#1341 new
Andrew Gizas

Evaluation of Javascript Frameworks

Reported by Andrew Gizas | April 28th, 2012 @ 07:24 PM

We are a researching team from University of Patras Greece, and we introduce our latest work regarding Comparative evaluation of JavaScript Frameworks (JF).

We presented our Job in the last WWW 2012 Conference in Lyon and we would like to inform you that our results are available for further analysis in the url:
http://150.140.142.212:100/JFMetrics/ipresentation.html

We provide quality, validity and performance tests for the most common JavaScript Frameworks today. Our main intention was not to name the best JF, but to reveal to their supporting community their drawbacks and help them in producing high quality, full-featured JFs for the web developers.

We would extremely appreciate your analysis on our job, further recommendations on how we should proceed our research and whatever advise do you think is appropriate! We provide a list of functions with big complexity and validation issues for prototype in order to be consider from its community

Of course we are ready to provide any kind of help in any open project/problem you have right now!

We would like to provide our conclusions to your new version of JF if it is possible! We would be more than happy to have an affiliation with you.

Kind Regards
Andrew Gizas
Computer Engineer & Informatics
MSc., PHD Cand.
Email:abg@hpclab.ceid.upatras.gr

Comments and changes to this ticket

  • Victor

    Victor May 26th, 2012 @ 12:15 PM

    Source metrics are incorrect because comments are stripped from resulting prototype.js file (speaking your words it is just Compact Version). You should download documented sources from github and count comments there. IMHO files in src/prototype folder have 14326 source lines, 1138 of them are empty, 8674 are comments. It has 60% comment ratio - actually the BEST comment ration from frameworks under your test.

  • Victor

    Victor May 26th, 2012 @ 12:26 PM

    And again the very questionable "Performance tests".

    • Your tests (actually not your but SlickSpeed) are testing ONLY speed of CSS selector engine; speed of DOM traversal methods or string/array manipulations is still untested.
    • Prototype can be built with at least 3 different selector engines, providing different CSS speed tests.
    • Currently these CSS speed tests are almost useless because user often cannot see difference between 50ms and 100ms selector execution time.
  • Victor

    Victor May 26th, 2012 @ 12:45 PM

    Validation tests are VERY questionable - the most of "Critical Errors" is JSLint's "Statement Missing Semicolon", the rest of critical errors is raised on browser-specific code (there should be no compatibility errors because this code is intended to work on exactly this specific browser).

    Also 5 warnings on AJAX ("Potentially Dangerous Technology: AJAX"), and one of them is false warning in JS comment - this is very humorous.

    Also many warnings on single-quoted strings "Multibyte character constants are not portable", and false positives on for in loop.

    Validation settings are requiring manual tuning and inspection to eliminate false positives.

Please Sign in or create a free account to add a new ticket.

With your very own profile, you can contribute to projects, track your activity, watch tickets, receive and update tickets through your email and much more.

New-ticket Create new ticket

Create your profile

Help contribute to this project by taking a few moments to create your personal profile. Create your profile ยป

The Prototype JavaScript library.

Shared Ticket Bins

People watching this ticket

Pages