Simon recently wrote a blog post about the utility computing market and some of the new entrants.
When we launched Zimki, in early 2006 (with the service going public in March 2006), we hope that it would the start of a new market and would spawn some competition to develop choice for the consumer. This seems to be happening with a new kid on the block, BungeeLabs.
The idea for Zimki was simple and based around:-
• an online development environment which took care of all the "Yak shaving" which normally occurs with software development
• a "pay as you consume" model for charging for the use of our computing cloud
• the creation of a competitive utility computing market through the open sourcing of Zimki.
Later that year Amazon's EC2 launched, we knew we had competition - so it's good to see another company enter the same market space.
Why good? Well, it validates the market, it creates competition and I have to agree with Sha Agassi's sentiment that utility computing clouds are the most important developments in the software industry for the last ten years.
Agassi however refers to Amazon's EC2 directly whereas my view is the really important step is the establishment of a competitive utility computing market - not just one provider - hence the reason for open sourcing Zimki to try and kick-start this process (see my earlier posts on open sourcing Zimki, large scale disruption, utility based grids etc)
So how about the new kids on the block with their next-generation on-demand environment? What do they provide?
A web-based IDE, on-demand scalable deployment, highly instrumented infrastructure and utility computations that combine computing, storage and network interaction - hey it sounds like a very familiar concept.
Unfortunately they are not going to be into beta phase until later this month apparently - but it is direct competition at the same level of the stack – which is good news.
The IDE they demonstrated at Web 2.0 looks good, although I didn't get a chance to play with it myself and from Alex Barnett's blog they've got the concepts right - utility computing and removing "yak shaving" (note, must buy shares in the little company that makes our Yaks for us). Of course, I'd have to agree since we've been doing and talking about this for over a year, I wouldn't exactly call it revolutionary though. We didn't even think it was revolutionary back when we started and that was some time ago.
However a couple of disappointments for me. First they seem to have created a new language - BungeeLogic. I wish they hadn't, there are enough of them out there, and that’s why we chose JavaScript.
Secondly from what I was told by them, they are only open sourcing the connection component not the entire environment and engine. This is a shame, as the real value is to have many providers in the same space with developers freely able to move between providers.
Still, it's exciting though.
