Web/Tech

Let’s face it: less than a week into the game, the shortcomings of Google Buzz are crippling its use. Even proponents are critical of the platform. As Robert Scoble put it, “They made some horrid mistakes.”

Nonetheless people (even Robert!) want it to succeed. Massively. The web is overflowing with calls to fix Buzz.

Google must react at lightning speed. But what can busy engineers do to draw signal from the jumble of rants and raves scattered across the web?

The Gmail help center is a dead end. The Suggest a Feature page contains nothing Buzz-related. The list of Known Issues has two entries.

As a quick solution, I set up an unofficial How to Fix Google Buzz series on Google Moderator. Anyone can participate by adding ideas and voting them up or down. I seeded it with Scoble’s “12 worst features Buzz copied from Friendfeed.”

It has now been live 24 hours, so let’s see if the experiment worked.

  • How many ideas have been submitted? 209.
  • How many votes have been cast? 7,016.

Not bad considering it took about 5 minutes to set up.

Here are the top 10 requests:

  1. “Provide a way to hide all comments until I want to see comments” (205 votes)
  2. “Filter by content type. (i.e. don’t show me twitter from anyone, or for a specific user)” (174 votes)
  3. “If I read comments in Buzz mark them as read in Google reader and vice-versa” (162 votes)
  4. “Provide a way to group friends into lists. Lack of this makes using Buzz with more than small groups very frustrating” (143 votes)
  5. “Let me see all the likes by a single person. Over on FriendFeed I can see what Mike Arrington has liked, or commented on. I can’t do that in Google Buzz” (128 votes)
  6. “Add comment moderation. It should be possible to block people right from comments (like on FriendFeed)” (111 votes)
  7. “I’d love to see Gmail filters applied to Buzz. Keep most of the keyword based filters that are available on Gmail filters but add new ones like Number of Comments, Number of Likes, etc. Actions should include Mute for hiding noisy threads” (99 votes)
  8. “A collapsed list view like what’s available in Google Reader” (97 votes)
  9. “Duplicate posts filter / grouping. So I don’t have to scroll past / see all those duplicate Reader shares. e.g. Of Mashable stuff” (92 votes)
  10. “Introduce lists (like in Twitter) to Buzz” (92 votes)

Note that #10 duplicates #4. Too bad Google Moderator doesn’t have a way to merge entries (feature request!)

Feel free to add in your votes and ideas.

On Wednesday we posted on the Google Code Blog that:

As we mentioned last April, we are in the process of porting Jaiku over to Google App Engine. After the migration is complete, we will release the new open source Jaiku Engine project on Google Code under the Apache License. While Google will no longer actively develop the Jaiku codebase, the service itself will live on thanks to a dedicated and passionate volunteer team of Googlers.

With the open source Jaiku Engine project, organizations, groups and individuals will be able to roll-their-own microblogging services and deploy them on Google App Engine. The new Jaiku Engine will include support for OAuth, and we’re excited about developers using this proven code as a starting point in creating a freely available and federated, open source microblogging platform.

Understandably, some people took this to mean that Jaiku was going away entirely.  This impression was exacerbated somewhat by a blogger reporting early that Jaiku was being closed alongside some other products.

The reality is a bit more nuanced, but it is significantly more interesting in my opinion.  First, the jaiku.com domain and the Jaiku user accounts (and their friend graph and their messages) continue to live on just as they have today.  The biggest difference is that behind the scenes Jaiku is moving away from its original proprietary hosting model and on to App Engine.

Personally I love this for several reasons — it is a tremendous validation of the power of the App Engine platform, and another great learning opportunity for the engineers here to work with a very real service.

But the second, and perhaps even bigger news, is that all of the code used to power Jaiku on App Engine is going to be released under the Apache license.  Combine these two changes — Jaiku on App Engine, and open source Jaiku — and you can start see the opportunity that emerges here.

Soon, anyone, for free and with little effort, will be able to install and modify the Jaiku code, launch it on App Engine, and run their own microblogging platform.  Combine that decentralization with standards such as OAuth and the forthcoming activity stream standards, and what we’re seeing here is the accelerating trend away from microblogging being a destination to microblogging being a pervasive and ubiquitous part of the fabric of the web itself.

Now that’s cool.

Will Google have a team of 20 working on Jaiku?  No, and we’re not going to sugar coat it, which is exactly why we posted such an honest an open
letter about the future of the product.  Are there many of us who passionately care about Jaiku and about the possibilities of microblogging?  Absolutely.  And we want you to participate.