Node 3

Due to a number of administrative and technical issues, the original Meshverse Journal, now known as Node 0 has been frozen for the time being at New Meshverse Journal content will flow from Node 3. All of the old content at Node 0 is available here at Node 3, but posts containing links will still point back to Node 0 where the pages may at times load slowly or not at all(though usually they will after 15 minutes or so). If it’s taking too long, you can copy the link and change the portion to Node 3 will provide the same info – quickly. Node 3 exists in order to provide much better response times for you and greater reliability for me so I can create on demand using the best of breed tools I already know. It may very well turn out that “the time being” ends up being a very short period but in the meantime, you’ll need to subscribe to this feed location to stay with the new flow. Hopefully this won’t be too inconvenient or confusing for folk and in time Node 0 will reactivate and be synchronized with Node 3.

Why is this happening? Well, in a nutshell, I write or host nearly 20 blogs on my server and the cumulative traffic levels are now causing performance problems and requiring far more administrative overhead than I’m willing to commit to both financially and time-wise. I’m not surprised. I knew this would happen since the shared hosting I’m using is inherently a compromise. I’d also seen signs of this coming for quite some time which was part of why I said at the beginning of the year that I was reassessing things. Overall it’s really a good problem to have – it’s more an issue of timing actually. Some folk may be interested in how this relates to my other meshverse related projects, the Server Commoditization and the Coming Boom or the state of WordPress hosting options so I’ll provide a bit more info that should help connect the Nodes. Also, you’ll have to read to the end to find out about Nodes 1 and 2.

Commoditization, like a force of nature is hard to ignore. It is impacting hardware and software choices for blogging, social networks and the emerging meshverse so having a sense of how and why leading edge capabilities become commodities can be helpful. I’ve written about this in more detail, including references to the hard, bottom line numbers from the likes of giants such as Google and Amazon. In the context of Rhythmeering, this provides an example of why understanding the Terrain is important. Node 0 is running on a copy of WordPress hosted at Dreamhost because they provided the best mix of capability, flexibility and price while avoiding getting bogged down in what would soon become commodity. I’ve been using Dreamhost for over two years and prior to that had used a variety of blogging software and blog-hosting services including Radio Userland, Manila, WordPress, Blogger, and Typepad/MoveableType. Since I needed more than just blogs, I knew I had to have lots of server capacity with a minimum of administrative overhead for domain hosting and simple web sites, discussion boards and chat. I’d been using the Web Crossing platform which I was familiar with from my days leading the development of Community Catalyst, but was seeing the commodity handwriting on the wall with the combination of open low-cost hosting and Ruby on Rails.

Back in 1999, we’d anticipated that the underlying platform would become a commodity so Community Catalyst was intended long term to be independent of what was under the hood. I considered the possibility of reimplementing Community Catalyst using Ruby on Rails, but realized that the time windows were too short and that it didn’t matter in the bigger scheme of GriotVision. Besides, I needed immediate capabilities starting with a rock-solid blogging tool. Web Crossing’s entry price point was comparable to that of a virtual private server but I didn’t really need all of that capacity right away. Plus the administrative overhead of maintaining multiple WordPress installations was too high. After trying out a few low-cost WordPress web-hosting services and reviewing countless others I decided on Dreamhost knowing that with WordPress Multi-User I could evolve whatever mix I needed. In the interim has begun offering premium features and VIP hosting. Eventually, the WordPress developers will probably take some of the $30M investment they just got to add social networking options to WordPress Multi-User. For those who can’t wait, there are the Buddypress and Xiando social networking enhancements which you can add to your own self-hosted WordPress multiuser site, though neither appears ready for prime-time.

I’d been hoping that Buddypress would mature quickly as I’ve got a service in the pipeline that really could use it and it would be a good way to consolidate many of my blogs and reduce some administrative overhead. Although Dreamhost’s 1-click intstallation and upgrade tools are great, they’re never just one click and can take a half-dozen or more clicks not to mention wait times. Adding in the fact that WordPress updates come out frequently and multiplying by 20 blogs results in more overhead than I’m willing to manage manually. Plus the gotchas(plugins that don’t work with the new upgrade, recovering from database errors etc.) can eat up huge amounts of time. When outages started to really disrupt my flow, I had to find ways to improve the situation which meant more time spent looking at stats, upgrading or adding tools for optimization and monitoring etc. so I began re-evaluating my hosting options.

Dreamhost has a “Private Server” offering that falls in between a virtual private server and their shared service(this DreamhostPS review and Facebook developers discussing DreamhostPS). The really nice thing about it is that you can go to a control panel and increase/decrease server capacity on demand but all of the other administrative services of shared hosting remain. There are other utility oriented providers like Slicehost and Joyent Accelerators(check this interesting review) that are driving the trend towards server commoditization which don’t quite match the administrative features of Dreamhost, but offer better developer options. At the other end of the server spectrum there are the white label social networks like Ning, People Aggregator and Vox which offer very strong social networking and administrative features which developers can access via standards-based software such as the Open Social API. These are being used by Node 1 and Node 2.

The meshverse however, encompasses all connected devices – not just servers so we’ll need to factor in the upheaval emerging on the desktop. Obviously there’s much more to this story but if you’ve come this far, you’re probably ready for a break – we’ll get to in the next episode soon. Let me wrap up by noting that as the meshverse awakens over the long term, FUNK will be used to replicate Meshverse Journal Nodes as well as other sites/services – in effect creating a high level, p2p, next generation content distribution network – a necessary ingredient in the secret sauce of the GriotVision platform.


1 Comment »

  1. […] Node 3 […]

RSS feed for comments on this post · TrackBack URI

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: