Weiqi Gao's Observations - web tag

Kevin Nilson: Pushing Data To The Browser With Comet

Fri, 10 Jul 2009 13:40:08 GMT

Last night's presenter at the St. Louis Java Users Group is Kevin Nilson who gave a fascinating talk about Pushing Data to the Browser with Comet.

I came into the presentation not knowing what Comet is, exception what I heard on the internet. Kevin did a great job explaining what Comet is. The most valuable part of the presentation is the demonstrations Kevin did, which are fun and informative.

I'll let you read an article by Kevin himself to get the gist of what the subject matter is. I'll just show the words I wrote down during the talk:

Jetty
Pub/Sub
Bayeux
JASONp
script tag
XSS
Cometd
Long polling
Channel
Service channel
Meta channel
Server crash
Client auto-reconnect
JSON on the wire/HTTP

The background conversation at the JUG is also informative:

Is Scala for real? Kevin told us its the hottest thing at Silicon Valley. Mark and I still have some lingering doubts. Mark is focusing on something called persistent data structures. I'm more of a Luddite, fearing the years of learning that I have to go through to be proficient.
At the end of the session, when Brian went up the stage to run the Birthday Selector application to give away the goodies, Kevin Nilson mentioned the Wheel of Fortune application that Jim Weaver wrote for his JUG. Small world.
Ken Totton (of OCI, where I work) is still looking for top notch Java architects and developers. Send me an email at "weiqigao at speakeasy dot com" if you are interested. A couple of other places are also recruiting.
Brian asked a question about MigLayout. He's reading the OCI May Java News Brief on MigLayout. I gave him an meta-answer: "Just blog about how it doesn't work, and the MigLayout author will comment on your blog and show you the right answer."

Coming back to Cometd, I do have one reservation: It's another messaging protocol. This is the rare occasion when my Java duties intersect with my messaging protocol-watching duties. It reminds me something Paul says around the office: "You can always define a Pub/Sub protocol on top of a Request/Response protocol. And you can always define a Request/Response protocol on top of a Pub/Sub protocol."

That why this post will also be available under the title "Another JUG Meeting, Another Messaging Protocol". :)

Oh, Really? (Joining the Crowd in Deriding Wolfram Day)

Sat, 16 May 2009 17:25:21 GMT

Fresh evidence that the web has made us stupid, even the geniuses at Wolfram Research:

How could they not know this?

Google Chrome: The Three Month Report

Thu, 11 Dec 2008 14:10:11 GMT

I almost forgot about my promise to write about my experience with Google Chrome.

That was 98 days ago. It's a long time to be evaluating a product. I'd like to say that I did not forget to write the evaluation, but that's exactly what happened. After the first batch of complaints, I simply kept on using Google Chrome.

Chrome "worked" for most of the web sites that I visit on a daily basis. And it did a good job of fading into the background. Most of the time, I don't think thoughts like "This is Chrome." By the time I realize I still haven't written a review, three months have passed.

So, what kept me from falling back to Firefox 3.0.4? Start up time. Maybe it's because I have too many plugins installed in Firefox, the few times when I tried to use Firefox its start up time is three times that of Chrome. Whereas Chrome starts in one second, Firefox starts in three.

I know what you are thinking: "It's only three seconds." You can't believe how long that three seconds feel, especially when there is a one second alternative that does the same thing. "I can't believe how I tolerated this before," was my thought every time I start Firefox.

Now, my complaints. I have only two. First, if I click on a link and Chrome starts a download, that download "belongs" to that tab. And the download will be canceled if that tab or the whole browser is closed. And this is a real-life issue. I have been bitten by it several times.

Second, there are still some financial institutions that tell me only IE or FF are supported. This is not Google Chrome's fault, but it is a problem for the Chrome user nevertheless. My work around is to fall back to IE.

So, I'm quite a happy Google Chrome user.