, then unless you had a *.properties file in your current directory, grep would assume that there are no other matches lower down in the structure, so you’d find nothing. Hmph.

So of course, find comes to the rescue. You can run find . -name *.properties -exec grep nifty.java.property {} \;

and that finds all your matches – with one problem: this doesn’t tell you which file it found the match in, so that’s pretty useless as it stands (in any case, remember to escape the ; with \ if you run this from the shell to prevent your shell from interpreting the ; as a command separator and not passing it to find).

So, there are a couple of solutions to this.

• find . -name *.properties -exec grep nifty.java.property /dev/null {} \;
  
  
  

The reason why this works: grep normally only shows the file names in which it matched a string if it has multiple files to process. Since the find command launches one grep per file name matched when it is ended with ; (in this case \;, otherwise the shell would interpret it instead of it being passed to find), grep assumes which file name you just called it with and therefore doesn’t spit out the file name.

So, adding the /dev/null means another file to be searched (which is always going to return 0 results), therefore grep decides it needs to show the file name, which is really cool, because otherwise find gives you no clue as to which file it executed grep for. You could of course pass -print to find, but that gives you every single file name matched and would be totally useless as you would potentially drown in matches.

This is probably the most portable option since it does not rely on any special options for grep or find.

• find . -name *.properties -exec grep nifty.java.property {} +
  
  
  

For the longest time I totally missed that most (if not all) find versions offer the option of ending the command with +. This has the effect of passing as many paths as will fit as an argument, i.e. we’re back to grep having multiple files to match and it prints the path name again. Neat!

• find . -name *.properties -exec grep -H nifty.java.property {} \;
  
  
  

Some versions of grep offer the -H option, which forces grep to act as if it was matching multiple files and it prints out the file name (with full path) once again.

So there you have it – three methods of getting the results you need. Happy grepping! ;]

PS: If find barfs on a whole bunch of “permission denied”‘s and spews out a bunch of extra stuff that prevents you from easily seeing the results you’re interested in, remember that you can add a 2> /dev/null

to redirect stderr to that great bit-bucket in the sky and get rid of all that annoying stuff.

That is, Until I found this post. Once again, the inimitable AutoHotKey comes charging in on its ugly white horse (the language is a bit of a pain ;] ) to the rescue!

Simply run a script as follows:

#space::SendInput !{Space}

Now put a link in your Start menu (which if you’re like me you spend a bit of time hunting around for in Windows 7 – you can find it under C:\Users\<username>\AppData\Roaming\Microsoft\Windows\Start Menu (seriously!? can you make this stuff harder to find please, Microsoft!?) and every time you boot Windows, your Win-Space with the IMHO useless “desktop peek” functionality is overridden to send an Alt-Space hotkey instead, which if you configure Launchy to listen for that, will bring up its sweet goodness, just like it used to! ;]

Enjoy – I know I will!

Wait! What was that? That looks like… nah, it can’t be! Or is it? Damn me to hell and back, that does look an awful lot like SpongeBob SquarePants – what’s he doing in “The Adventures of Buckaroo Banzai Across the 8th Dimension”!? Here’s a closer look:

Wow! Looks just like him, don’t you think?

What can I say except: wow!

’nuff said… ;]

I have to say, I was simply astounded at the complexity of the mechanism. I mean, sure, I knew that it was complex, but the intricacies of the mechanism floored me. I had no idea that such am everyday item possessed such engineering beauty!

First of all came the tray mechanism with all of its plastic-geared beauty… Well, maybe that bit wasn’t so beautiful after all… ;] However the worm gear drive and stepper motor was pretty cool.

The direct drive CD spinning motor was pretty cool, too – at least I think it is when you think about how fast it had to spin that little plastic platter.

The real marvel of the thing comes in the read-write head, however. That is a precision-cast little bit of aluminium (to keep it light) that is machined to some apparently pretty tight tolerances to keep the optical path accurate. This is wrapped in a miniature 3D circuit board whose equal I have yet to see elsewhere, including laptop designs. These were little bits of board connect with filigree ribbon (I hesitate to call them “cables”) that wrapped around all sides of the aluminium substructure.

Not stopping there, we have a wonderfully complex optical path that includes the laser diode, a detector and the lens that is suspended on the teeniest imaginable copper springs which also service as current-carrying conduits that energize the circuit boards on either side of the lens. These  circuits create electric loops which then help to drive it up and down inside the strong magnetic field generated by the two tiny rare-earth magnets on either side.

All in all a complete engineering marvel that I picked up for a measly $40.

I think I’m beginning to understand why these things cost over $1,000 at inception… ;]

Image processing in Mathematica

OK, yeah, I’m a huge nerd – what’s your point? ;]

1. Does it support jQuery easily enough (CakePHP uses Prototype/Scriptaculous as its default framework, but has provisions for using others)
2. Is the Ajax-y goodness baked in enough?
3. Will it really be flexible enough to accomodate my needs?

AS part of this I also ended up installing Xdebug, since that’s the default debugger for CakePHP. I must admit, the things I really like about it so far are:

1. Ease of install (instructions here got me up and running and a couple of tweaks to my Eclipse PDT setup (which supports Xdebug out of the box – nice!) found here and here got me the rest of the way). The tailored install instructions that analyze your php.ini, provide the correct download and customized install instructions is the coolest thing for a newb like me to get up and running quickly! ;]
2. Code coverage analyzer! I’m not sure if Zend Debug is able to do the same thing, but that’s way cool!

All in all, looks like I’ve found me a framework I can be happy with – finally! ;]