I've done a bit of work with the packages available in the JVM with both LylaCaptcha and Mach-II. First, you're mileage will vary on theme depending on the OS of the host system.
I can't speak directly for ImageMagick "speed", but it basically the default standard for image processing. ImageMagick is written in C so it will probably out perform Java. The JMagick or im4java interfaces for Java are just thin APIs to the command line version of ImageMagick. In general, Adobe CF image functions are the slowest out of the bunch of available image processing packages for CFML. I recently wrote some functionality to read image dimensions for Mach-II <view:img> custom tag. Adobe CF was the slowest at reading the dimensions, reading the file via java.awt.* was faster and the fastest (because it only reads metadata) is java.awt's default toolkit. However, the default AWT toolkit would cache the results of image dimensions until the CFML engine was restarted. We ended using the default AWT toolkit because it was crazy fast, but had to pass in the image file path with a tick count at end to the it stops caching the results. Calling the flush() method in the default toolkit did *not* actually flush the cache. I guess what I'm saying is if ImageMagick will add new features to Open BD -- I'm all for it. However, I suggest that Open BD should maybe think twice about adding 50+ image functions to the root namespace -- therefore polluting the function namespace. I suggest using the Image construct like image.resize(200, 200) instead of the Adobe CF way of imageResize(image, 200, 200). I guess it depends on how compatible you want to be. Best, Peter P.s. Let Open BD be the people's CFML engine. ;-) Alan Holden said the following on 10/03/2009 12:42 AM: > Any comments would be welcome - while we're on the subject - on imagecfc > (http://www.opensourcecf.com/imagecfc/) which as far as someone of my > intelligence can tell, just extends some classes that are already in the > JVM. > > Does OpenBD pretty much use the same or a similar thing? If one had > legacy code (CF6.1 for example) that uses imagecfc, how much efficiency > does one gain by converting this code to the native OpenBD cfimage tags? > I hear that it's fairly substantial in the case of CF8. > > ImageMagick is actually faster than the native CF tags and/or imagecfc? > > Well, you get what I'm asking in general... > > Alan K. Holden > > > > > --~--~---------~--~----~------------~-------~--~----~ Open BlueDragon Public Mailing List http://groups.google.com/group/openbd?hl=en official site @ http://www.openbluedragon.org/ !! save a network - trim replies before posting !! -~----------~----~----~----~------~----~------~--~---
