Entries from May 2008 ↓

My dear friend Yaniv Golan, CTO of Yedda, had given a brilliant presentation regarding Incentives In Online Social Communities a few weeks ago at The Marker COM.vention and since it’s unfortunately in Hebrew, I wanted to translate it, include some of my own additions, and share it with you.

Online community participation

Yaniv Golan

Let’s start with the obvious question….Why?

Why do users comment? Why do they write blogs? Why do they upload pics to Flickr? Why do they send links to friends?

What are the motives behind user participation in social communities? Understanding why users participate can lead us to understand further how to engage users and increase their participation in online communities. Let’s first learn a bit more about our users.

Membership life cycle for online communities

Amy Jo Kim was the first to propose the idea of a member’s life cycle in an online community (2000). The cycle suggests five phases of a user’s lifecycle within a community:

1. Peripheral (i.e. Lurker) – An outsider, unstructured participation
2. Inbound (i.e. Novice) – New user, invested in the community, on his way to full participation
3. Insider (i.e. Regular) – Committed participator, member of the community
4. Boundary (i.e. Leader) – A member brokering interactions and encouraging/sustaining participation
5. Outbound (i.e. Elder)  - On his way to leaving the community, perhaps to another community due to a particular change in the community or personal choice

Power Law of Participation

According to Ross Mayfield:

“The vast majority of users will not have a high level of engagement with a given group, and most tend to be free riders upon community value.  But patterns have emerged where low threshold participation amounts to collective intelligence and high engagement provides a different form of collaborative intelligence……

Digg is the archetype for low threshold participation.  Simply Favorite something you find of interest, a one click action.  You don’t even have to log in to contribute value, you have Permission to Participate. Del.icio.us taps both personal and social incentives for participation through the low threshold activity of tagging.  Remembering the URL is the hardest part, and you have to establish an identity in the system.  Commenting requires such identity for sake of spam these days and is an under-developed area.  Subscribing requires a commitment of sustained attention which greatly surpasses reading alone.  Sharing is the principal activity in these communities, but much of it occurs out of band (email still lives).  We Network not only to connect, but leverage the social network as a filter to fend off information overload.  Some of us Write, as in blog, and some of us even have conversations.  But these are all activities that can remain peripheral to community.  To Refactor, Collaborate, Moderate and Lead requires a different level of engagement — which makes up the core of a community…..Participation in communities plots along a power law with a solid core/periphery model — provided social software supports both low threshold participation and high engagement.”

All users activities in online communities whether low threshold or high engagement activities co-exist within a community to create a form of collective intelligence. Therefore it is key for virtual communities to allow both low threshold and high engagement participation so that users in all 5 phases of their lifecycles will be made to feel comfortable within the community.  

Participation Inequality

Social Platforms – the 1% rule

1. 90% of users are lurkers
2. 9% of users contribute sometimes
3. 1% of users actively participate and are responsible for almost all the action

On Wikipedia for example, participation inequality is even higher. More than 99% of Wikipedia’s users are lurkers. Only 0.2% are active participants. Wikipedia’s most active 1,000 people — 0.003% of its users — contribute about two-thirds of the site’s edits. 

We see here that small groups of people often turn out to be the main value creators of social communities. Over time, their actions fuel widespread interaction that engages the lurkers and attracts new users. If continually nurtured, the community can become a self-sustaining generator of content and value.

So let’s go back now to our initial question:

Why do users participate in virtual communities?

According to Peter Kollock in The Economies of Online Cooperation: Gifts and Public Goods in Cyberspace, there are three major reasons for why users contribute in online communities:

1. Anticipated Reciprocity – A user is motivated to contribute to the community in the expectation that he will receive useful help and information in return. Indeed we have seen such active users receiving more help than lurkers.
2. Increased recognition – individuals want recognition for their contributions. the desire for prestige is one of the key motivations for individuals’ contributions in an online community. Contributions will likely increase if they are visible to the whole community and are credited to the contributor. … the powerful effects of seemingly trivial markers of recognition (e.g. stars, ranking) are overwhelming.
3. Sense of efficacy – Individuals may contribute because the act results in a sense that they have had some effect on the community. Wikipedia is a good example of this.

Yet there are also other elements which can motivate users to become active in online communities:

1. Connections within the community – the more friends a user has within a given community, the more important it becomes for him to participate in. Therefore it’s important for online communities to allow users to form friendships easily and encourage a high level of interaction between users.
2. Emotional Safety – a sense of belonging and identifying with the community. Once users become regulars in a community, just like in any offline community, they stop feeling fearful and begin to feel a sense of safety in and identification with the community. The key here is to get these individuals to become regular users in your community and create a cozy and ”feel good” environment for them.
3. Common emotional connection – niche communities that are built around a particular emotional connection/cause between members tend to become more cohesive and experience lower percentages of participation inequality.
4. Altruism - Yossi Vardi coined the term “Dopamine Over IP” – each user transfers dopamine to another user….by contributing content, a user knows that he will cause pleasure to those who view it and those users that forward this content onwards, know the same.

For more reasons why people become active participants in social online communities and the key to Web 2.0’s success, please see my posts:

• What’s Behind the Success of Web 2.0? A Psychological Interpretation
• Web 2.0 and the new tribalism

So now you ask…

What are the ways that online communities can overcome participation inequality and increase users’ participation?

1. Make it easy for users to contribute, make them feel confident with their contributions, and share their contributions with other members in the community -> Feeling of influence
2. Make participation a side effect. Let users participate with zero effort by making their contributions a side effect of something else they’re doing. For example, Amazon’s “people who bought this book, bought these other books” recommendations are a side effect of people buying books. You don’t have to do anything special to have your book preferences entered into the system.
3. Reward users’ contributions and allow for markers of their contributions. Promote and feature top contributors -> Sense of recognition, sense of community, fulfill anticipated reciprocation

Allow users to rank each other within the community and comment on contributions -> sense of community, feeling of influence

4. Platform should be flexible enough to transform with the changing needs of its members -> feeling of influence
5. According to virtual community pioneer Jonathan Bishop, online community managers need to also change the beliefs of lurkers on their site in order to increase participation. Lurkers, believe that they do not need to post messages or that they are being helpful by not posting. Such beliefs prevent them from carrying out their desires to be social and participate in the community. Therefore it is up to the community managers to change this attitude by use of persuasive text or by other means.

A few more useful tips for community managers

1. Simplicity is key – participating in the community should be simple for the user. The simpler it is, the higher the participation rate will be.
2. Allow some actions to be performed by non-registered users.
3. Give people something good to talk about – as always, content is king. If your content is interesting and appealing enough, people will be eager to contribute.
4. Display the activity on your site. No one likes to go into an empty restaurant. Already on the homepage show users all the great stuff that’s happening within the community.
5. Offline events are a great way to make a community even more cohesive and virtually active.

Paul Duncan announced today the release of PersistJS, a client-side JavaScript persistent storage library.

Currently the only reliable cross-platform and cross-browser mechanism for storing data on the client side are cookies. Unfortunately, using cookies to store persistent data has several problems:

* Size: Cookies are limited to about 4 kilobytes in size.
* Bandwidth: Cookies are sent along with every HTTP transaction.
* Complexity: Cookies are difficult to manipulate correctly.

Modern web browsers have addressed these issues by adding non-Cookie mechanisms for saving client-side persistent data. Each of these solutions are simpler to use than cookies, can store far more data, and are not transmitted along with HTTP requests. Unfortunately, each browser has addressed the problem in a different and incompatible way.

Trying to address the need for client-side storage sans browser-specific techniques or browser plugins, Paul has created an abstraction layer that allows developers to use most of the most common client-side storage mechanisms via a common interface. It currently supports persistent client-side storage through the following backends:

• flash: Flash 8 persistent storage.
• gears: Google Gears-based persistent storage.
• localstorage: HTML5 draft storage.
• whatwg_db: HTML5 draft database storage.
• globalstorage: HTML5 draft storage (old spec).
• ie: Internet Explorer userdata behaviors.
• cookie: Cookie-based persistent storage.

Other notables features include:

• Small (9.3k minified, 3k gzipped)
• Standalone: Does not need any additional browser plugins or
  JavaScript libraries to work on the vast majority of current
  browsers.
• Consistent: Provides a consistent, opaque API, regardless of
  the browser.
• Extensible: Custom backends can be added easily.
• Backwards Compatible: Can fall back to flash or cookies if no
  client-side storage solution for the given browser is available.
• Forwards Compatible: Supports the upcoming versions of Internet
  Explorer, Firefox, and Safari (Opera too, if you have Flash).
• Unobtrusive: Capability testing rather than browser detection, so
  newer standards-compliant browsers will automatically be supported.

Erick Schonfeld / TechCrunch:
Google Backers Back Aster Data Systems.  It's All About The Clusters.  —  Clusters are the way to go.  Google and Yahoo run their Websites on distributed databases spread across vast clusters of servers.  Now Aster Data Systems, a startup that is coming out of stealth mode today …

There’s some confusion as to what the RAW photo format actually is, and, like any good photographic fact, it can incite forum flame wars as quickly as the mention of the words Leica and Bokeh in the same sentence. Although it comes in various flavors — seemingly one for every different camera model — RAW is essentially the raw data from the camera’s sensor, hence the name.

If your camera has a RAW setting, you should be using it, no excuses. Here’s why.

Dynamic Range

Dynamic range is the difference between the lightest and darkest parts of a scene. Unless the lighting is very flat (lacking in contrast), your camera’s sensor will only capture a subset of that range. A RAW file, which contains all the data from the sensor, will give a dynamic range of around eight stops. A JPEG will will give you a couple of stops less, which usually translates to blown out, or over-exposed highlights and loss of details in the shadows. So, while you still need to be careful with exposure, RAW will record the maximum information available to you.

Also, the histograms displayed on most cameras are based on a JPEG preview (even when you are shooting RAW). So a histogram that shows your picture as overexposed (the graph is pushed up against the right-hand side) might still have some detail left.

No In-Camera Processing

One drawback of RAW is that you can get flat-looking previews on both the camera’s LCD screen and when you load up your images into editing software. This is because, unlike JPEGs, the camera is doing no processing to the file; no sharpening, no fancy tricks to boost the colors, no nothing. All of the important decisions are left to you to apply later, on a big screen with a much more powerful computer than the one in the camera.

To get the maximum data from a scene, common advice says that you should expose for the highlights, just like with slide film back in the day. Once the highlights have blown, there’s no getting them back. With the shadows, however, you can often pull details out of the murk. The flat looking preview will show you just what you captured. It might not be pretty now, but you are shooting to record the maximum information.

Adjust later

Next to capturing the maximum info from the sensor, the best thing about RAW is the post processing that can be done. Because the camera doesn’t bake any of its settings into the image, you have a clean slate on which to work. Using non-destructive editing software like Apple’s Aperture or Adobe’s Lightroom, you can make endless adjustments to the exposure, white balance, contrast and just about anything else you could do in a real darkroom and change your mind later.

These programs never touch the original RAW file; they keep a small text file (just a few kilobytes in size) which contains the adjustments you have made. Each time you look at the photo, these settings are re-applied in real time (although usually there is a preview to keep things quick). Even cropping, dust spotting and sharpening can be undone, years later, with the original file unaffected.

The (Few) Disadvantages

As you’d expect, there are some disadvantages. RAW capture is slower. Hold down the shutter release of a DSLR and it will happily shoot jpegs until the memory card is full, barely slowing down. Try that with RAW and even pricey cameras will slow to a crawl. Also, RAW files are bigger. That, though, is a poor excuse. Hard drives are cheap, and getting bigger all the time. Of course, some cameras don’t let you shoot RAW files. The manufacturers want you to buy a more expensive camera. If you own a Canon, though, you might be in luck. The CHDK (Canon Hacker’s Development Kit) will let you install hacked firmware onto some models, adding RAW capture amongst other goodies.

So if your camera has a RAW setting, go switch it on now. The advantages far outweigh the small drawbacks, and it is the only way to be sure you are getting all you can from your camera. A RAW file isn’t called a digital negative for nothing.

Eugene Lazutkin has written a very thorough post on dojox.lang.aspect a module that takes AOP seriously in JavaScript.

As someone who has been to a couple AOSDs (the AOP software conference) and was excited to see AOP on the scene, it is good to see someone who gets it working on the JavaScript side.

Of course, we all get the interception piece. It is very easy to just wrap a method to do something new in a dynamic language such as JavaScript. But the power (and complexity!) of AOP lies in the world of joinpoints, pointcuts and worm holes :)

For these, a dynamic language isn’t as helpful. Eugene delves into the world and takes us to the module from first principles. He then builds a timer aspect which can be used to profile an application. Along with a memoizer and other useful cross cutting concerns, you end up with good ole Fibonaci:

Memoization got results too:

On my computer with Firefox 3 the calculation of 1-order (regular) Fibonacci number of 15 (987) took ~48ms without memoization and 0-1ms after memoization. The calculation of 0-order Fibonacci number of 15 (32768) took ~1155ms without memoization and the same 0-1ms after. As you can see this technique can work wonders without much investment of time.

Just like with Java, we won’t see every developer worrying about pointcuts day to day, but instead, simple usage of existing advice will become very useful indeed.

Apple screen savers are pretty much just Quartz Compositions (graphics created via Quartz Composer, a node based visual programming language), and, while there are some nice ones that come pre-installed, they get old rather quickly. You could opt to dedicate those idle CPU cycles to an altruistic effort (or just put your system to sleep and save energy). However, if you are looking to add some glitz and glamor to your workstation you need go no further than your own system.

Apple has squirreled away some gems right on on your own hard drive and the first place you should look is /System/Library/Compositions (in the Finder, hit Command-Shift-G and paste that path in & hit return). With that folder open, go to System Preferences and select Desktop & Screen Saver and select the Screen Saver tab. Now, you can test each one by dragging a .qtz file right onto the test panel in the Preferences window and either wait a bit or click Test. Notable ones include Defocus.qtz, Stix.qtz and Travelator.qtz. Some may require you to select a base image (click Options… to see if any of the Composers have options you can tweak).

Things get really interesting if you head on over (in the Finder) to /Developer/Examples/Quartz Composer/Compositions (which only exists if you install the Developer Tools from the Leopard Install DVD). You have to dig into folders, but you’ll find Image TV.qtz under Conceptual which mimics various Apple ads and the Apple TV intro quite nicely. There’s even a token Screen Savers directory where you’ll find a Quartz Composer that uses motion-detection called Security.qtz (it may not be the best choice for a screen saver if there isn’t a great deal of motion around your system).

Screen Savers DIY

If you do install the Developer Tools, definitely head on over to /Developer/Applications and startup Quartz Composer. Apple made it pretty easy to create your own visual delights. It’s as simple as dragging nodes around, taking input and specifying how data should be processed.

You can open up any of the examples to use as a starting point.

If you do give it a try, drop a note in the comments and share your creation with the world!

[via Mac Tips & Tricks]

Just when you thought that datepickers had been played out, along comes Filament Group and puts a whole new spin on it. Working from Mark Grabanski’s jQuery UI DatePicker control, the team substantially enhanced the UI with a host of new features including:

• shortcut links to preset date ranges, for example, “Past 30 days” or “Current YTD,”
• links to “All dates before…” and “All dates after…” to simplify selecting a range of values where the data set is very large or the high or low end value is an unknown, and
• only showing the number of calendars needed for choosing a particular range (i.e., you only need one calendar to choose “All dates before…”, but you’ll need two to select a custom range).
• Use of progressive enhancement for graceful degradation /li>

Check out the demo here.

Blu has created one of the most incredible stop-motion animations I’ve seen. Painted on walls in Buenos Aires and Baden over this past winter, the piece features a black-and-white creature that morphs into various blobby forms.

Starting off as a multi-armed monster, it constantly shifts with the most consistent tropes involving head changes—from spiky to cubed to round, etc. and sometimes devouring/birthing itself. There's also plenty of other mind-bending evolutions involving things like crawling teeth and creepy bugs. It gets even better when the creature(s) interact with the surrounding details, crawling into corners and snatching pieces of paper off the wall with a frog-like tongue.

via Dreamers, thanks Schatz!