The preferred noisemaker of the 2010 World Cup: the vuvuzela. For some reason these were distributed to fans attending World Cup matches, ruining the match for everyone watching at home.

When played by a stadium full of people the vuvuzela sounds like angry bees. I'm always up for a small project to help improve (or build) my BlackBerry development chops, and a vuvuzela app seemed like a perfect candidate. There are a slew of vuvuzela apps for the iPhone, but I only saw one other vuvuzela app for the BlackBerry, and that one only has a single vuvuzela sound.

Mine? It has 7.

I built the app with the 5.0 JDE. I tested it on a Storm2 9550 and Bold 9700 both running OS 5 and an 8820 running 4.5 and it worked on all three. If it doesn't work for some ancient OS, get a new phone.

Usage

Visit http://j.mp/cfg-vuvuzela using your BlackBerry to install the app. On newer devices the icon can be found in your Downloads folder.

Scroll and click a button to play the sound, or press the keys 1-7. You don't need to hold the ALT down; pressing W is the same as pressing 1. I haven't tested this on a SureType (Pearl) device.

Known Issues

• Multiple audio threads – This is a J2ME limitation. I use javax.microedition.media.Player to play the audio, and you can only have one Player instance on CDMA devices and two on GSM. I read this on the BlackBerry support forums, and confirmed with a member of RIM's media development team.
• Initial launch – the first time you launch the app on OS 5 it can take up to two minutes before the screen displays. Oddly enough it launches immediately on an old 8820 running OS 4.5. I'm looking into why this happens.
• Volume control – I'm actually not sure where the volume setting comes from, I just play the sound and it works, and it's loud. The way the VUVUZELA should be.
• The folks over at SistMan Software also have a vuvuzela app, and there is a conflict since both of our primary modules are named Vuvuzela (Vuvuzela.cod, etc). Until that's fixed, you can't have both apps installed. I like to think mine is better since it has more sounds, but they launched first and I'll publish a minor update with a unique module name.

Download

Download it OTA – Please link to this page, and not the .jad directly.

It took me much too long to figure out that it wasn't easy (or maybe even possible) with Automator, even though it's simple to do from the command line.

Each Automator action returns a result. That result is passed to the next item in the chain, assuming the item is accepting input. The key is to chain a series of Get Value of Variable calls together, which are passed to a shell script as $1-$n and the special $@ variable. I've written a short primer on using Arrays in Bash that may be helpful.

A crappy picture is worth at least 68 words, so here is what a sample workflow looks like:

You can download the sample workflow here.

If you want to loop through selected Finder items one-at-a-time, the best way is to use Nyhthawk Productions's excellent Dispense Items Incrementally action.

array.sh

#!/bin/bash
 
echo "<<< Load a file into an array"
echo "# Set the IFS (Internal Field Separator) to a newline
IFS='
'
# Load file.txt from the current directory
arr=( \$( < file.txt ) )
"
 
 
IFS='
'
arr=( $( < file.txt ) )
 
 
echo "<<< Addressing individual array elements"
echo "\${arr[0]} = ${arr[0]}" # the first line of the file
echo "\${arr[1]} = ${arr[1]}" # the second line of the file
 
 
echo ""
echo "<<< \${#VARNAME[@]} will always return the number of elements in an array"
echo "\$arr contains ${#arr[@]} (\${#arr[@]}) items"
 
echo ""
echo "<<< Loop through the array (\${arr[@]}), loading each item as \$foo."
num=1
for foo in "${arr[@]}" ; do
	echo "Loop iteration $num: $foo"
	num=$((num+1))
done
 
 
echo ""
echo "<<< Loop through the array, addressing each item with an index"
num=0
while [[ $num -lt ${#arr[@]} ]] ; do
	echo "Array index $num (\${arr[$num]}): ${arr[$num]}"
	num=$((num+1))
done

file.txt

file line 1
file line 2
file line 3
file line 4

Saving the two files above as array.sh and file.txt, and running array.sh yields:

$./array.sh 
<<< Load a file into an array
# Set the IFS (Internal Field Separator) to a newline
IFS='
'
# Load file.txt from the current directory
arr=( $( < file.txt ) )
 
<<< Addressing individual array elements
${arr[0]} = file line 1
${arr[1]} = file line 2
 
<<< ${#VARNAME[@]} will always return the number of elements in an array
$arr contains 4 (${#arr[@]}) items
 
<<< Loop through the array (${arr[@]}), loading each item as $foo.
Loop iteration 1: file line 1
Loop iteration 2: file line 2
Loop iteration 3: file line 3
Loop iteration 4: file line 4
 
<<< Loop through the array, addressing each item with an index
Array index 0 (${arr[0]}): file line 1
Array index 1 (${arr[1]}): file line 2
Array index 2 (${arr[2]}): file line 3
Array index 3 (${arr[3]}): file line 4

I needed an excuse to add my new signing to the JDE, and quickly threw together an app (with an ugly icon) to lock your screen.

ScreenLock OTA (.jad) Installation
ScreenLock Offline (.alx) Installation

And then came Snow Leopard, OS X 10.6. They included built-in IPSEC VPN, freeing users from the shackles of the horrible Cisco IPSEC client. But yet again, Apple decided that users don't need to copy and paste text into the password field.

For anyone with a OTP soft token (like a RSA SecurID), this is a pain in the ass. Half the benefits of generating a passcode on the same device that you're using it on, are not having to type it.

Enter AppleScript, everyone's favorite natural language that thinks it's acceptable to use apostrophes when referencing variables. I've never actually written an AppleScript longer than tell application "iTunes" to playpause so this was new ground for me. Joe mentioned he has a similar script for junky non-RSA soft token solution he uses, but he left the country immediately after taunting me with the prospect of getting the code.

If you're doing any moderately serious AppleScript development, I highly recommend Script Debugger from Late Night Software. It was invaluable to seeing what objects and properties were really available, and troubleshooting errors.

What it does

• Prompts for your SecurID PIN (you can save this in the script, I highly recommend you NOT do that)
• Launches the SecurID application (it if it's running it will quit it first), generates a token, stores the token in a variable, and then quits the SecurID application
• Kicks off the VPN connection process, types in the generated passcode when the window is available, and logs in.

If there's any type of a banner with terms to accept, it won't click OK there.

Download the script, view the source code and read more at the project page.

My first Seagal-themed project is off to a roaring start, and I built a custom DMG packaging script purely so I could force the beta testers to look at Seagal's (self described?) "striking and somewhat boyishly handsome" mug while they rush to install the latest build of my project.

Beta testers, this is what you have to look forward to for the next release!

It's useful for things like extracting the version number out of your Info.plist during a Run Script Build Phase.

PlistBuddy wasn't in my existing path, I found it at /usr/libexec/PlistBuddy

Example Usage

/usr/libexec/PlistBuddy -c "Print :CFBundleVersion" Info.plist

Outputs: 0.14, or whatever the value of the CFBundleVersion key in Info.plist is.

PHP Warning: mysqli::mysqli(): [2002] A connection attempt failed because the connected party did not (trying to connect via tcp://localhost:3306) in xxx.php on line 2
PHP Warning: mysqli::mysqli(): (HY000/2002): A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond.
in xxx.php on line 2

PHP bug #45150 suggests that this is due to issues resolving localhost on Windows. This is partially correct, more specifically it's an issue with localhost resolving to ::1 when IPv6 is enabled.

While it feels like Windows has shoehorned in IPv6 support, the blame seems to lie with MySQL, which won't support IPv6 until version 6.0.

I'm not sure why the timeout isn't instant, the port is closed, so it could be partially a problem with PHP or Windows.

> nmap -sT -p 80,3306,3389 -6 ::1
 
Starting Nmap 5.00 ( http://nmap.org ) at 2009-11-20 15:48 Mountain Standard Time
Interesting ports on cfg64 (::1):
PORT     STATE  SERVICE
80/tcp   closed http
3306/tcp closed mysql
3389/tcp open   ms-term-serv
 
Nmap done: 1 IP address (1 host up) scanned in 3.14 seconds

The Fix

Open up %windir%\system32\drivers\etc\hosts with a text editor and comment out the line that looks like:

::1             localhost

Prefix it with a #, like so:

#::1             localhost

Save it and your PHP/MySQL connections will immediately begin working. You could also use 127.0.0.1 in your connection string instead of localhost, but I didn't want to change code in innumerable files.

To view the Event Log on a Storm you need to switch to the Home screen, show the keyboard, and activate alt/num lock by holding down the '123' button. Once alt is 'locked', press ", /, ", /. That's double-quote, forward-slash, double-quote, forward-slash. On all the QWERTY devices you can type ALT+LGLG.

The OS released approved for release by Verizon (application version 5.0.0.328, bundle 508) doesn't have a Show Keyboard menu item on the home screen, leaving users with no way to view the event log or configure any event log filters.

I thew together a quick little application that does one thing, and one thing only, launch the Event Log Viewer.

OTA install: http://coreygilmore.com/ota/eventviewer/EventViewer.jad
Desktop Manager install: http://coreygilmore.com/ota/eventviewer/EventViewer.zip

One paragraph in particular jumped out at me, emphasis mine:

We worked closely with Apple to bring Latitude to the iPhone in a way Apple thought would be best for iPhone users. After we developed a Latitude application for the iPhone, Apple requested we release Latitude as a web application in order to avoid confusion with Maps on the iPhone, which uses Google to serve maps tiles.

So Google was working "closely" with Apple, built a native iPhone application¹ for Latitude, and then Apple essentially rejected it and told them to build a web application. Isn't building an entire application only to have it wholly rejected one of the leading complaints about the nightmarish App Store approval process?

1. I'd love to see a build of this leaked out so jailbreak users could run it

The links on Atomeo are dead, and the cached copy of the code threw some errors with JDE 1.6.  Eventually I stumbled across a post by Anand Rajasekar containing ExportPrivateKey.zip which can be used to export the private key from a JKS keystore.

The syntax is:

java -jar ExportPrivateKey.zip {keystore_path} JKS {keystore_password} {alias} {target_file}

java -jar ExportPrivateKey.zip web.keystore JKS KEYSTORE_PASSWORD bas bas.key

Related Resources

Mark Foster maintains an excellent wiki with resources on Keytool to OpenSSL and OpenSSL to Keytool conversion (among other things).

<cfscript>
wsConfig = {
 root = "https://example.com/baaws/",
 corews = wsConfig.root & "core/ws?wsdl",
 utilws = wsConfig.root & "core/wsutil?wsdl",
 dispatcherws = wsConfig.root & "dispatcher/ws?wsdl",
 emailexchangews = wsConfig.root & "emailexchange/ws?wsdl",
 emaildominows = wsConfig.root & "emaildomino/ws?wsdl",
 emailgroupwisews = wsConfig.root & "emailgroupwise/ws?wsdl"
};
</cfscript>

You're able to reference a member of the wsConfig struct before the struct is even created! At least before I'd expect it to be created. To accomplish the same thing (avoid using temporary variables to create the object, but define my root URL for reuse) in PHP (without using unset()) I'd do something like:

<?php
$wsConfig = array(
 root => 'https://example.com/baaws/',
);
 
$wsConfig = array(
 root => $wsConfig['root'],
 corews => $wsConfig['root'] . 'core/ws?wsdl',
 utilws => $wsConfig['root'] . 'core/wsutil?wsdl',
 dispatcherws => $wsConfig['root'] . 'dispatcher/ws?wsdl',
 emailexchangews => $wsConfig['root'] . 'emailexchange/ws?wsdl',
 emaildominows => $wsConfig['root'] . 'emaildomino/ws?wsdl',
 emailgroupwisews => $wsConfig['root'] . 'emailgroupwise/ws?wsdl',
);
?>

In this case the second time I reference $wsConfig I'm using the values from the first copy. The first copy isn't destroyed until after the second copy is set.

Performance Impact

The first question I have is if there is any performance impact to creating a large struct implicitly versus using a Java object. In Adobe ColdFusion it seems the largest struct you can implicitly create outside of a function is one with 16319 members. Inside of a function the limit appears to be 3451 members.

Build yourself a testcase using the following code, or download a pre-assembled testcase:

<cfsetting enablecfoutputonly="yes" showdebugoutput="no">
<cfscript>
writeoutput("&lt;cfscript&gt;<br />struct = {<br />");
for( x = 1; x LT 16319; x = x+1 ) {
	writeoutput("abc#x# = 'abc#x#',<br />" );
}
x = x+1;
writeoutput("abc#x# = 'abc#x#'<br />" );
writeoutput("};<br />&lt;/cfscript&gt;");
</cfscript>

I'm not sure if you can use Java to implicitly create a struct (coldfusion.runtime.Struct in Adobe CF) or implicitly create a java.lang.object. Implicit creation of a large struct is much faster than creating a struct repeatedly calling StructInsert() (download testcase).

1. OpenBD doesn't support implicit creation yet.

99% of these instructions apply to JBoss as well as the BlackBerry Administration Service (BAS) which is running on top of JBoss. I read a few guides on installing a SSL Certificate in JBoss, but ran into issues with each of them.

For anyone who only needs vanillla JBoss instructions I'll flag BAS-specific items that can be ignored. Some guides will have you extract the private key from your certificate and generate a PKCS12 keystore using openssl. I'm intentionally not taking that approach to avoid modifying the BAS server.xml.

Installation

Outline

The process goes like this:

1. Generate a new key, optionally create a new keystore
2. Generate a CSR
3. Purchase the SSL certificate
4. Prepare chained certificates – gather root certificate, intermediate certificates, reply
5. Import certificates
6. Install new certificate

Prerequisites

Everything will be easier if you keep all the related files in a single location initially.

1. Create a directory called "req" and open a new command prompt inside of it.
2. Choose a keystore name and certificate alias. In the examples I'm using web.keystore and bas respectively.
   (BAS only) the keystore name must be web.keystore
3. Locate keytool.exe in the JRE bin directory. On my server it was in C:\Program Files\Java\jre1.6.0_07\bin\keytool.exe
4. (BAS only)Back up the current keystore file located in <drive:\program files>\Research In Motion\BlackBerry Enterprise Server\BAS\bin\web.keystore
   Note: Most likely this is in <drive>:\Program Files\ unless you're on 64-bit Windows, in which case it will be <drive>:\Program Files (X86)
5. (BAS only)You'll need to know your current keystore password, use that whenever prompted for a password. If you want to set a new password you will need to change the cacerts keystore password in the BlackBerry Server Configuration.
   

Step 1 – Generate the Key

keytool -genkey -alias bas -keyalg RSA -keystore web.keystore

This will generate a new key with the alias bas and create a new keystore in web.keystore (if it doesn't exist). (BAS only) When prompted for a password us your current BAS cacert password.

Press RETURN when prompted for a key password to use the keystore password. Replace bas.example.com with the hostname you are purchasing a SSL certificate for. Replace bas and web.keystore with the keystore name and certificate alias you chose in the Prerequisites section.

C:\req>keytool -genkey -alias bas -keyalg RSA -keystore web.keystore
Enter keystore password:
Re-enter new password:
What is your first and last name?
  [Unknown]:  bas.example.com
What is the name of your organizational unit?
  [Unknown]:  bas.example.com
What is the name of your organization?
  [Unknown]:  bas.example.com
What is the name of your City or Locality?
  [Unknown]:  Denver
What is the name of your State or Province?
  [Unknown]:  CO
What is the two-letter country code for this unit?
  [Unknown]:  US
Is CN=bas.example.com, OU=bas.example.com, O=bas.example.com, L=Denver, ST=CO, C=US correct?
  [no]:  yes
 
Enter key password for
        (RETURN if same as keystore password):

Step 2 – Generate the CSR

Next you must generate a Certificate Signing Request (CSR) to submit to Go Daddy (or whoever you've chosen).

keytool -certreq -keyalg RSA -alias bas -file bas.example.com.csr -keystore web.keystore

Provide your keystore password when prompted. This will generate a CSR from web.keystore for the bas entry.

The CSR should look something like:

-----BEGIN NEW CERTIFICATE REQUEST-----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==
-----END NEW CERTIFICATE REQUEST-----

Step 3 – Purchase the SSL Certificate

Open the generated CSR up in a text editor( eg, notepad bas.example.com.csr) and paste the contents into the CSR Data field on Go Daddy.You're given the choice of which issuing organization to use (Starfield Root or Godaddy Root), and the Web Server Type. I chose Starfield Root. The Web Server Type must be Tomcat.

Once you've verified the ownership of the domain you will receive a zip file containing the SSL certificate and the intermediate certificates.

Step 4 – Prepare chained certificates

Extract the contents of the the zip file (mine was named certbundle.zip) into your req working directory. If you chose Starfield Root for the issuing organization you should have the following files in req:

• web.keystore – Your keystore file
• bas.example.com.csr – Your generated CSR
• bas.example.com.crt – Your certificate response
• sf_bundle.crt (or gd_bundle.crt) – Bundled intermediate certificates
• sf_intermediate.crt (or gd_intermediate.crt) – Intermediate certificate
• sf_cross_intermediate.crt (or gd_cross_intermediate.crt) – Cross intermediate certificate

I'm not sure if this step is 100% necessary, but I did it and my certificate worked. Download the appropriate root certificate for your issuing organization and save it into your req working directory – Starfield Root (sfsroot.crt) or Go Daddy Root (gd-class2-root.crt).

Step 5 – Import Certificates

Once you've gathered the various certificates they must be imported into your keystore. The root and intermediate certificate are used to establish a chain of trust which eventually ends with your self-signed keystore.

1. Import the root certificate (sfsroot.crt or gd-class2-root.crt)

   keytool -import -alias root -keystore web.keystore -trustcacerts -file sfsroot.crt

2. Import the cross intermediate certificate (sf_cross_intermediate.crt or gd_cross_intermediate.crt)

   keytool -import -alias cross -keystore web.keystore -trustcacerts -file sf_cross_intermediate.crt

3. Import the intermediate certificate (sf_intermediate.crt or gd_intermediate.crt)

   keytool -import -alias intermed -keystore web.keystore -trustcacerts -file sf_intermediate.crt

4. Import your reply. The alias you use here MUST match the alias you used when you generated the certificate in Step 1.

   keytool -import -alias bas -keystore web.keystore -trustcacerts -file bas.example.com.crt

   The output should be similar to:

   C:\req>keytool -import -alias bas -keystore web.keystore -trustcacerts -file bas.example.com.crt
   Enter keystore password:
   Certificate reply was installed in keystore

Your request is now complete, and your keystore contains a trusted, signed request. You can verify this:

C:\req>keytool -list -keystore web.keystore -storetype jks
Enter keystore password:
 
Keystore type: JKS
Keystore provider: SUN
 
Your keystore contains 1 entry
 
bas, Jun 2, 2009, PrivateKeyEntry,
Certificate fingerprint (MD5): F4:63:44:DA:33:BA:C2:2A:C3:0E:6C:92:AB:E9:45:12

Step 6 – Install the Certificate

JBoss

Install and activate your new keystore using the instructions on the JBoss SSL Setup wiki page. If you've followed all the steps properly you will have a JKS-type keystore, not PKCS12.

BAS

1. Back up your web.keystore file if you haven't already.
2. If the new keystore has a different password than the current keystore change the password in the BlackBerry Server Configuration tool, under the Administration Service – Cacerts keystore tab.
3. Stop the BAS services – BlackBerry Administration Service – Native Code Container and BlackBerry Administration Service – Application Server.
4. Rename <drive:\program files>\Research In Motion\BlackBerry Enterprise Server\BAS\bin\web.keystore to web.keystore.original
5. Copy web.keystore from your req working directory into the BAS\bin directory.
6. Restart the BAS services – starting the Application Server service will automatically start the Native Code Container service.
   

Browse to your BAS using HTTPS and verify the server is working correctly. To undo the changes stop the BAS services, delete web.keystore and rename web.keystore.original to web.keystore. Restart the BAS services, which are now using the original, working keystore.

Troubleshooting

• When importing your reply you must use the same alias as you did to generate the original key which was used to generate the CSR.
• The key password must match the keystore password.
• (BAS only) The keystore filename must be web.keystore
• (BAS only) The keystore password must match what is entered in BlackBerry Server Configuration > Administration Service – Cacerts keystore
• (BAS only) Open the highest-numbered <servername>_BBAS-AS_01_YYYYMMDD_00##.txt log file and search for error, ssl and keystore for information about why the server didn't start. Most commonly this will be a password mismatch between the keystore and the BAS or a missing certificate.
• Verify the keystore using keytool -list.
• If the keystore is valid but the certificate is shown to be issued by bas.example.com instead of Go Daddy or Starfield you most likely did not import your reply (Step 5, keytool -import … bas.example.com.crt) using the same alias as the generated key (Step 1, keytool -genkey).

Resources

The following sites were helpful:

• Go Daddy Tomcat SSL Installation instructions
• keytool documentation
• JBoss SSL Setup
• How to install a Go Daddy Wildcard Certificate on JBoss
• Import a new SSL certificate for the BlackBerry Administration Service and BlackBerry Web Desktop Manager
• Go Daddy Certificate Repository
• Portecle – A wonderful Java GUI application for managing keystores and certificates.

I suspect that no-one at Apple knows how genuinely torturous the app store approval process is for developers personally after a rejection. When they hold the key to the only distribution pipe for something you’ve spent a lot of your time on – in my case a year – something you’re hoping could provide you with a livelihood – and polite email enquiries are not replied to – not even with an autoresponder, it is extremely frustrating. I don’t think I’ve ever felt as powerless in my life (and I’ve had to deal with US immigration authorities…).

I suspect quite a few people at Apple know what a nightmare the App Store approval process is, but I doubt a single one of them is in a position to do anything about it. Apple Inc. is no more likely to publicly acknowledge a problem than the Catholic church, and the App Store approval process is Apple's abortion is sure to be a sensitive topic.

When the basis of approval is subjective, you need talented, intelligent and rational people making decisions and passing jdugement. Good luck finding that combination in someone who is willing to earn $30k/year to run iPhone apps all day searching for boob pictures or profanity.

Objectionable Content

For your enjoyment, here are a few of the titillating terms that Apple objected to. I've no doubt that by posting this text my site will be masturbation fodder for teens everywhere.

• Auparishtaka or mouth congress – Another way of saying fellatio, but I can see how the term mouth congress is totally hot.
• Lingam – Penis. It's about time someone came up with a non-scientific way to refer to the male genitalia.
• Yoni – Vaginia – definitely the most offensive 4-letter word you can use to describe one.

Fill in the Blanks

Sometimes it's what is left out that counts. See if you can guess what the missing words are.

Assets

I've removed three words from the following paragraph.

The qualities of good ______(1.1) are that they should be bright, well set, clean, entire, convex, soft, and ______(1.2) in appearance. ______(1.1) are of three kinds according to their size:

Small
Middling
Large

Large ______(1.1), which give grace to the ______(1.3), and attract the hearts of women from their appearance, are possessed by the Bengalees.

Good Qualities

5 words this time, have at it.

The qualities of good ______(2.1) are as follows: They should be equal, possessed of a pleasing brightness, capable of being ______(2.2), of proper proportions, ______(2.3), and with ______(2.4) ends.

The defects of ______(2.1) on the other hand are that they are blunt, protruding from the ______(2.5), rough, soft, large, and loosely set.

There you have it. Smut! Pornography! It's obvious that this needs to be kept out of the hands of young children, so they can safely peruse Google Images with their morality intact.
a

Answers

Fill in the blanks or use your imagination.
Assets
1.1: nails
1.2: glossy
1.3: hands

Good Qualities
2.1: teeth
2.2: coloured
2.3: unbroken
2.4: sharp
2.5: gums

I'm not sure what percentage of my time I spend working with ColdFusion, but whenever I do I always run into arbitrary limitations or inconsistencies that are incredibly obnoxious. It gives ColdFusion the feeling of a language that was started by one group of people, picked up half-finished by another group with many features are implemented very shallowly in order to meet a ship date. And all too often Adobe never looks back to address those flaws. Hopefully the CFML Advisory Committee will address these flaws.

A great example blatantly rushing a feature is the implicit creation of structs and arrays. With the initial release of ColdFusion 8.0 you could use { } and [ ] notation to create both inside of a CFSCRIPT block, but you couldn't nest them until the release of ColdFusion 8.0.1

<cfscript>
// This only works in Adobe ColdFusion 8.0.1+
someStruct = {
 key1 = {
  subKey1 = 'xyz'
 }
};
</cfscript>

I'm hesitant to email my entire wishlist the Committee as Sean suggested, but I did want to write it out and solicit some feedback. Some of these are specific to Adobe ColdFusion (such as var declaration positions), and there is the possibility that some of the issues are fixed or under consideration to be fixed in the upcoming release of CF 9 Centaur.

ColdFusion/CFML Wishlist (or gripes)

• Consistency between cfscript, cfml functions. cfquery, cfdump, and all of the other functions missing from CFSCRIPT
• Allow var declarations anywhere (Note: Supported in OpenBD)
• Fully document CFQUERYPARAM type mapping (cf_sqltype_* to actual SQL type)
• Fully document the SQL commands/functions available for CFQUERY Query-of-Queries
• Ordered structs – I can't think of a good reason not to preserve the order of items.
• Implicit creation of structs with case sensitive names and names containing spaces

  <cfscript>
  struct = {
   'one key' = 'abc',
   'mixedCase' = 'def'
  };
  </cfscript>

• Support a more standard struct declaration format.

  <cfscript>
  struct = {
  	key: 'val',
  	orKey =&gt; 'val'
  };
  </cfscript>

• Pass implicitly created objects as function arguments.
• Allow [] notation to append to arrays.

  <cfscript>
  someArr = [ 'one, 'two', 'three' ];
  someArr[] = 'four';
  // someArr is now [ 'one, 'two', 'three', 'four' ]
  </cfscript>

• NO WHITESPACE RESIDUE FROM ANY CFML CODE – Seriously, don't leave a mess when you're parsing. Every byte counts.
• Defined way to pass any variable by reference
• Allow optional args and default values in cfscript function declarations
• True eval() function. I want to be able to dynamically parse code. DE and EVALUATE are about as healthy as cigarettes.
• Public/private key encryption added to cfencrypt w/o needing Bouncy Castle. OK, this one is probably just for me :)
• Zero-based arrays. I bet people that like 1-based arrays also hang their toilet paper under instead of over. Savages.
• A better way to handle caching, such as named caches so that you can expire individual caches easier if/when they need to change. (Note: Supported in OpenBD and Railo)
• Dynamic function calls. Also check out the PHP call_user_func_array function.

  <cfscript>
  function printPrefix(str) {
  	writeoutput("PREFIX: #str#\n");
  }
  func = "printPrefix";
  #func#("test");
  // Outputs:
  //  PREFIX:TEST
  </cfscript>

• Ternary operator. WTF? Anyone who is thinking about suggesting IIF should stop right now. IIF doesn't even come close to being true ternary operator.
• Improved For-In loops – Loop through structs with keys/values, like foreach in PHP

  <cfscript>
  for( key =&gt; val IN someStruct ) {
  	writeoutput("someStruct[#key#] = #val#\n");
  }
  </cfscript>

• StructValueArray – Return an array of all the values from a struct (just as StructKeyArray returns an array of all the keys). (Note: Matt has requested this for OpenBD.)

In a perfect world things like zero-based array indexes and ordered structs could be set at the application level to maintain backwards compatibility, but I'd even settle for server-wide.

Leave a comment if I'm missing compatibility for one of the CFML engines and I'll update the post accordingly.

The VAR parameter to the renderOutput function is required but was not passed in.

I didn't have any functions named renderOutput.  Nor did I have any arguments named VAR.

Apparently this is the error you get when you've got CFDUMP call inside of a function with output="no".

Did I say apparently? I meant obviously.

Error Executing Database Query.
[Macromedia][SQLServer JDBC Driver]Unsupported data conversion.

Relax, your code is (likely) fine.  This typically happens when you add a column to your table and use "SELECT *" in the query.  Somewhere along the line the list of columns is cached and since the list of columns you're selecting in the query hasn't changed ColdFusion chokes when what it receives doesn't match what it expects.

The fix couldn't be simpler – just toss a semi-colon at the end of your CFQUERY and watch your problems go away.  If you make another change just remove the semi-colon.

So a query that looks like this:

<CFQUERY NAME="qryUsers" DATASOURCE="dsn">
	SELECT * FROM users
</CFQUERY>

Becomes:

<CFQUERY NAME="qryUsers" DATASOURCE="dsn">
	SELECT * FROM users;
</CFQUERY>

I fully expect that I'll be stumbling across this page in 6 months.

/**
 * Out a MySQL-style table of data
 *
 * @param	array  $data         Associative array of data to output. 
 * @param	array  $header_keys  Optional; Assoc array of display names to use for headers. Keys must match those of $data. Defaults to keys of $data.
 * @param	string $glue         String to join lines with, defaults to newline.
 * @return string
 *
 */
function build_table($data, $header_keys=array(), $glue="\n") {
  $table = '';
  $data_fmt = array();
  $divider_row = array();
  $header_fmt = array();
 
  $keys = array_keys($data[0]);
  $col_lengths = array_flip($keys); // used to determine the max column width
 
  if( empty($header_keys) ) {
    $header_keys = array_combine($keys, $keys);
  }
 
  // set the base max length to the length of our header keys
  foreach( $keys as $key ) {
    $col_lengths[$key] = strlen($header_keys[$key]);
  }
 
  foreach( $data as $row ) {
    foreach( $keys as $key ) {
      $col_lengths[$key] = max($col_lengths[$key], strlen($row[$key]));
    }
  }
 
  foreach( $keys as $key ) {
    $data_fmt[] = '%-' . $col_lengths[$key] . 's';
    $header_fmt[] = '%-' . $col_lengths[$key] . 's';
    $divider_row[] = str_pad('', $col_lengths[$key]+2, '-'); // fill the spacing
  }
  $data_fmt = '| ' . implode(' | ', $data_fmt) . ' |';
  $divider_row = '+' . implode('+', $divider_row) . '+';
  $header_fmt = '| ' . implode(' | ', $header_fmt) . ' |';
 
  // assemble the table
  $table .= $divider_row . $glue;
  $table .= vsprintf($header_fmt . $glue, $header_keys);
  $table .= $divider_row . $glue;
  foreach( $data as $row ) {
    $table .= vsprintf($data_fmt . $glue, $row);
  }
  $table .= $divider_row . $glue;
 
  return $table;
}

The output will be similar to:

+--------------------+--------------------+---------------+
| First              | Second col         | Random Column |
+--------------------+--------------------+---------------+
| 60386980488376     | 597655305189574649 | 51            |
| 959432             | 459191             | 396802        |
| 73874213           | 570702             | 771           |
| 144579584678722975 | 892300317939345    | 59180         |
| 293172             | 314                | 127725766727  |
+--------------------+--------------------+---------------+

It only works with associative arrays, but that's easy enough to fix. You can proudly test it out using this glorious snippet:

$data = array();
 
function get_rand() {
  $max=rand(1,6);
  $str='';
  for( $i = 0; $i < $max; $i++ ) {
    $str .= rand(10, 1000);
  }
  return $str;
}
 
for( $i = 0; $i < 5; $i++ ) {
  $row = array(
    'one'   => get_rand(),
    'two'   => get_rand(),
    'three' => get_rand(),
  );
  $data[] = $row;
}
 
$disp = array(
  'one'   => 'First',
  'two'   => 'Second col',
  'three' => 'Random Column',
);
 
echo build_table($data, $disp);

There are two bugs in the current release of DU Meter (4.0.3076) that really aggravate me.  The first is position related; the window remembers it's last position, and doesn't check to see if that might be offscreen.  If it remembered it's position relative to the nearest corner this wouldn't be an issue, instead if I go between my 15" laptop screen and a 30" LCD it's either right in the middle of the window or hidden.  The second bug involves the display of the numeric graph.  Every time you open the Options window it resets the display mode setting to "Graphical and numeric display".

I was thrilled to see that there was a Vista Sidebar Gadget available (sadly only for 32-bit Windows), and hoped that it would offer a numeric-only display.  By virtue of being in the sidebar I wouldn't have to worry about positioning either.

Nope.  Only the combination graph and numeric display.  I can't fix the app, but I've got some control over the gadget.  I grabbed uncompressed gadget directory (%Program Files%\Windows Sidebar\Shared Gadgets\DUMeter.gadget) and began editing.  The background was a fixed size so I created a new one that was the correct height of the numeric display.  The stats window comes from a COM control which I very hackily modified the height of, to make it fit my new background.

The new gadget is supposed to be 28 pixels high, but there's an undocumented minimum height of 57 pixels for gadgets.  This is due to the height of toolbar being 57 pixels.  As a result of this the gadget will look a little odd if it's not at the bottom of the Sidebar – it's 28 pixels of content followed by 29 pixels of empty space.

Adding the DI Meter and DU Meter Mod Gadget

The modified and original DU Meter gadgets

You can download the gadget from my DU Meter Gadget project page.

Geoffrey McGill, one of the Datejs developers, had posted some code in late 2007 to get Datejs working with an older version of Ext, but it didn't work with the latest release of Ext, version 2.2.

It didn't take long to update Geoffrey's changes to work with Ext 2.2, and I've put a demo online along with the Ext override code.

The code will override properties of all Ext.form.DateField objects and allow you to use the Datejs syntax. If this isn't the behavior you want you can easily extend and create your own class using Ext.extend.