Advertisement
Advertisement

More Related Content

Advertisement
Advertisement

Mobile JavaScript

  1. Mobile JavaScript javascriptz right in yo hands, mawfocker
  2. token “me” slide Web developer / Master Exploder at Nitobi Software SFU graduate Big fan of the open web Bigger Canucks fan
  3. Mobile: How’s it doing? Browsing ~3% of all browsing takes Today, by Device Category Trend place on a mobile device; November, 2009 to September, 2010 Apparently 6% in Asia. Month Desktop Mobile Consol November, 2009 98.65% 0.04% December, 2009 98.36% 0.05% January, 2010 98.19% 0.05%
  4. Browsers everywhere!
  5. WebKit FTW The major browser platform moving forward Apple started it Android jumped on Even RIM pulled their head out of their ass and made the switch
  6. But... Still have: Internet Explorer Mobile; now with extra non-compliance! Old BlackBerry Browser, it’s so bad it doesn’t have a name
  7. so wtf do we do? (other than drink beer)
  8. codes that help Peter Higgins’ has.js is rad SunSpider is sweet Sputnik is also neat
  9. has.js Peter Higgins, Dojo project lead, put this out there as of JSConf.eu 2010. A JavaScript feature detection library, using JavaScript. http://github.com/phiggins42/has.js
  10. SunSpider A JavaScript benchmark. Part of the WebKit project.
  11. Sputnik Google project, a JavaScript conformance test suite against the ECMA-262 spec Over 5000 tests
  12. The numbers
  13. Android (2.2) has.js: apparently a few css and DOM tests fail! function bind fails (?), and full <input> support is not there. No es5 or svg. SunSpider: 6121ms Sputnik: All tests ran, 226/5246 failed (4.3%)
  14. iPhone (4) has.js: same css, dom and function bind issues as Android. Better <input> support too. es5 is lacking, but full SVG support (woot)? SunSpider: 10252 ms Sputnik: Crashed mobile safari... 3.7% tests fail (137/3746)
  15. Shitty BlackBerry, slide 1 BlackBerry OS 5.0
  16. Shitty BlackBerry, slide 2 has.js results array has no reduce or isarray methods, some regexp tests fail no audio support, no canvas there are lots of css bugs/lack of support: overflow, border-radius, transforms, text- overflow, opacity. super weak date support; no toisostring, now or tojson. no native JSON tools (just grab json2.js from json.org) no xhr upload events, any type of offline storage (OTHER THAN GEARS, YES!!!!) no svg, video or vml
  17. Shitty BlackBerry, slide 3 SunSpider and Sputnik SunSpider: Didn’t finish, each test ran on average 4000ms (compared to Android’s ~100ms). Sputnik: Can’t even get to the test runner!
  18. BlackBerry 6 (WebKit) has.js: better than both Android and iPhone. Full event support (unlike Android/iPhone). Websockets and worker support, that is fucking awesome. Full storage options. SVG is super performant on it SunSpider: 26131 ms Sputnik: Crashed browser at 1%
  19. Windows Phone 7 has.js: All array tests fail, no audio support, getelementbyid ignores case, getelementsbytagname returns comment nodes, so much more fail! css bugs galore; see IE6. no event.stopPropagation or preventDefault BUT... VML!!!! SunSpider: 145050 ms... Sputnik: NOPE
  20. Palm has.js: pretty solid overall. createAttribute apparently fails. canvas, audio, video works. sqlite and localstorage options are there too. SunSpider: 39193 ms Sputnik: nope
  21. Final words Know your enemy Pick your battles
  22. BEER TIME NAO Thanks!

Editor's Notes

  1. other stuff: - one of many contributors/maintainers of PhoneGap, XUI, lawnchair
  2. other interesting tidbits: - financial firm Morgan Stanley predicts mobile browsing will outstrip desktop by 2015
  3. web as a platform; fuck native if we could harness the power of browsers we could get out to so many people
  4. -closest thing we have to a unified platform on the web
  5. - initially, WP7 had IE 6/7 hybrid. latest version has 7/8 hybrid. sources say the browser is separate from the rest of the platform and can be upgraded much easier than the OS. Also MSFT has hinted that it will be moving WP7 to IE9.
  6. - checks for proper behaviour in your basic types, functions (binding et al), common browser bugs.
  7. tons of benchmark tests, run a bunch of times for each test to come up with a 95% confidence interval
  8. hard to read tho. you need to know the ecma spec to make any real sense of it. but a good guide for general ecmascript conformance.
  9. so much fail I need many slides to show this off
  10. re: css. it DOES have position: fixed though!
  11. no surprises here
  12. know your enemy: look at your analytics. if youre building a mobile site, who visits your &amp;#x2018;desktop&amp;#x2019; site? pick your battles: some of these environments really suck. building a web app on one platform might take four or five times more time than building it for another one.
Advertisement