http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/ Techniques Behind Modern Web Wed, 27 Aug 2008 20:57:38 +0000 http://wordpress.org/?v=2.6.1 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-282 Close Look into “Include” JavaScript Compression Lib - A Client-Side Solution | Just Talk About Web Mon, 10 Mar 2008 06:30:48 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-282 [...] Mar A few weeks ago, I posted a comparison of Dojo ShrinkSafe, YUI Compressor and Jasob Obfuscator. Then, I received comment from Brian introducing an interesting project: Include; some day later it [...] [...] Mar A few weeks ago, I posted a comparison of Dojo ShrinkSafe, YUI Compressor and Jasob Obfuscator. Then, I received comment from Brian introducing an interesting project: Include; some day later it [...] ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-259 Ganesh Thu, 06 Mar 2008 04:34:11 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-259 Very good article. It would be great if the author can include dean edwards packer also as part of the benchmark. Very good article. It would be great if the author can include dean edwards packer also as part of the benchmark. ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-258 Jasob Support Wed, 05 Mar 2008 22:52:48 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-258 That's not true, Dave. In Jasob you can use various criteria to automatically mark or unmark certain names for obfuscation, including automatically unmarking global names (variables and functions) for obfuscation. There's no way an application could automatically detect which global variables or functions can be obfuscated and which cannot because there's simply no such information inside a single .js file as you are saying YUICompressor is doing. What do you mean by "worst results"? We don't know how the author of this review initially came up with obfuscated file size of 2,945 bytes. Worst results could be without obfuscating any name and without removing comments and whitespaces. At the end, any given .js file will be part of an application and then this distinction between pubic and private variable and function names would be irrelevant in the terms of the obfuscation process which will give you the opportunity to obfuscate all those public names as well and with Jasob you can do that. That’s not true, Dave. In Jasob you can use various criteria to automatically mark or unmark certain names for obfuscation, including automatically unmarking global names (variables and functions) for obfuscation. There’s no way an application could automatically detect which global variables or functions can be obfuscated and which cannot because there’s simply no such information inside a single .js file as you are saying YUICompressor is doing. What do you mean by “worst results”? We don’t know how the author of this review initially came up with obfuscated file size of 2,945 bytes. Worst results could be without obfuscating any name and without removing comments and whitespaces. At the end, any given .js file will be part of an application and then this distinction between pubic and private variable and function names would be irrelevant in the terms of the obfuscation process which will give you the opportunity to obfuscate all those public names as well and with Jasob you can do that. ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-257 Dave Sanderson Wed, 05 Mar 2008 16:24:26 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-257 The point is that YUICompressor can automatically detect what to obfuscate while Jasob requires you to do it manually. That's why the results could be the worst (2,945 bytes) or the most-optimized (of 2,243 bytes - but not the best for public variable and function names would be lost.) The point is that YUICompressor can automatically detect what to obfuscate while Jasob requires you to do it manually. That’s why the results could be the worst (2,945 bytes) or the most-optimized (of 2,243 bytes - but not the best for public variable and function names would be lost.) ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-253 Jasob Support Wed, 05 Mar 2008 06:52:39 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-253 That depends on your project and your preferences. By obfuscating all names (or as much as you can) you will not only make obfuscated file size smaller but overall obfuscation will also be better. Sure, if you have JavaScript library that you want to distribute to other developers then you may choose not to obfuscate certain public functions or global variables which at the same time doesn't mean that end user (developer) cannot obfuscate those public functions and variables as well. Regarding the summary in your review, part of the sentence "...and find YUI Compressor is the best one in both terms of compression ratio" is not correct. Thanks. That depends on your project and your preferences. By obfuscating all names (or as much as you can) you will not only make obfuscated file size smaller but overall obfuscation will also be better. Sure, if you have JavaScript library that you want to distribute to other developers then you may choose not to obfuscate certain public functions or global variables which at the same time doesn’t mean that end user (developer) cannot obfuscate those public functions and variables as well. Regarding the summary in your review, part of the sentence “…and find YUI Compressor is the best one in both terms of compression ratio” is not correct. Thanks. ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-252 Jimmy Wed, 05 Mar 2008 06:12:04 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-252 I don't think selecting everything for obfuscation is wise choice. You may want keep public function names as original, especially in case of SHA-1 lib. Of course, I know that if we use Jasob for a project, all other pages will "know" obfuscated function names so we can choose to obfuscate all. Correct me if I'm wrong. I don’t think selecting everything for obfuscation is wise choice. You may want keep public function names as original, especially in case of SHA-1 lib. Of course, I know that if we use Jasob for a project, all other pages will “know” obfuscated function names so we can choose to obfuscate all. Correct me if I’m wrong. ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-251 Jasob Support Wed, 05 Mar 2008 05:49:43 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-251 Thank you for updating test results. In the given example (index.html and sha1.js files) you can simply mark ALL names for obfuscation which would give you even better results (2,243 bytes for the obfuscated sha1.js file). Please remove or modify the following in your Jasob review as it is simply not correct: "...but my trial to compress a script file for SHA-1 hash generator resulted in a slight bigger file in comparison with the one compressed by YUI Compressor (2,945 over 2,596 bytes). However, I suppose it’s not too large different." and in the Cons section "...NOT the best compression ratio". Thank you for updating test results. In the given example (index.html and sha1.js files) you can simply mark ALL names for obfuscation which would give you even better results (2,243 bytes for the obfuscated sha1.js file). Please remove or modify the following in your Jasob review as it is simply not correct: “…but my trial to compress a script file for SHA-1 hash generator resulted in a slight bigger file in comparison with the one compressed by YUI Compressor (2,945 over 2,596 bytes). However, I suppose it’s not too large different.” and in the Cons section “…NOT the best compression ratio”. ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-249 Jimmy Wed, 05 Mar 2008 02:12:26 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-249 Thanks for pointing out the functionality. I did a new test: generating new names for all variables and private functions. My result is the obfuscated file size of 2,310 bytes. Jasob is WINNER! Thanks for pointing out the functionality. I did a new test: generating new names for all variables and private functions. My result is the obfuscated file size of 2,310 bytes. Jasob is WINNER! ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-248 Jasob Support Tue, 04 Mar 2008 21:23:50 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-248 Jasob has the BEST compression efficiency. The size of the obfuscated file (sha1.js) of 2,945 bytes is WRONG. The correct file size after obfuscating it with Jasob is 2.243 bytes which clearly makes Jasob a winner in this comparison. To achieve best compression ratio, names from both name lists should be first marked for obfuscation and then new names should be generated from the main menu (Obfuscate -> Generate New Names) or obfuscation process should be started (Obfuscate -> Obfuscate). Please update comparison table with correct obfuscated file size (2.243 bytes) for Jasob obfuscator. Jasob has the BEST compression efficiency. The size of the obfuscated file (sha1.js) of 2,945 bytes is WRONG. The correct file size after obfuscating it with Jasob is 2.243 bytes which clearly makes Jasob a winner in this comparison. To achieve best compression ratio, names from both name lists should be first marked for obfuscation and then new names should be generated from the main menu (Obfuscate -> Generate New Names) or obfuscation process should be started (Obfuscate -> Obfuscate). Please update comparison table with correct obfuscated file size (2.243 bytes) for Jasob obfuscator. ]]> http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-188 Jimmy Mon, 25 Feb 2008 03:02:05 +0000 http://justtalkaboutweb.com/2008/02/20/obfuscating_javascript_css/#comment-188 Thanks for your insightful explanation. I'll try to integrate "include" into one of my projects for in-depth knowledge of how it works. Probably, I'll have a review on it soon. Thanks for your insightful explanation. I’ll try to integrate “include” into one of my projects for in-depth knowledge of how it works. Probably, I’ll have a review on it soon. ]]>