s it. Say bye bye to Yahoo Toolbar…

I am not responsible if following what I say here makes you do illegal stuff , violate Licenses or burn down your linux box.

Filed by kargig at 16:18 under Linux
No Comments | 4,730 views

Rox-filler is an excellent(for me) file manager that I use under my favorite wm…fluxbox (of course).

The latest emerge world broke my rox-filler. Rox-filler 2.2.0 has a known bug that makes it crash with gtk+ versions 2.8.11, 2.8.12, etc. Until now I had gtk+ 2.8.8 version and the latest emerge world updated it to 2.8.12 which had became “stable” in portage. Unfortunately rox-base/rox inside portage is not up to date with the latest rox-filler version (2.4.X) that fixes the bug…so it just crashes.
There is an open bug about it inside gentoo’s bugzilla: http://bugs.gentoo.org/show_bug.cgi?id=123675

For now I downgraded gtk+ to 2.8.8. I might create a custom ebuild for the latest rox-filler, as someone in bugzilla says, to circumvent the problem..but I am not really fond of that solution. I’d prefer if the rox-base/rox maintainer created a new ebuild….. sooner or later.

Filed by kargig at 21:53 under Linux
No Comments | 2,748 views

I had for over 2 months a friend’s left over laptop at my house. It’s an old amd K6 333Mhz with 64Mb RAM. Useless ? maybe yes maybe not…I just wanted to see what I coulld do with it…

I’ve booted it with DSL and I had a nice desktop to work on…now what ? A few days ago I had started preparing a new linux based access point for our local wireless network in the city I study. It had 2 Winstron CM9 802.11a/b/g cards and one netgear MA311 802.11.b card. I also happen to have an orinoco gold 802.11a/b/g pcmcia card I was given some time ago as a gift (even though I don’t own a laptop, yet). So why not mess with these two ? I plugged in the pcmcia in the laptop and it was instantly recognized. Then I just added through ifconfig the IP that I wanted the wireless card to have…and wow! I was associated instantly with the remote linux based AP in the next room. That was easy 🙂 It’s probably a good time now to install dhcp to that AP too so I won’t even need to add an IP manually 🙂

So all I had to do was give an IP…no drivers…nothing. The card was automatically recognized, and here’s the lsmod part about it:

root@ttyp0[root]# lsmod | grep ath
ath_pci 31508 1
wlan 48520 1 [ath_pci]
ath_hal 111792 1 [ath_pci]


By the way…this card is NOT recognized in windows XP without drivers. And if you happen to forgot the drivers cd while travelling…you are probably doomed.

Good job linux people 🙂

Filed by kargig at 22:44 under Linux,Networking
No Comments | 3,001 views

I had some problems making my new Corsair Flash Voyager 1Gb work under udev. Until now I used the SYSFS{serial}= part from udevinfo to distinguish my usb devices. Corsair Flash Voyager doesn’t have one. So I put the following line in /etc/udev/rules.d/10-local.rules to make it work:


KERNEL="sd*", SYSFS{vendor}="Corsair", SYSFS{model}="Flash*Voyager*", NAME{all_partitions}="corsair%n", GROUP="disk"


The first partition of the usb stick is now recognized as /dev/corsair1, and if I had more they would be named /dev/corsair2, /dev/corsair3, etc

Filed by kargig at 15:50 under General,Linux
No Comments | 4,527 views

We all have nowdays pile of cds and dvds with thousands of files inside. The big problem is…you know that you have the files somewhere…but where ? In which cd/dvd ?
A solution to this is a cataloging application like Disksearch. It is very simple, has an easy to use gui, it’s written in python and can be used in both Linux and Windows. It even has regular expression support for searching files.
I’m giving it a try and I hope it won’t let me down.

Just a thought:
sqlite support wouldn’t be bad, it’s sometimes more usefull to have an sql database than a simple index text file…but then you might lose compatibility with windows. A patch could be handy though. Any programmers to look into it ? 🙂

Filed by kargig at 04:46 under General,Linux
5 Comments | 7,943 views

I own a netroute2 and I have an asterisk at home to serve me as a pbx. I use it primarily for testing, and I only have a working sipdiscount trunk on it so far. What I wanted to do was plug a normal phone on the netroute and set it up so that I can call land lines in Greece for free through asterisk, using a sip trunk with sipdiscount.

add to your /etc/asterisk/sip.conf

[sipdiscount]
type=peer
host=sip1.sipdiscount.com
dtmfmode=inband
;allow-g726
canreinvite=no
fromdomain=stun.sipdiscount.com
username=USERNAME
fromuser=USERNAME
secret=PASSWORD
;qualify=yes
[5000]
; netroute
type=friend
regexten=5000
callerid="netroute2" <5000>
host=dynamic ; This device needs to register
secret=PASSWORD
allow=ulaw
allow=alaw


add to your /etc/asterisk/extensions.conf

exten => 5000,1,Dial(SIP/5000,30,rm)
exten => _7.,1,Dial(SIP/${EXTEN:1}@sipdiscount,30,rm)
exten => _7.,2,Congestion
exten => _7.,3,Busy


Netroute is more complicated than asterisk. Why ? because there is no documentation at all. So it’s a trial and error kind of thing. My current working solution IS very very crude and NOT thoroughly tested.

Edit your /etc/call_route.conf and comment out the line that says:
plugin load gr

Also, make sure your last lines of this file look like these:
# default routes
route_pattern add default/@ gr filt_last rl_rg_sip0
# note: default route is required
route_pattern add default/! default filt_last rl_rg_sip0


Now go to the web interface of netroute and in the voip section add asterisk’s sip domain/realm and IP settings and the settings you entered in the section [5000] of asterisk’s sip.conf. You might even want to lower the registration interval, if netroute and asterisk is in the same lan the extra traffic is insignificant. Then go to dialplan configuration and click on “+Add pattern”. In the Pattern field add 7X and don’t tick any of the other boxes. In the Prefix field just add 7.
When finished, do a
/etc/init.d/checkpoint
/etc/init.d/rc-voip restart

Now pick up a handset plugged into the first netroute’s FXS port and dial a string such as: 700302101234567
7 in the beggining is used by asterisk to send the call to sipdiscount trunk and 0030 is used by sipdiscount to call Greece. So in order to call 2101234567 we need to dial 700302101234567 from the netroute.
If you just want to dial Greek land line though sipdiscount change

exten => _7.,1,Dial(SIP/${EXTEN:1}@sipdiscount,30,rm)

with

exten => _7.,1,Dial(SIP/0030${EXTEN:1}@sipdiscount,30,rm)


Then you would only need to call 72101234567 from netroute.

Check asterisk’s log file for errors. A successful call should look like this:
Executing Dial("SIP/5000-49ab", "SIP/00302101234567@sipdiscount|30|rm") in new stack

🙂

Filed by kargig at 16:34 under Linux,VoIP
2 Comments | 6,239 views

Before I begin let me state my opinion. I am a big fan of Gentoo. I have installed gentoo more times and in more machines than I had installed win98 🙂

I am currently making an effort, oh and I mean it..it IS a effort, to install and configure a debian system. I want to use primarily as an Access Point/Client with 3 wireless cards. Sounds hard ? maybe…but debian makes it harder.
First of all…the damn development branches. Debian has stable, testing and unstable branches. I understand what stable is…but why should a decent distro have more than one branches in testing mode? The packages in stable branch are older than my grandmother. So most people actually use…the testing branch, and everyone on the net says it is “stable” enough to use it. So it’s more like debian has the stable-old-like-my-grandma, the testing-people-say-I-am-stable-enough and the unstable-don’t-use-me-I-will-crash your box branches. Gentoo just gives you arch and ~arch. Stable and unstable. Nice and easy.
Next…aptitude interface sucks. It just sucks. It so not user friendly…you actually get lost if you don’t learn all the keyboard shortcuts by heart. Yes, Gentoo’s emerge is a bit complicated too, but at least it is sane enough not to provide a crappy ncurses based gui. It’s better to stay command line only than having an awful gui. Maybe when you only have a few packages to update it might be a bit easier to distinguish what’s going on…but having more than 20-30 packages to update, like when you have just finished a fresh install and want to update,then it’s more like a killer task. It’s almost impossible to identify a blocking package unless you know the keyboard shortcuts. But then again…if you are experienced enough…why do you need a gui since everyone knows tasks made by guis are done slower ? Aptitude’s ncurses gui needs a drastic redesign if it ever wants to become user friendly and not a tool for the lazy admins that are just fond of blue and purple colors.
What I liked ? Updates did not hurt any of my conf files, yet. That’s where gentoo sucks. etc-update really really really sucks. dispatch-conf is a lot better…but it could have been a bit more optimized, especially when it replaces just headers and comments from files.

Filed by kargig at 02:08 under Linux
16 Comments | 7,767 views