I like the IDE. It has the features I want, chiefly, the ability to work on a Remote site (or a Local site if you prefer).

When you open a file (on the Remote site) the following happens (essentially):

1) Open the file for edit
2) Dreamweaver GETS the file and saves it locally
3) You edit the file (the local copy)
4) You save the file (locally)
5) Dreamweaver PUTS the local version over to the remote site. (NOTE: There is an option when configuring a new site which asks to “put files on save”. If you neglect to tick this then Dreamweaver will not automatically PUT the file.)

What is sweet about this procedure is that when you open the remote file for editing, Dreamweaver will compare the remote file against the local file (if it already exists) and warn you if they are different. But the larger issue here is that it actually will GET the file!

I use both my desktop PC and my laptop (via wireless) to edit the same projects that are stored on my local development webserver. I don’t have to worry that the local versions on the desktop and laptop will almost always be out of snyc. Dreamweaver takes care of that headache for me.

I just GET the file (if I know it is out of date) or just open the file and let Dreamweaver GET the file.

Life is good. I once tried to switch to Netbeans since it handles remote projects, but alas it only does a PUT on Save (if so configured) and not a GET on edit. That is vital functionality for me. A showstopper without out.

I work on many remote servers maintaining websites and projects. With Dreamweaver I can be assured that I am editing the most up to date version of the remote file.

Ok, so what is the point of this post?

Recently I upgraded both my laptop and my desktop from Ubuntu 11.10 to 12.04 LTS. Overall the experience was painless. However, suddenly lost SFTP ability with Dreamweaver.

Oh nose!

This is a fairly common issue and typically has to do with cached ssh keys. Not in this case. This was a fresh install (and I mean… I reformatted and installed 12.04 new — then installed wine and everything else). There were no cached keys. No, this was worse.

I noticed that I could connect to a remote site via FTP, but not SFTP. Interesting. With Ubuntu 11.10 I didn’t have problems connecting to remote sites via SFTP, and I had recently gone from Ubuntu 11.10 to Mint 12 back to Ubuntu 11.10 then to Ubuntu 12.04 on my laptop, each time reformatting. There were no cached keys, and with 11.10 and older connecting to remote sites via SFTP was ok.

I looked in my sysiog (/var/log/syslog) and found these entries whenever Dreamweaver started and then tried to connect remotely via SFTP):

May 8 09:09:56 coco kernel: [34928.411034] ptrace of non-child pid 12243 was attempted by: wineserver (pid 12215)

Well, I’ll cut to the chase and explain how to resolve the issue.

Edit your sysctl.conf file

$ sudo vi /etc/sysctl.conf

at the end of the file add this line
kernel.yama.ptrace_scope=0

Save, then restart your PC (log out may be sufficient, but I did a restart).

Now open Dreamweaver and connect to your remote site via SFTP!

The other day I rsync’d one of my folders to back up the documents to an external harddrive. Later, I copied the files back (not using rsync — but cp).

The permissions on the files changed, setting the executable bit on all the files. Now, when I open a file in gedit I am asked if I want to display the contents or run the file. Obviously it is text file, but because the executable bit is set linux treats it differently.

It is easy enough to change a few files:

$ chmod 644 *

To recursively change all of the files, I would use this set of commands:

$ find . -type f | xargs chmod 644

This works great…unless there are filenames which contain spaces in their names. In this case, the output from find will break up the filename into several pieces (separated by the space) and send each one over to xargs.

To prevent we use use the -print0 flag on find and the -0 flag on xargs:

$ find . -type f -print0 | xargs -0 chmod 644

Try it!

So what happens?  Well, there is a library conflict which causes my desktop to revert to a low-graphics, single-monitor mode.  Yuck.

Fortunately, it is easy to correct.

Step 1. After booting up, open up a command prompt

Step 2.  Stop the gdm (x-windows) daemon.

$ sudo service gdm stop

Enter your root password and x-windows will stop and you revert to a terminal login.

Step 3.  Login.

Step 4.  You have to be root to install the driver so go ahead and change to root.

$ sudo su -

Enter your root password

Step 5.  Now, at some point you would have already downloaded the NVIDIA driver.  Hopefully you kept it laying around somewhere.  Change to the directory containing the driver. In my case, it is the Downloads directory.

# cd /home/scott/Downloads

My NVIDIA driver is named NVIDIA-Linux-x86-173.14.25-pkg1.run

Step 6. As root, install the driver

# ./NVIDIA driver is named NVIDIA-Linux-x86-173.14.25-pkg1.run

This starts the installer.  Follow the onscreen instructions.  The installer will detect that the driver is already installed.  Select ‘Yes’ to continue.

It will report an error (which is what caused the problem in the first place!)

ERROR: File ‘/usr/lib/xorg/modules/extensions/libglx.so’ is not a symbolic link.

Accept the error and continue.

You will then get a prompt asking you if you want to run nvidia-xconfig to automatically update your X configuration.  I select No.

Installation of the graphics driver is complete!

Step 7. Exit root

# exit

Step 8. Restart X-Windows

$ sudo service gdm start

Enter your root password

Congratulations!  Your NVIDIA graphics driver is (re) installed and your dual-head display is back the way it was.

Not only do I use SSH, but I often use SCP (secure copy) to copy files from one server to another. Both commands require a password.

There is a way, however, to create a unique key pair so that you can login to a remote server without having to enter a password.

SSH Without a Password

The steps are simple and straightforward.

1. On the client run the following commands:

$ mkdir -p $HOME/.ssh

$ chmod 0700 $HOME/.ssh

$ ssh-keygen -t dsa -f $HOME/.ssh/id_dsa -P ”

NOTE: Those are a pair of single quotes after the -P, which will set an empty password.
This will create two keys, a private key and a public key, the .ssh folder.

2. Copy $HOME/.ssh/id_dsa.pub to the remote server.

3. Log in to the remote server and run the following commands:

$ cat id_dsa.pub >> .ssh/authorized_keys

$ chmod 0600 .ssh/authorized_keys

That’s it! You can log out of the remote server then log in using SSH and will not be prompted for a password.

scott@blackjack:~$ ssh-keygen -t dsa -f $HOME/.ssh/id_dsa -P ”
Generating public/private dsa key pair.
Your identification has been saved in /home/scott/.ssh/id_dsa.
Your public key has been saved in /home/scott/.ssh/id_dsa.pub.
The key fingerprint is:
a5:3c:b6:e9:13:85:62:34:ee:6f:56:47:46:6f:dd:8b scott@blackjack
The key’s randomart image is:
+–[ DSA 1024]—-+
| |
| o . |
| o . … . ..|
| +..o. o o o|
| o .S. o .. .|
| …+. .E . |
| .oo . |
| .= |
| o.. |
+—————–+
scott@blackjack:~$ cd .ssh
scott@blackjack:~/.ssh$ ll
total 24
drwx—— 2 scott scott 4096 2010-08-17 09:01 ./
drwxr-xr-x 55 scott scott 4096 2010-08-17 08:58 ../
-rw——- 1 scott scott 668 2010-08-17 09:01 id_dsa
-rw-r–r– 1 scott scott 605 2010-08-17 09:01 id_dsa.pub
-rw-r–r– 1 scott scott 4714 2010-06-25 14:55 known_hosts
scott@blackjack:~/.ssh$ scp id_dsa.pub sstanger@midnight:/home/sstanger/
sstanger@midnight’s password:
id_dsa.pub 100% 605 0.6KB/s 00:00
scott@blackjack:~/.ssh$ ssh sstanger@midnight
sstanger@midnight’s password:
[sstanger@midnight ~]$ cat id_dsa.pub >> .ssh/authorized_keys
[sstanger@midnight ~]$ chmod 0600 .ssh/authorized_keys
[sstanger@midnight ~]$ exit
Connection to midnight closed.
scott@blackjack:~/.ssh$ cd
scott@blackjack:~$
scott@blackjack:~$ ssh sstanger@midnight
[sstanger@midnight ~]$

NOTE: On some systems the ssh policies may not allow this. If this is the case then you can make this change on the remote host:

[sstanger@midnight ~]$ mv .ssh/authorized_keys .ssh/authorized_keys2
[sstanger@midnight ~]$ chmod 700 .ssh
[sstanger@midnight ~]$ chmod 640 .ssh/authorized_keys2

Manually mounting a USB drive is achieved with a few simple steps in a terminal window:

First we need to know what the name of the file system is. With the device plugged in we use the fdisk command.

scott@blackjack:~$ sudo fdisk -l
[sudo] password for scott:

Disk /dev/sda: 400.1 GB, 400088457216 bytes
255 heads, 63 sectors/track, 48641 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0×6669894e

Device Boot Start End Blocks Id System
/dev/sda1 * 1 47888 384660328+ 83 Linux
/dev/sda2 47889 48641 6048472+ 5 Extended
/dev/sda5 47889 48641 6048441 82 Linux swap / Solaris
Note: sector size is 2048 (not 512)

Disk /dev/sdb: 1016 MB, 1016858624 bytes
16 heads, 32 sectors/track, 969 cylinders
Units = cylinders of 512 * 2048 = 1048576 bytes
Disk identifier: 0×00000000

Device Boot Start End Blocks Id System
/dev/sdb1 * 1 969 992182 6 FAT16
scott@blackjack:~$

The device is called sdb1, and it is this that we will mount. The sequence of commands to mount and then unmount are:

scott@blackjack:~$ cd /media
scott@blackjack:/media$ sudo mkdir thing
scott@blackjack:/media$ sudo mount /dev/sdb1 thing
scott@blackjack:/media$ sudo umount thing
scott@blackjack:/media$ sudo rmdir thing
scott@blackjack:/media$