State of the browser 2011

#lwsbrowser panel speakers have been great

Yesterday I dragged myself across town to State of the browser and I'm glad I did.

Amazingly the Jubilee line was working this weekend and it was a pleasant journey from Baker street to North Greenwich while listening to The Bugle via Google listen.

Upon arriving I introduced myself to a Mozillan Gerv and we had a varied discussion from Bugzilla to the Mongol Rally. Bugzilla definitely looks like the premier BTS right now. It's frankly insane I have to use Redmine at work. :/

Once the room opened I must have spent the next two hours working out how to connect to the Ravensbourne wireless.

The /etc/wpa_supplicant.conf snippet that had me connected for minutes at a time was:

network={
	ssid="eduroam"
	scan_ssid=1
	key_mgmt=WPA-EAP
	eap=PEAP
	identity="RGuest057@rave.ac.uk"
	password="ChufJelQue"
}

I don't know if this a ja.net requirement or not, but I must say I was fucking seething. Why does it have to be so hard to get Internet? :( Unfortunately my Vodafone SIM was only giving me GPRS, so I couldn't tether.

I did manage to get my Android device connected more reliably to the Ravensbourne wireless, but I bizaarely couldn't tether my laptop over USB to it! :(

So I listened to @mahemoff and noted:

Next I was very impressed by Paul Rouget, who came across as a very competent hacker. The Audio API and other demos looked great. Later at the pub he was talking about his slide software which can control what other people see. Must look into that. :-)

There was really quite a good representation of Mozillans, fellow Durbanite cyberdees was around and two on my table, Patrick Finch a UK expat in Sweden and the Romanian @irinasandu. I did try Firefox on my Android Nexus S and that turned out to be a embarrassing waste of time. It's practically unusable and I'm worried for Mozilla because their mobile ports suck so bad.

Thankfully there was lunch and I met Kiwi @toni_b who seemed sane. I am not sure. She told me I'd like Wellington if I like wind and earthquakes. Gulp.

After lunch I attended two great break out sessions. The fantastically eccentric Simon Stewart gave a great talk about "regression testing" with selenium or selenium2 or WebDiver. I asked several questions and he answered them well. What a great guy. :) Importantly for my own ego, the approach I was taking in my own testing at BONDI & WAC didn't seem so bad.

1) Interrogate the DOM
2) Ability to execute arbitary javascript
1&2 = "browser reading"
3) Ability to interact -- can't be done with JS

logical failure

browser specific bugs

saucelabs -- jason huggins
browsermob

record people remotely - no

going back to JS API - use qunit

Next I accidentally found myself listening to the Dos and Don'ts of the mobile Web thinking this would be a waste of time. I was surprised to hear a familiar sounding name "Staikos". Turns out Matthew Staikos, brother of George Staikos of Torch mobile was speaking! Holy cow! RIM have sent one of the the real webkit browser guys! His talk was great (avoid position fixed was a key take away, a problem I've seen with Opera) and I want to give WebWorks a whirl, however there doesn't seem to be an SDK for Linux. :(

The final panel discussion was good, I asked a general question about which features are typically not enabled on mobiles. The best question IMO came from @pornelski, who asked why didn't RIM, Chrome & Firefox use the W3C widget packaging format. Opera of course already use the format, because as Chris Mills mentions, they wrote it. ;) IIRC RIM/Chrome dodged the question whilst Paul Rouget from Firefox was saying the standard needed more, which I was surprised to hear. I expected him to say, Firefox extensions came first or something. ;)

Later I went to a pub for a swift pint of ale to shockingly see Ireland beating England at rugby. Oh well. @pornelski and @ernesto_jimenez were trying to convince me node.js is a good idea. cramforce tried earlier in the day and scared me with his "10x the amount of code" statement. I was not impressed by "coffee script" generation of Javascript and the callback orientated code. A sloccount of Node.js suggests 337k SLOC. Hmm, no thanks guys!! Nonetheless I must give Server sent events a whirl real soon now, hopefully on http://greptweet.com/ to make the feedback sane.

Advertisement

If you like this, you might like the stateless Web kiosk software I develop. Webconverger typically replaces Windows on PCs and is deployed in public and business environments for ease of deployment and privacy. Once installed it auto-updates making it painless to maintain. Try it where you exclusively use the only viable open platform... the Web!