Archive for October, 2009

As we near completion of the 2.9 milestone, it’s that time of dev cycle again, when we ask all you community developers who’ve been putting off contributing to core to dust off your dev environments and help us get closer to being release-ready. How? Bug hunts! Yes, that time-honored tradition (in the time of WordPress, anyway) of everyone pitching in to test patches and report the results, working on solutions to major bugs, and helping to clear out Trac has come around again, and we’re scheduling not one, but two bug hunts over the next couple of weeks to ensure that everyone has enough time to prepare and participate.

#1 – The first bug hunt of 2.9 will be Thursday through Saturday, November 5-7, 2009. This should give people a few days to plan for it, upgrade their dev environments if they haven’t been following trunk, and figure out how to allot their time. We’re stretching over both weekdays and weekend to try and accommodate everyone’s schedule.

#2 - The second bug hunt will be a week later, Saturday through Monday, November 14-16, 2009. This should make it possible for anyone who needs more than a week to set some time aside to participate. This bug hunt will coincide with WordCamp NYC, where a special Hacker Room will be set aside for people to go and work on 2.9 bug tickets alongside regular core contributors including Mark Jaquith and Matt Martz (sivel from IRC).

The Goals

Test, test, test existing patches! You can see all tickets with patches that need testing by checking this report. When you’ve tested a patch, report your results in the ticket comments, so core committers can see how the patch is faring.

Fix known bugs! You can see the bugs that need patches by checking this report. Look for the ones that seem that they’ll affect the most people or have the biggest impact by being fixed. Edge case bugs should be lower priority.

Report new bugs! As you’re testing out the development version, if you come across a bug, search trac to see if someone has reported it yet. If so, add a comment with your experience to the ticket so we’ll know it’s affecting more than one person. If no ticket exists yet, create one.

Core committers will be around (in the #wordpress-dev channel at irc.freenode.com) both weekends to review patches that have been thoroughly tested, answer questions as needed, and give feedback on patches that need more work before being commit-worthy.

If you’ve never participated in a WordPress bug hunt before, but you’d like to get involved, we’d love to have you join us! To prepare, you’ll want to set up a test environment, start using the current development version/maybe install the beta testing plugin, join us in the #wordpress-dev IRC channel, and read up on automated testing.

Last week on October 15 I attended The ReadWrite Real-Time Web Summit (try saying that three times fast), along with Tom Hughes-Croucher, and several other Yahoos from the Updates, Messenger and Search teams.

Ironically, considering the subject matter of the summit, this was also the day Balloon Boy didn’t take off into the air, and this ‘real time’ news story, breaking in front of our eyes on CNN, Yahoo!, and Twitter was a perfect illustration of both our need for immediate information and our annoyance with current delivery and discovery methods.

Tom’s session was about coming up with the concept of “Comet Tornado Reverse HTTP Proxy,” which provides a way to use Tornado to add real-time stuff to existing architecture. When Tom got back from his session, the Github guys were also talking about an implementation of his idea in NginX.

The session that most interested me was one on Gender and the Real Time Web. As with most tech conferences, there were only a handful of women, and most of them were in that session. We discovered that, of those in attendance, the women were less likely to share location details in real time, but otherwise, their real-time web habits didn’t differ dramatically from the men.

Real-time information is obviously already upon us, but we struggle with how to best digest it. Do we need filters? Will important information bubble up out of the stream? How much time is someone willing to spend customizing their experience so that important information won’t be missed? How does all of this affect how and what we build and design? I’m not sure much got firmly decided at the summit, but the conversations that took place will give us all some great places to start figuring out the answers.

Robyn Tippins
Community Manager, YDN

We’re putting the finishing touches on our new Lists feature and we’re really excited about the folks who have already taken a lot of time creating awesome lists. From the @time list of funny people to your own list of people who make you laugh—it’s easy to see how this feature increases discovery and adds value in lots of ways.

Lists also make it easier to curate tweets into meaningful real-time experiences on your own sites via the Lists API. Media companies are already taking advantage: for example, check out @huffingtonpost’s use of the Lists API in their World Series coverage.

@Bettydraper/rolodex: A collection of fan-created Mad Men characters
@NYTimes/staff: The colorful people behind The Gray Lady
@BBC/radio1-1xtra: Turns out BBC radio hosts are a chatty bunch
@Joesebok/poker: A list of professional poker players by a professional poker player
@jayrosen/mindcasters: A list of the some of the best new media thinkers by an NYU professor
@Stocktwits/suggested: A list of traders for stock market fanatics

We’ve been taking our time rolling out the lists feature to make sure things go smoothly and developers have a chance to begin experimenting with our Lists API. For example, TLISTS will provide tools to efficiently build, measure and distribute Lists, while Listorious hosts a directory of ‘awesome lists’ on Twitter.

Anyone can curate and publish lists, so if you have an idea for one, just click “New list” in the sidebar of your Twitter account and you’re on your way. Add accounts to a list using the “Lists” drop drown on a profile page. We believe Lists will be a new discovery mechanism for great tweets and accounts.

Caridy Patino, developer of the YUI Event Bubbling Extension, gave a great talk today on building modules in YUI 3 at YUIConf 2009.

Some elements that stood out:

• Modules are contained pieces of functionality that can be used in YUI 3 via the use() method
• Modules are not plugins, though the plugin functionality is defined as a module. Plugins add functionality to a pre-existing object and are added via the plug() method.
• There are three ways to register modules with YUI:
  • Specify the modules to load on the seed object and let YUI pull it in, e.g.,

    YUI().use('module'...

  • Include the script defining the module on the page, and then add it to YUI, e.g.,

    <script src="module.js"></script> ... YUI().use('module'...

    This is a quick way to start using a custom module, but it is the least performant because it adds a blocking http request to the page

  • Specify the location and dependencies for a module in a configuration object passed into the YUI seed object. This is way offers the highest performance. To illustrate:

    YUI({
        modules:{
            'module': {
                fullpath: "http://example.com/module.js",
                requires: ['node', 'event']
            }
        }
    }).use('module'...
    

In general, module architecture looks something like this:

• 
  

  YUI.add(‘module’, function (Y) {
      //module code
  },
  /* version */,
  /* requirement list */);
  

For a good example of module structure and usage, check out Dav Glass’s YQL module in the YUI Gallery. Here’s the source code. Another good example is in tutorial for Dynamically Adding YUI and External Modules on YDN.

Slide decks and video from the YUIConf presentations should be up soon. Please follow the YUI blog for more info.

Find Caridy online:

• github.com/caridy
• twitter.com/caridy
• caridy.name

—

Erik Eldridge

Yahoo! Developer Network

On October 17th, YDN sponsored yet another grassroots (un)conference, this time BarCamp Blackpool. I went along to talk and mingle and talk some more.

Blackpool is a great place for an event. For those who don’t know, it is a seaside tourist town in the North of England. It is known for its yearly illuminations, tower and amusement park. This explains why directly outside my hotel room window was the UK’s largest roller coaster.

There was a good turn out, but as usual with free events many people who asked for a place didn’t come along. If you are one of these people and you didn’t let the organisers know then shame on you! Seriously, people who don’t get paid put a lot of hard work in to organising these events. We need to fix the problem of non-attendance somehow. </rant>

There was a nice variety of subjects covered on the day, from the non-technical but geeky talk on Geocaching from Alistair MacDonald to the more hardcore techy Amazon EC2 and Mono framework talk from Tim “tagwalk” Hastings. Rich Quick gave some useful advice on HTML emails in a typically entertaining fashion.

Probably my favourite session was an introduction to British Sign Language by Lalita D’Cruze. After some encouragement at the BarCamp, she has created a blog about learning British Sign Language, which I think will be an interesting read as Lally continues her training.

Evening entertainment was provided by the hilarious Paul Sylvester and a game of Werewolf, rounded off a great day.

Many thanks to Gemma Cameron for organising the event, and to everyone who attended and talked.

Ian Pouncey

Web Developer, Yahoo! London