The sections we’re really interested in are 3.1.1 and 3.1.2. These are detailed below.

Regarding production software on a per-CPU and Virtual Machine (VM) basis:

3.1.1 Adobe grants Licensee a license to install and use the Software as Production Software on a per-CPU basis as provided in a separate writing. The total number of CPUs of all of the Computers on which the Software operates may not exceed the total number of CPUs separately licensed. For purposes of this License Metric, (a) all CPUs on a Computer on which the Software is installed shall be deemed to operate the Software unless Licensee configures that Computer (using a reliable and verifiable means of hardware or software partitioning) such that the total number of CPUs that actually operate the Software is less than the total number on that Computer, and (b) a CPU shall mean a single, physical chip with the capability of housing multiple processor cores. If Licensee purchases a 2-CPU Production Software license, then each CPU to which it applies such license must reside in the same Computer. In the event that Licensee desires to apply 2-CPU Production Software licenses to Virtual Machines, then (c) for the Standard version of the Software, the number of 2-CPU licenses required shall be the greater of (i) the number of available physical CPUs for all instances of the Software divided by two (any fractions shall be rounded up for purposes of this provision), or (ii) the total number of Virtual Machines on all Computers on which the Software is installed, and (d) for the Enterprise version of the Software, the number of 2-CPU licenses required shall be the number of physical CPUs on which the Software operates divided by two (any fractions shall be rounded up for purposes of this provision).

The key information here is:

• The Standard license is charged per 2-CPU or per-VM, which ever is greater.
• The Enterprise license is charged per 2-CPU.

But what does this actually mean for Joe Bloggs? Fortunately the guys at Adobe, via Twitter, were on hand to help.

Here is a little table that should whet your appetite:

Regarding production software on a cloud network:

3.1.2 Adobe grants Licensee a license to install and use the Software as Production Software on a Cloud Network on a per-instance basis as provided in a separate writing. The total number of instances that may run on a Cloud Network may not exceed the total number of instances separately licensed. Licensee must purchase and apply 1 instance license to each use of the Standard version of the Software. Licensee may apply 10 instances for each license purchased of the Enterprise version of the Software.

The key information here is:

• The Standard license supports cloud computing, but for each instance, 1 license must be applied.
• The Enterprise license supports up to 10 cloud instances, e.g. on Amazon EC3, Azure, Joyent, Rackspace Cloud etc.

In essence, if you’re going to run a [decent] virtual machine environment, or indeed a cloud environment, it would be far more cost efficient to run ColdFusion Enterprise, since it is costed by the number of virtual machines or increments of 10 cloud instances.

You can download the original Adobe ColdFusion 9 EULA (PDF 400K) from the Adobe website.

UPDATE: You can download the beta project from the Google Chrome website.

Instead of me talking you through the project, Google and Scott McCloud have put together a cool little cartoon.

However, as a brief summary:

• Google Chrome is Google’s open source browser project.
• The browser will use the popular Webkit HTML rendering engine used in Safari and Adobe AIR.
• The browser will include a brand new JavaScript Virtual Machine called V8.
• The browser will include Gears to allow developers to enhance the user experience.
• Google Chrome will use special tabs, like more traditional browsers, but set above the address and menu bar.
• Each browser tab will run on its own process. If one tab fails for some reason, the whole browser will not need to be restarted, losing valuable work or tabs. This is similar to functionality found in Internet Explorer 8.
• The browser has an address bar which includes a more intuitive auto-completion feature called ‘omnibox’. It is said to be less ‘irritating’ than current auto-complete/suggest functionality common to Firefox 3 or Google Suggest.
• As a default homepage Chrome presents you with a kind of speed dial feature, similar to the one found in Opera.
• Chrome has a privacy mode, which allows you to create an incognito window and nothing that occurs in that window is ever logged on your computer. Again, this is similar to functionality found in Internet Explorer 8.
• Web apps can be launched in their own browser window without address bar and toolbar, much like Mozilla’s Prism project.
• To fight malware and phishing attempts, Chrome constantly downloads lists of harmful sites.

(Click on the images to see a larger view)

You can find out more information from Google’s blog post on the subject.

Performance refers to the response time of requests to CFML pages and Components (CFCs), whilst scalability refers to how the performance results change when the load on the server increases.

Factors Affecting Performance

• Hardware and software configuration
• Database and other external system requests
• Web server performance
• CFML application

Macromedia

• JRun 4.0: Configuring individual JVM settings for each JRun Server
• The Moment of Truth: Tuning Performance in JRun Apps and Environment
• JRun 4: General troubleshooting techniques for an unresponsive JRun server
• ColdFusion MX: Explanation of Request Timeout and threadWaitTimeout settings
• Server settings: the jrun.xml file
• ColdFusion MX: Tips for performance and scalability

Sun

• Frequently Asked Questions about Garbage Collection in the HotspotTM JavaTM Virtual Machine
• Tuning Garbage Collection with the 1.4.2 Java[tm] Virtual Machine
• Java Tuning White Paper
• jvmstat 3.0

Brandon Purcell

• ColdFusion Peformance Tuning Presentation
• Obtaining a Thread Dump with ColdFusion or JRun running as a Windows Service
• JRun Connection Closed Errors
• Performance Tuning and Monitoring Coldfusion or JRun
• Java Thread and How does it work?
• Advanced Debugging In JRun 4

Steven Erat

• Steven Erat’s blog: activeHandlerThreads or Simultaneous Requests: Less is More
• Simultaneous Requests in CFMX 7
• Keep Library Loaded – Check this box to retain the library in RAM
• SeeStack – Parse ColdFusion Thread Dumps Like a Pro!
• Request timed out waiting for an available thread to run
• Troubleshooting Tips For ColdFusion MX
• Help! My ColdFusion MX Server is Slow!!!!

Peter Freitag

• Peter Freitag’s blog: Tuning Garbage Collection Outline
• ColdFusion Garbage
• ColdFusion Memory Usage Stats
• Getting ColdFusion Server Metrics
• Have you tuned your JVM on ColdFusion MX yet?

Daemon

• Daemonite blog: Maximum JVM heap size for CFMX
• Enhancing CFMX logging

Others

• CFMX JVM Tuning – The difference between MaxPermSize and Heap Size
• Sarge’s blog: CFMX Performance Settings (JRun 50% CPU thread)
• Robi Sen’s blog: Hung Servers, Java.lang.OutOFMemory errors and Tuning CF JVM
• Misc: Java Garbage Collection Tuning
• Java Performance Tuning
• SeeFusion :: ColdFusion Server Monitoring