Showing posts with label ubuntu. Show all posts
Showing posts with label ubuntu. Show all posts

Tuesday, October 21, 2008

Amazing -- my EeePC is working perfectly again!

It's almost a week since my dualboot system on the EeePC 901 stopped functioning (as described here) and I was back to only having Xandros (advanced mode) working.
I had contacted Asus about sending the machine back to them and was waiting for a response which, according to their 48 hour policy, was actually due this morning.
My feeling up to now was that some unfixable hardware problem had arisen which caused the screen to go black when the DCIn cable was plugged in or unplugged which situation was only recoverable by rebooting.
Then today I came across this thread referring to a similar, but not identical, problem which offered this solution reported to work for many poeple:

1. Unplug power cable and remove battery
2. Hold down power key for 30 seconds
3. Leave the battery out and plug in power cable
4. Hold down power key for 5 seconds during whuich time the machine should boot
5. If everything looks good, shut down as normal and replace the battery

So, I tried this and when the computer rebooted in step 4, the Grub menu came up. Afetr replacing the battery, and booting into Ubuntu, it seemed that everything was back to normal. Even unplugging and replugging the power cable caused no problems.
Looks like everything is completely back to normal which is marvelous given that I had totally lost all hope of a recovery.

Monday, October 13, 2008

Installing Tasque in Ubuntu Eee

I had already posted on the installation of tasque in Ubuntu but, as I discovered in trying to install it today on my EeePC, I didn't write it very clearly.
So here are steps again with more precision:

1. Download the tasque deb (tasque_0.1.5-1_i386.deb)from here and install.
2. Add this repo to /etc/apt/sources.list
deb http://ppa.launchpad.net/banshee-team/ubuntu hardy main
3. After updating, install these three dependencies:


mono-gmcs
libnotify0.4-cil
libgtk2.0-dev


...and that's it. Tasque should now work although you will have to connect to RTM first and login.

Ubuntu Eee now looking fabulous.

Yesterday I had some reservations about Ubuntu Eee particularly due to problems I had with 3D effects and Yakuake.
Know what the culprit was? Yes, the so-called Netbook Remix which seems to have been designed to make Ubuntu Eee as ungeeky as possible and probably ties in with EeePC's originally targetted audience of schoolkids.
Luckily this is easy to get rid of although I did run into a small problem which is explained in this post.
Essentially, the solution was, as it often is, to reboot.
After this, Ubuntu Eee started to look amazing. In everything I've done today, I cannot see anywhere where Ubuntu Eee is lacking anything in comparision to the full blown Ubuntu Hardy on a full-sized box.
Simply, everything works great -- 3D effects, Yakuake, AWN, Livestation, Brazilian characters
I installed AWN from the Reacocard repo and. again, everything worked fine.
The only problem I had, if you can even call it a problem, relates to a slight, but easily resolvable, difficulty to get sound working when an audio app starts up. I posted to the forum on this but have yet to receive a reply.
I also had another sound problem in that the headphones gave a lot of static-type noise even with no audio app playing. However, to resolve this was just a matter of muting the Microphone in the sound mixer.
All-in-all, a very pleasing day.

Sunday, October 12, 2008

Yes, we're dual-booting on the EeePC

Once I had managed to get the new partitions on the 16 GB drive, installation of Ubuntu eee from a USB key was relatively easy.
The same precautions are required every time you want to boot from anything other than the internal HDs as the Boot Device Priority setting is not remembered from boot-to-boot.
Interestingly, the /dev/sdb3 partition that wasn't seen initially by Xandros presented no such problems to the Ubuntu installer.
The install went smoothly (without any of the "no-migration-assistant" problems experienced with Intrepid Ibex). Note that I chose to install no bootloader.
he problems started when I tried to configure the /boot/grub/menu.lst of Xandros to take care of the Ubuntu boot. I first tried this from within Xandros. Everything seemed to go well but when I rebooted (after having commented out the "hidden menu" line in /boot/grub/menu.lst), no grub menu appeared no matter how many times I tried it.
I eventually found what the problem was when I stumbled on this guide (see the section titled "Changing the original GRUB menu"). It seems the changes I had made went to /dev/sda2 while the only changes that will have an effect must be made to /dev/sda1. This device can only be accessed from another partition.
So, I had to go into a LiveCD version of Ubuntu eee (from the bootable USB key) and ad this line to the /etc/fstab

/dev/sda1 /media/sda1 ext2 defaults 0 0

Then I created the sda1 folder in /media and mounted it. Then I could cd into that folder and edit /boot/grub/menu.lst.
Having done this, it was also necessary within the same folder to store the vmlinuz and initrd.img files of Ubuntu eee to /boot/ubuntu. It seems that /dev/sda1 views the 16 GB HD as an external drive.
With this done, I now got a grub menu on starting up and could boot into Ubuntu eee by choosing it from the menu.
Ubuntu eee worked fine and, after adding 36 new updates, I was able to install Opera 9.60 (note that this time, libqt3-mt was downloaded and installed by Opera itself rather than me having to do it as was the case with the LiveCD version).
I installed Yakuake from the repos and Livestation from here. Livestation works pretty well bu Yakuake doesn't seem to like being at 50% width as pressing F12 drops down the 50% width consle but a "shadow" console also drops down in the remaining 50%.
I aso tried 3D effects but, although, I got the wobbly windows and emerald themes, I couldn't get the cube to rotate. I have feeling that /etc/X11/xorg.conf might need to be modified.
All in all, however, this was a major stpe forward and looks very promising.

Saturday, September 20, 2008

Dragbox -- had a look.

Dragbox is described here but the most up-to-date version can be downloaded here as a .tar.gz file.
I tried it today in Ubuntu Hardy on my Dell.
To install you must untar (tar -zxvf file.tar.gz), CD to the new directory and configure (./configure). For me the configure completed without a problem. Then its just compile (make) and install (sudo make install).
Dragbox has a reasonable manpages which I reproduce here:

EXAMPLES
dragbox . file1.txt
Adds the current directory and file1.txt to the currently
running dragbox, or creates a new one. If file1.txt does not
exist, its name is added as text.

dragbox -t "Dragbox is cool"
Adds a text snippet to the currently running dragbox, or cre‐
ates a new one.

dragbox --get -0 | xargs -0 tar cfz backup.tar.gz
Get current items and pipe to xargs to create a tarball. Note
that this simple command will have problems if contents
include text items.

dragbox --name remember .
Put the current directory on a shelf called "remember"

TIPS AND TRICKS
Dragging in gnome works very well, but there are some hidden tricks
that can make some things easier.

Pressing modifier keys can switch between different types of drags;
dragging a file to nautilus normally copies it, but you can hold
shift to move it, or shift-ctrl to make a symlink.
You can drag from any window without activating it by holding the
altgr or super key when you drag. This is often very useful.

You can bring up a minimized or obscured window while dragging by
hovering momentarily over it’s button in the window list (in the
gnome-panel)

Just like minimized windows, you can switch workspaces while drag‐
ging if you hover over the workspaces switcher for a short while.

If you selected something in a window and focus another window, it
looks like the selection disappeared. In many cases it’s still
there, you can try to drag it without focusing the window (see
above)

Type escape while dragging to cancel the drag

ACCESSING DRAGBOX
Dragbox needs an X session to run and display its shelves. However,
to access dragbox shelves from non-graphical sessions, all you need
is to make sure dragbox knows which X session to connect to. It
might be enough to define the DISPLAY variable.


Dragbox seems to work fine, although it gives errors in the terminal (I launched it from a terminal) about some dbus error which don't seem to affect it at all.
So, for example I can use the command
dragbox file.txt

to bring a text file to the dragbox without physically dragging anything.
OK, so it works but I have to admit that, right now, I just can't see that this is going to be of immense use to me. Indeed, I just can't think of any circumstance where I would use this at all.
Maybe I'm just too ignorant to spot the benefits but I'm very willing to learn.

Edit (21-09-2008): Just installed it by same method in Foresight 2.0.4 on the Dell. Works just the same. Now if only I could find a use for this tool.

Edit (21-09-2008): This is a useful review which goes someway to explaining the benefits of Dragbox. Indeed, one immediate benefit is that it brings a multi-item clipboard to Linux (much like Shadow in OS X).

Monday, September 08, 2008

Google Chrome in XP (VM) under Ubuntu

Once I had my VM version of Windows XP running under Ubuntu on the Dell, it was a very simple matter to get Google Chrome up and running.
The other alternative was to run Chrome with Wine in Ubuntu but I understand that this is not as fast as the normal Chrome in Windows.
However, in the VM Windows, it seems to be absolutely fine with no shortcomings that I have seen so far.

Problems installing VMware Server in Ubuntu Hardy

For some reason, VMware Server is no longer installed in Hardy on the Dell. So, today, expressly for the purpose of running Google Chrome within/under Ubuntu, I tried to install it again.
I used this guide but I still had problems:
1. VMware Server would not launch either from dock, from Applications>System Tools or from a terminal (vmware-server-console). In this latter, just got a command not found error.
The solution was as shown in post #3 of this thread which is to run these two lines in a terminal AFTER install

sudo ln -sf /usr/lib/gcc/i486-linux-gnu/4.2.3/libgcc_s.so /usr/lib/vmware/lib/libgcc_s.so.1/libgcc_s.so.1

and
sudo ln -sf /usr/lib/libpng12.so.0 /usr/lib/vmware/lib/libpng12.so.0/libpng12.so.0

Note that I had initially installed VMware Server 1.0.6 and then ran these two lines. Then I uninstalled this and installed 1.0.7 in its place. However, this too would not launch UNTIL I had run these same two lines (not at all changed) once again.
2. Permissions problem in running .vmx file Error message said something about insufficient permissions to copy the .vmx to the "inventory".
I then changed the owner of my VM folder (~/virtualmachines) from root to me (same with group) and used the Apply to Enclosed Files button.
However, still got permissions error.
Only when I ran this command
sudo chown -R paul:paul ~/virtualmachines

did the problem go.
3. It is worth noting that without the installation of xinetd, I kept getting a "Can't open inetd" error during the vmware server compile.

Monday, September 01, 2008

Remote login OS X

This enables logging into the Mac from a remote computer.
The steps are:
On the Mac (as server):
1. System Preferences >> Sharing : Here check the box Remote Login and select who you want to use this facility in the "Allow access for" box. Note the remote login command. In my case it's

ssh paul@192.168.1.5

In my case, this didn't work unless I used
sudo ssh paul@192.168.1.5

Without sudo I got a lot of stuff about unknown keys after which it went back to the command prompt.
With sudo, on the other hand, I was given the option to accept, or not, the authenticity of the host. After typing "yes", the connection was made.
Now I could delve into whatever directory on the Mac from Ubuntu on the Dell and use scp (without sudo) a file to the Desktop of the Dell with
scp file.ext 192.168.1.12:~/Desktop

This caused me to be asked again to accept or not the authenticity of the host (this time the Dell was the server or host).

Wednesday, August 20, 2008

Well, does pdnsd make browsing faster?

After at least 14 hours of browsing in Ubuntu on the MacBook with pdnsd turned on, my conclusion is that I just cannot at all see any evidence for faster browsing. Honestly, to me it looks exactly the same, no slower and no faster.
And yes, I am sure that pdnsd is running and the selected DNS server is 127.0.0.1. Here's what I get from nslookup:

$ nslookup google.com
Server: 127.0.0.1
Address: 127.0.0.1#53

Non-authoritative answer:
Name: google.com
Address: 64.233.167.99
Name: google.com
Address: 72.14.207.99
Name: google.com
Address: 64.233.187.99


So, this is disappointing. Nevertheless, given that pinging my usual DNS servers only took 65 ms, that doesn't provide much leeway for speeding up page retrieval.
I'm really going to have to try and find out if others REALLY are getting faster browsing?
On the same topic, here are two interesting articles from SourceForge. The first covers how DNS works in Linux and the second looks at a more elaborate local DNS server setup for the purpose of, yes, speeding up browsing.

Tuesday, August 19, 2008

Use a local DNS server to speed up browsing

Well, this sounds like a good idea. After all, we would all prefer to have any clicked link or icon or speeddial applet cause the desired webpage to show up as near to instantaneously as possible.
As all of the usual means of calling up a webpage involve adding a "word" address to the address bar, the first stop nearly always has to be to the DNS server to translate those words into the numbers which the computer can understand.
So, how long does this take?
I use OpenDNS (208.67.222.222 and 208.67.220.220) and if I ping either of these, I get a reply after about 65 ms. So, if I could use a local DNS server that allowed me to retrieve the IP address of the webpage I could perhaps save up to 65 ms of time per page.
If page typically takes 2 seconds to open, am I really going to notice if it opens 0.065 seconds more quickly? I really don't think so.
What I'm not sure of in this scenario is how long the OpenDNS servers take to find the IP of the web address compared to how long it would take to retrieve the same information from a small local cache.
So, there may be a possibility here to speed up my browsing so I though I'd give it a try.

The first one I tried was dnsmasq which is available in the Hardy repos. I used this guide.
The installation and configuration went smoothly (I listened to 127.0.0.1) but trying to retrieve a new web page was intolerably slow and sometimes didn't happen at all. Seemingly, when nothing relevant was in the LocalHost cache, it didn't then look in the OpenDNS servers (although I didn't have these included in /etc/resolv.conf -- only 127.0.0.1 here -- 208.67.222.222 and 208.67.220.220 are the DNS servers in my router).
Certainly, running, for example, dig google.com showed the 127.0.01 was being used as the DNS server. Also, the first attempt at this gave maybe up to 300 ms as the retrieval time, the second time gave typically 0-2 ms retrieval which is certainly a vast improvement.
Unfortunately, these encouraging results did not translate themselves into good browsing performance. Indeed, browsing became almost impossible.
There are several other howtos on dnsmasq for Ubuntu such as this and this, but nothing I found allowed me to eliminate this problem.
Also, because dnsmasq saves the cache to RAM, this means that you have to start building your cache again after a reboot -- which means slow browsing until all your usual page have been called up at least once.
Given these disadvantages, I decided to give pdnsd a try. This saves the cache to disk so it's not lost.

I first used this guide but this simply didn't work. In trying to start pdnsd, I consistently got an error about the absence of the file

/etc/cache/pdnsd/pdnsd.status

which indeed was not present.
Also, I could't find any reference to the absence of this file in google.
So, next I tried this Ubuntu howto although I first tried the guide in this post from much later in this thread.
This however, didn't give a working pdnsd as this command
$ sudo /etc/init.d/pdnsd restart

always gave an error on the start part of the command.
Finally, I tried the guide in post #1 and this worked without problems.
What I did was as follows:

sudo apt-get install pdnsd
##I did NOT install resolvconf
sudo gedit /etc/pdnsd.conf


My /etc/pdnsd.conf is given here:

// Read the pdnsd.conf(5) manpage for an explanation of the options.

/* Note: this file is overriden by automatic config files when
/etc/default/pdnsd AUTO_MODE is set and that
/usr/share/pdnsd/pdnsd-$AUTO_MODE.conf exists
*/

global {
perm_cache=1024;
cache_dir="/var/cache/pdnsd";
run_as="pdnsd";
server_ip = 127.0.0.1; // Use eth0 here if you want to allow other
// machines on your network to query pdnsd.
status_ctl = on;
paranoid=on;
// query_method=tcp_udp; // pdnsd must be compiled with tcp
// query support for this to work.
min_ttl=15m; // Retain cached entries at least 15 minutes.
max_ttl=1w; // One week.
timeout=10; // Global timeout option (10 seconds).

// Don't enable if you don't recurse yourself, can lead to problems
// delegation_only="com","net";
}


server {
label=OpenDNS;
ip=208.67.222.222;
ip=208.67.220.220;
timeout=30;
interval=30;
uptest=ping;
ping_timeout=50;
purge_cache=off;
}

// This section is meant for resolving from root servers.
server {
label = "root-servers";
root_server=on;
ip = 198.41.0.4
, 192.228.79.201
, 192.33.4.12
, 128.8.10.90
, 192.203.230.10
, 192.5.5.241
, 192.112.36.4
, 128.63.2.53
// , 192.36.148.17
// , 192.58.128.30
// , 193.0.14.129
// , 198.32.64.12
// , 202.12.27.33
;
timeout = 5;
uptest = query;
interval = 30m; // Test every half hour.
ping_timeout = 300; // 30 seconds.
purge_cache = off;
exclude = .localdomain;
policy = included;
preset = off;
}


source {
owner=localhost;
// serve_aliases=on;
file="/etc/hosts";
}

rr {
name=localhost;
reverse=on;
a=127.0.0.1;
owner=localhost;
soa=localhost,root.localhost,42,86400,900,86400,86400;
}

/*
neg {
name=doubleclick.net;
types=domain; // This will also block xxx.doubleclick.net, etc.
}
*/

/*
neg {
name=bad.server.com; // Badly behaved server you don't want to connect to.
types=A,AAAA;
}
*/

/* vim:set ft=c: */

I will admit that I have no idea why the root-servers section is required particularly as the servers involved have no relevance to my setup. I'll try a test where I take this section out.

Next, I amended /etc/dhcp3/dhclient.conf where I took out the line
prepend domain-name-servers 208.67.222.222, 208.67.220.220;

and added
prepend domain-name-servers 127.0.0.1;

Now I rebooted (and this seems to be important and started pdnsd with this
sudo /etc/init.d/pdnsd restart

and it started without a hitch.
Now everything worked perfectly. For example
dig google.com

gives this
; <<>> DiG 9.4.2-P1 <<>> google.com
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 44898
;; flags: qr rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 4, ADDITIONAL: 4

;; QUESTION SECTION:
;google.com. IN A

;; ANSWER SECTION:
google.com. 855 IN A 72.14.207.99
google.com. 855 IN A 64.233.187.99
google.com. 855 IN A 64.233.167.99

;; AUTHORITY SECTION:
google.com. 172704 IN NS ns1.google.com.
google.com. 172704 IN NS ns2.google.com.
google.com. 172704 IN NS ns3.google.com.
google.com. 172704 IN NS ns4.google.com.

;; ADDITIONAL SECTION:
ns1.google.com. 172704 IN A 216.239.32.10
ns2.google.com. 172704 IN A 216.239.34.10
ns3.google.com. 172704 IN A 216.239.36.10
ns4.google.com. 172704 IN A 216.239.38.10

;; Query time: 0 msec
;; SERVER: 127.0.0.1#53(127.0.0.1)
;; WHEN: Wed Aug 20 00:23:44 2008
;; MSG SIZE rcvd: 212

Note that the first time I used dig google.com, the Query time was something quite high (350 ms) but the second time it reverted to 0 ms presumably because of the local DNS cache.
Another useful command is
sudo pdnsd-ctl server OpenDNS retest

which gives this output
Opening socket /var/cache/pdnsd/pdnsd.status
Succeeded

and
sudo pdnsd-ctl status

which gives this output
pdnsd-1.2.6-par running on ubuntu-mac.

Cache status:
=============
1024 kB maximum disk cache size.
133062 of 1058816 bytes (12.6%) memory cache used in 513 entries.

Thread status:
==============
server status thread is running.
pdnsd control thread is running.
tcp server thread is running.
udp server thread is running.
476 query threads spawned in total (0 queries dropped).
0 running query threads (0 active, 0 queued).

Configuration:
==============
Global:
-------
Cache size: 1024 kB
Server directory: /var/cache/pdnsd
Scheme file (for Linux pcmcia support): /var/lib/pcmcia/scheme
Server port: 53
Server ip (0.0.0.0=any available one): 127.0.0.1
Ignore cache when link is down: off
Maximum ttl: 604800
Minimum ttl: 900
Negative ttl: 900
Negative RRS policy: auth
Negative domain policy: auth
Run as: pdnsd
Strict run as: on
Use NSS: on
Paranoid mode (cache pollution prevention): on
Control socket permissions (mode): 600
Maximum parallel queries served: 40
Maximum queries queued for serving: 60
Global timeout setting: 10
Parallel queries increment: 2
Randomize records in answer: on
Query method: udp_only
Query port start: 0
Query port end: 65535
TCP server thread: on
TCP query timeout: 30
Delegation-only zones: (none)
Server 0:
------
label: OpenDNS
ip: 208.67.222.222
server assumed available: yes
ip: 208.67.220.220
server assumed available: yes
port: 53
uptest: ping
timeout: 30
uptest interval: 30
ping timeout: 50
ping ip: (using server ip)
interface:
device (for special Linux ppp device support):
uptest command:
uptest user: (process owner)
force cache purging: off
server is cached: on
lean query: on
Use only proxy?: off
Assumed root server: no
Randomize server query order: no
Default policy: included
Policies: (none)
Server 1:
------
label: root-servers
ip: 198.41.0.4
server assumed available: yes
ip: 192.228.79.201
server assumed available: yes
ip: 192.33.4.12
server assumed available: yes
ip: 128.8.10.90
server assumed available: yes
ip: 192.203.230.10
server assumed available: yes
ip: 192.5.5.241
server assumed available: yes
ip: 192.112.36.4
server assumed available: yes
ip: 128.63.2.53
server assumed available: yes
port: 53
uptest: query
timeout: 5
uptest interval: 1800
ping timeout: 300
ping ip: (using server ip)
interface:
device (for special Linux ppp device support):
uptest command:
uptest user: (process owner)
force cache purging: off
server is cached: on
lean query: on
Use only proxy?: off
Assumed root server: yes
Randomize server query order: no
Default policy: included
Policies:
exclude: .localdomain.
Succeeded

The big question, of course, is "is browsing any faster"? Well, it's certainly not a whole lot slower. However, I have the impression that pages which are opened for the first time are just a little bit slower than they would have been with the OpenDNS servers alone. Thereafter, I believe they might be just a shade faster.
I have the impression that I should not have stopped prepending the servers 208.67.222.222 and 208.67.220.220 from /etc/dhcp3/dhclient.conf as now, these servers are missing from /etc/resolv. I believe that when 127.0.0.1 can't resolve the address, the OpenDNS servers should then be used to do the resolving and thereby avoid untoward delays with pages whose adresses are not yet in the cache.
Nevertheless, I believe I would need a few days of browsing to get a better feel for whether or not there is a performance boost.

Saturday, August 16, 2008

Speed and performance in Ubuntu Hardy

During my Windows days (long time ago) I had always been interested in trying out all sorts of half-assed, snake-oil type schemes to speed up performance on my computer. However, I soon learnt that the vast majority, if not all, of these so-called tips, were simply nonsense.
Nevertheless, I came across this blogpost and decided I'd try out the suggestions just to see. My reasoning was that I have a lot more faith in suggestions made in the Linux arena than I have for Windows which seems surrounded by an army of gobshites and chancers.
Now I didn't do the writeback (in /etc/fstab) option due to the warnings given by the author.
The problem with these tweaks is that it's hard to truly judge whether or not there has been any performance improvement, but in general I can't say I noticed any difference, bad or good.
Nevertheless, it's best to leave it over a few days to see if anything at all has changed.
The boot time(from selecting Ubuntu from the Grub menu to the cpu reverting to 0% on the desktop) was 82 seconds. Unfortunately, I had never measured it before but certainly it doesn't seem like it booted any faster or slower than before.

Additionally, I now get about 80% of the bandwidth that I'm paying the ISP for. However, according to the ISP, anything over 78% is acceptable. Well, OK. At least, they didn't say anything over 0% was OK by them.
Nevertheless, I wanted to see if there was anyway to add a few bps to my bandwidth. I am of course aware of the plethora of total garbage, and fake claims and offers, that envelope this issue particularly in the Windows arena.
However, from the same blog as I used earlier in this post I got this and tried it.
I then tried three different speed tests (speakeasy, gambitdesign and the one from my ISP). Every single one of them showed absolutely no evidence for any improvement whatsoever.

So, a bit of a blank day on the performance front but at least things didn't get worse. In reality, I'd like to observe how things look over the next few days before drawing a final conclusion.

Wednesday, August 13, 2008

Sound problem Ubuntu Hardy (Mac)

This is very strange but today sound didn't work in Hardy on the MacBook despite having worked perfectly yesterday.
Luckily the fix was no more than what I had done here using the solution I found in this thread.
OK, great that it worked fine after this but what happened that I lost the sound? For the moment I have not the vaguest idea what went wrong.

Remote desktop access -- more experimentation.

Well, there certainly is no shortage of means to let you access and control a distant desktop from the comfort of your own desk as shown in this article.
As I had had some problems with xncviewer, particularly from KuruminNG (Dell), I decided to try one other app listed in this article. This was x11vnc.
First, I tried to access Foresight (on the Dell) from Ubuntu (on the Mac). The steps are very simple and are as follows:
1. Download/Install x11vnc on the remote desktop (Foresight-Dell). For this I used:

sudo conary update x11vnc=@fl:2-devel

2. Start up the remote server by typing this in a terminal
x11vnc -display :0
3. In the client (your own computer or Ubuntu-Mac in my case), type this
vncviewer 192.168.1.12:0
and almost immediately, Foresight's desktop appeared on the Mac. Note that sshd was NOT running on Foresight at the time and NO passwords at all were required. Absolutely nothing, very quick and problem-free.

Note that in Ubuntu, I needed to download xtightvncviewer (one of four suitable apps available) before vncviewer would work.
Really, this whole thing, including the downloads, installs and connections could easily be set up in less than 5 minutes.

Flushed with this success, I tried to see if I could control KuruminNG (Dell) from Ubuntu-Mac. Here the steps were:
1. Install x11vnc in Kurumin
sudo apt-get install x11vnc

2. In a Kurumin terminal, type
x11vnc -display :0

3. In Ubuntu-Mac, type
vncviewer 192.168.1.12:0


Once again, connection was almost immediate with no requests for passwords or anything else.

Now, the area where I had a problem with Kurumin before, controlling another Linux desktop from Kurumin as client. So, for this I tried:

1. Install x11vnc in Ubuntu-Mac
sudo apt-get install x11vnc

2. In Ubuntu type
x11vnc -display :0

3. In Kurumin type
vncviewer 192.168.1.1:0


For the third time, everything worked perfectly, absolutely no flaws whatsoever. However, there was a difference from the last two attempts, in that this time, after the third step in Kurumin (vncviewer), a window appeared in Ubuntu asking did I want to accept or refuse the request to control the desktop. Once I said yes, connection was immediate.

My conclusion based on this very brief and incomplete test is that x11vnc seems quite a marvelous app and well worth trying again.
One big difference from NXserver, is that the client sees exactly what is on the servers screen right now. Also, the clients cursor moves around the servers screen.

Saturday, August 09, 2008

Subversion 1.5 in Ubuntu Hardy

While I was trying to reinstall my madwifi driver in Hardy on the Mac, I had to use this command:

sudo svn co http://svn.madwifi.org/madwifi/trunk madwifi

However, this time I got this error message:
svn: This client is too old to work with working copy 'madwifi'; please get a newer Subversion client

Puzzling, as I had the latest version of subversion that was available in the Ubuntu repos (1.4.6).
In googling around, I came across this post which explained what the problem was (basically there's a new version 1.5 which is available in the Ubuntu Intrepid repo but NOT in Hardy).
Unfortunately, the repos in this post don't work (at least, they didn't for me). However, I came across this thread that discusses the same problem. The repos shown in post#3 work fine. here they are:
deb http://ppa.launchpad.net/clazzes.org/ubuntu hardy main
deb-src http://ppa.launchpad.net/clazzes.org/ubuntu hardy main

Intrepid Ibex bites the dust

After updating Intrepid (Mac) today, I found it would no longer boot (to anything usable, that is). Just kept going to a black screen after the login page.
If I tried the recovery option, and did the X server fix, I might sometimes go as far as actually seeing the desktop wallpaper, but no icons, no panels, nothing else.
First I tried to reinstall II which went fine and then installed all of 330 uṕdates. This seemed to go fine too, until I rebooted, that is. Now, it exactly the same, just a black screen.
I should mention here that every time I re-installed Intrepid, all of the partitions on my HDD became invisible such that nothing at all appeared on the rEFIt menu page -- just a white screen. Today when this happened, there was no boot flag against ANY partition (I had previously attributed this problem to the presence of the boot flag as when I got rid of the problem wet.)
This time I just rescanned the disk and that brought everything back to visible. I really have no idea what's going on here but luckily it's easy to resolve.
What was good about Intrepid in the few days I was using it?
1. Booted a bit faster (actually maybe 30-40 seconds faster than Foresight on the Dell). This, of course, only applies when it does actually boot to completion which was not often.
2. The recovery mode boot option includes options like "fix X server" which is a useful addition
I honestly didn't see anything else that really excited me

What was bad about II?
1. The shutdown button on the panel only allowed logout.
2. Prepending preferred DNS servers in /etc/dhcp3/dhclient.conf didn't work. I was forced to put what I wanted in /etc/resolv.conf and make that file immutable.

Maybe I'll try it again when it gets to beta.

Thursday, August 07, 2008

More problems with Intrepid Alpha 3

After shutting down Intrepid (MacBook) with apparently everything functioning normally, this morning's stratup was less rosy.
Although it booted, it kept going to a black screen just after the Initial screen (with progress bar) dissappeared. So, couldn't ever login. Thought I was looking at another re-install.
But no, went to Recovery Mode in the boot menu and this gave me a list of recovery options (this is an upgrade from just the TTy1 console screen which was the only thing available historically).
Chose the Fix X Server (or something) option. Then when it had done this, I chose, boot and everything worked fine.
Same problem in later attempts, but after 2-3 blackscreens it eventually booted as normal.
Another problem I see now, is that the internet connection seems extremely slow. As usual this is a DNS problem. Despite having changed /etc/dhcp3/dhclient.conf to prepend the OpenDNS servers, I only have 192.168.1.254 as the only available DNS server on both my wired and my wireless connections to Intrepid.
Worse, there is NO network option available in Systems>>Administration. So, no other way to change DNS servers.
Wonder when this will be fixed.

Wednesday, August 06, 2008

Intrepid problems

Yesterday, I was congratulating myself that the upgrade I did in the morning brought Yakuake 2.9.3 back to Ubuntu Intrepid (alpha 3).
However, my jubilation was shortlived. When I shut down Ubuntu (as always had to go through Log Off and then Restart from Options) I noticed that in the login page the cursor wouldn't move and the keyboard seemed dead. However, I thought this was just a temporary problem that I could forget about.
Not so -- when I tried to boot to Intrepid this morning, both the mouse and the keyboard were totally inactive in the login screen. This effectively means you can do nothing other than a hard shutdown.
I then tried the Last Successful Boot option but this just brought me to a TTy console. Looked in /var/log/syslog but couldn't see anything obviously wrong.
I then tried an update from TTy1 (sudo apt-get update). However, nothing had changed after that -- still couldn't login.
Nothing for it but to re-install Intrepid.
Once again had to use

sudo ubiquity --no-migration-assistant

to get the install to go to completion.
In addition, I had the same problems with just a white screen showing up when I restarted the MacBook. And once again, the problem was a boot-flag causing every single one of the partitions to have a yellow warning sign (boot-flag was on first partition -- EFI). I took out the boot flag and applied. After this, all partitions became visible (no warning, data available).
However, then I put the bootflag back.
Now when I rebooted, I got the same problem again. So, back into Parted Magic and out with the bootflag.
Now I got the Apple on the rEFIt screen but no penguin.
So, I use the Hardy LiveCD to re-install the Ubuntu Grub in the MBR. I could do this while still asleep, now, but here are the steps:

1. In terminal in LiveCD, type
$ sudo grub

2. > find /boot/grub/stage1 # what partitions has Grub installed
3. > root (hd0,8) # pick the partition whose Grub you want to put in the MBR
4. > setup (hd0) # put this Grub in the MBR

You should get a seriesg of messages telling you that this has succeeded.
5. >quit # leave grub and go back to bash
and now reboot.

After all of this, Intrepid was back. After installing all 291 updates, everything worked fine, no problems.
Hard to explain why I had the problems but it's all back, at least even without an explanation.

Tuesday, August 05, 2008

Yakuake's back in Ubuntu Intrepid

The Yakuake developer replied to my email but basically said he couldn't see what the problem was from the error messages I got with both versions 2.9.3 and 2.8.1
However, a further 81 updates were available for Intrepid this morning. After the install, I tried again to install Yakuake 2.9.3 and this time it performed perfectly.
So, it seems there was some problem with the updates I installed yesterday. Oh well, it is still in alpha.

Monday, August 04, 2008

Remote access beyond your router

This is a nice tutorial on how to use NXserver and sshd in Ubuntu to connect to a remote computer anywhere in the world.
Strangely though, I seem to be falling at the first hurdle here as in the section called "STEP 1: Testing SSH", I get this

$ ssh localhost
The authenticity of host 'localhost (127.0.0.1)' can't be established.
RSA key fingerprint is 5e:13:99:b5:12:51:9c:1e:99:8e:25:ec:ef:42:33:ad.
Are you sure you want to continue connecting (yes/no)? yes
Failed to add the host to the list of known hosts (/home/paul/.ssh/known_hosts).
paul@localhost's password:
Linux ubuntu-mac 2.6.26-5-generic #1 SMP Sun Aug 3 01:25:54 UTC 2008 i686

The programs included with the Ubuntu system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Ubuntu comes with ABSOLUTELY NO WARRANTY, to the extent permitted by
applicable law.

To access official Ubuntu documentation, please visit:
http://help.ubuntu.com/
Last login: Mon Aug 4 23:32:31 2008 from localhost


Then, as per the tutorial, I type
ssh paul@xxx.xxx.xxx.xxx (where the x's represent my external IP)

and get this after quite a delay
ssh: connect to host xxx.xxx.xxx.xxx port 22: Connection timed out


Actually, I really don't understand this last command and maybe I've misinterpreted it. Right now it looks like I'm trying to ssh from Ubuntu on my Mac back to Ubuntu on my Mac. Doesn't seem right.

Problems with Yakuake 2.9.3 in Ubuntu Intrepi Alpha 3

When I first upgraded to Intrepid about a week ago, I installed Yakuake 2.9.3 from the repos and it worked perfectly as it has always done for me in Ubuntu.

However, this morning, there were a total of 79 updates available which I installed. Immediately afterwards, Yakuake would no longer launch for me. Trying to launch it from a terminal gave this error message:

(5872)/: Communication problem with "yakuake" , it probably crashed.
Error message was: "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." "



I uninstalled and re-installed Yakuake 2.9.3 but always got the same error.

After quite some time googling and trying various suggestions that I had found, I gave up as none worked and I couldn't see any light at the end of the tunnel.

So, I completely removed Yakuake 2.9.3 and installed Yakuake 2.8.1 from source. Although it installed, it too gave an error when I tried to launch it as follows:

kbuildsycoca running...
DCOP Cleaning up dead connections.
DCOP aborting (delayed) call from 'anonymous-23932' to 'yakuake'
paul@ubuntu-mac:~$ KCrash: Application 'yakuake' crashing...
Could not find 'drkonqi' executable.
KCrash cannot reach kdeinit, launching directly.


So, once again Yakauke failed to launch for me.

Please note that in both cases, Yakuake does actually appear to be running (running ps -e in a terminal) but I cannot get the terminal emulator screen to show itself.

It seems that some change resulting from the upgrade has made the use of Yakauke impossible. I have emailed the Yakuake developer (hein@kde.org) about this problem and maybe he'll have a suggestion.
In the meantime, I'm using Guake (which is a Gnome version of Yakuake). Although I didn't like it when I tried it first (mainly because the window width is not adjustable and I want a small window) this will give me an opportunity to assess Guake more thoroughly.