Of course, I’ve tinkered with tunneling X through SSH, tunneling VNC, the somewhat faster tight VNC, and I even tried the XRDP project which aims to provide a Windows compatible RDP server (remote desktop) for *nix systems. In the past I’d combine one of the above with a bare minimal window manager like ratpoison or evilWM, and I’d achieve somewhat usable speeds. Certainly tight VNC isn’t a bad option, and it can even provide for detachable and re-attachable X sessions, but nothing has really lured me away from screen till now.

Enter NoMachine’s NX. All I can say is I am simple stunned. The speeds are phenomenal. Indeed, nothing really prepared me for this kind of remote X goodness. I have a Gnome desktop running, and I am happily rediscovering X Windows and the Linux desktop after a long absence. As much as I love ncurses and screen, I think it would be foolish to deny that certain applications can benefit from a graphical interface. So, If you need remote access to your Linux desktop, I would definitely check it out. There is an open source offshoot called FreeNX, which is what I currently have installed on my Ubuntu system. NoMachine is also kind enough to offer their server with few strings attached, and clients are available for virtually all platforms.

Backing up your database

There are two approaches to backing up your mysql database. You can either backup the raw database files themselves (usually found somewhere in the proximity of /var/lib/mysql/), or you can dump the database (or individual tables) to a file that can be used to reconstruct it in the case of a disaster.

The perl script mysqlhotcopy is useful for doing raw copies. For creating a backup .sql script that can be used to recreate your database, I recommend mysqldump.

mysqldump dbname > dbname-bkup.sql

To restore using this backup:
mysql dbname < filepath_filename.sql

Note, before restoring, if a database with this name already exists, you may want to delete it. Alternatively, if you use the switch –add-drop-table to mysqldump when making the backup, all database tables will be deleted before restoring.

Other switches that may be of interest include:

Here is a crontab example:

0 5 * * * mysqldump –user=shortstat –password=NotTelling –opt blogstats > /var/www/blogstats.sql

mysqlhotcopy is a Perl script that physically copies files. MyISAM tables are stored as files, and can therefore be backed up and restored simply by copying the data, index and structure files. This is much quicker than mysqldump, but does not work for InnoDB tables, which are not stored individually as files. When copying files, you need to ensure that either the server is offline, or the tables are locked to ensure a consistent backup. mysqlhotcopy does all of this for you by obtaining a read lock on all the tables to be backed up, then copying the tables and releasing the lock. It is a good option for small, live databases, but not for larger, heavily used databases. mysqlhotcopy works as follows:

$ mysqlhotcopy dbname filepath

There are a number of prerequisites to running mysqlhotcopy.

To restore from mysqlhotcopy, simply copy the files back to the data directory:

$ cp filepath_filenames datadir_filepath

Search

/

Search the file downwards for the string specified after the /

?

Search the file upwards for the string specified after the?

n

Repeat last search given by ‘/’ or ‘?’

N

Repeat the last search given by ‘/’ or ‘?’, except in the reverse direction

F

Search the current line backwards for the character specified after the ‘F’ command. If found, move the cursor to the position

f

Search the current line for the character specified after the ‘f’ command. If found, move the cursor to the position

;

Repeat the last f or F

Replace

[address] s [/pattern/replacement/] [options] [count]

Address can be empty (current line only), or a range designated as follows:

Options

c Prompt for confirmation before each change
g Substitute all instances of pattern on each line
p Print the last line on which a substitution was made
i Ignore case for the search pattern

Examples:

:s/foo/bar/g
Replace every occurrence of the word foo with bar on current line.

:%s/foo/bar/c
For each line on the file, replace the first occurrence of foo with bar and confirm every substitution.

:1,10s/yes/no/g
Substitute on first 10 lines.

:%s/[Hh]ello/Hi/gc
Confirm global substitutions.

:%s/foo/bar/gi
Ignore the case of the pattern you want to substitute. This replaces foo, FOO, Foo, and so on.

:s/Fortran/\U&/ 3
Uppercase first instance of “Fortran” on next three lines.

:g/^[0-9][0-9]*/s//Line &:/
For every line beginning with one or more digits, add the “Line” and a colon.

The idea behind iTunes “genius” feature is to scour your music database and queue up tracks similar to the one you are currently listening to. How well it does this is debatable. iTunes seems to cook up it’s “similarity” data based on the listening and purchasing habits of it’s users. Interestingly, if everyone’s listening habits were completely eclectic, and if they always kept their player on random, there would be nothing to go on. Fortunately, this isn’t the case.

Sometimes these genius playlists are spot on, other times they are wildly divergent. Either way it’s kind of fun.
Unfortunately iTunes doesn’t make their “genius” info available via web services, so my plugin leverages Last FM’s web services instead, which will give you a list of similar tracks for any track, as well as a list of similar artists. Once again, results vary.
Here’s an example:

>>> Looking up “When We Get Famous” by Math And Physics Club
Merging with current Playlist

Found 149 Similar tracks according to Last FM
Found track: Mars by The Lucksmiths
Found track: Young and Dumb by The Lucksmiths
Found track: Get It Right by The Fairways
Found track: Twee by Tullycraft
Found track: Leaders of the New School by Tullycraft
Found track: Emblematic by Pipas
Found track: Run Run Run by Pipas
Found track: Popkiss by Blueboy
Found track: Talulah Gosh by Talulah Gosh
Found track: One Two Three Four by Acid House Kings
Found track: Springtime Sunshine by The Shermans
… and so on.

Here’s a screen shot:

I’ve tried this on J. River Media Center versions 12 and 13, and also on Jukebox. Here’s the installer:
J. River SmartAlec Plugin v. 0.1

Here’s the source code as well, in case anyone wants to poke around. I did this in MS’s Visual C# Express Edition. If you do anything cool or useful please send me a copy. The project uses Martin Lottering’s CheckBoxComboBox control which can be found on CodeProject, as well as an XML-RPC wrapper which can be found here.

J. River SmartAlec Plugin v. 0.1 source

The podcast helped me to conceptualize the issue in terms of causation. I’ll outline two different stances, one spiritual and the other materialistic. I won’t pretend to get either right. It’s interesting to note up front that historically there have been theological stances which do not believe in free will, Calvinism being one example. I think the issue partly being the difficulty in reconciling individual free will with the idea of God’s omnipotence.

When a conscious agent decides to take some action, how do you conceptualize the chain of causation that preceded that action? In the pro free will stance the chain ends in a kernel of free agency which one might call, alternatively, the self, soul, or spirit, and this kernel operates outside any physical laws that operate on the material world. In the Christian view the kernel was forged by God in His own image, and is usually considered to be immortal. The self becomes a well spring of action in the world that operates from outside the world. A kernel of agency unfettered by physical laws, but which is capable of generating a spark of action within the world. The first, uncaused node in a chain of causation. This freedom from the physical world adheres only in ones mental life, of course. In the attempt to exercise our will we are brought once again into the physical and material world.

Alternatively, in the scientific or materialist view, the chain of causation for every action essentially has one well spring… the big bang. Thoughts and decisions are synonymous with neural activity in the brain. The more rational the thought or decision, the more the pre-frontal cortex in involved, hopefully, but either way it all amounts to neural activity. This neural activity, in turn, was caused by outside stimulus, or perhaps internal biochemistry. It’s hard not to resort to the computer as an analogy here, and conceptualize the thoughts as algorithmic in nature. The algorithms and networks themselves are built upon templates ironed out over the course of evolutionary history. The experiential and developmental history of the organism fine tunes the algorithms, as well as creating a repository for data, in the form of memories, which might be used in making future decisions. Of course this explanatory path is no where near complete, scientifically, but in principle the chain of causation becomes a massively tangled web, winding it’s way back in time to the beginning of our universe.

This is a radically different notion of the self. One in which we are riders on a causal trajectory which started with the big bang. At the end of the podcast, Ginger, who seems to more or less agree with the materialistic notion of free will outlined above, makes a defense for the idea of personal responsibility, but I can’t help thinking that from within that framework such notions are simply useful fictions. At the very least, notions such as blame, praise, responsibility, guilt, etc… loose some of their gravity.

This isn’t necessarily a bad thing, but it’s a different way of thinking about the world and the self. Different, but certainly not new. Diderot was already writing about these things in works like D’Alembert’s Dream way back in 1769.

Anyway, I haven’t read much about flow but I imagine it’s an idea that is not typically called upon to illuminate the motivations and psychology of those striving for goals that don’t match up with societies.

You will want to first install all the helper apps. On Debian or Ubuntu :
apt-get install mp3info unrtf antiword xpdf-utils html2text pstotext dvi2tty cabextract

Then download the tarball, run make, and set up your .bash_profile file properly. I added the following line to my .bash_profile:

export LESSOPEN=”|/usr/local/bin/lesspipe.sh %s”

Now “re-source” this file from your prompt (. ~/.bash_profile) and try things out. To view a file from an archive use less archive.tgz:docs/README for example. To turn on the syntax highlighting feature use less -R mycode.c.