Communicat.us

When I was in college, companies often donated hardware and some software to universities. This was a mixed blessing. The donation was great but it also meant that the college had to use their limited IT staff to manage the systems and the new equipment was always vying for limited raised floor lab space.

Here is where a cloud service could be a huge win for both the corporate sponsor and the university.

IBM announced it will make portions of its software portfolio available in a cloud computing environment to more easily allow professors to incorporate technology into their curricula and students to focus on development vs management of the systems they require. The initiative is called the Academic Skills Cloud and provides the software and system capabilities at no charge and without the need for students or faculty having to install and maintain it themselves.

Here are some additional resources:

• full article -> Academic Cloud to Speed Delivery of Technology Skills to College Students
• more about  IBM Cloud University Initiatives 
• recording of the live webcast

In my morning “digital mail bag” was a pointer to The Register’s article on Oracle, Sun, and the departure of their CTO for cloud. In the article, they describe Lew Tucker as “former vice president of Salesforce.com’s AppExchange applications market place; had been with Sun for less than two years and was handed the role of building out Sun’s hosted cloud.” You can read the whole article here.

A few weeks ago, I sat down with representatives from a number of government agencies and much of the conversation surrounded cloud computing. What was evident were a few things:

• nearly everyone is investigating cloud options
• “cloud” is an over used term and means everything from raw CPU and storage to hosted custom applications to hosted commodity services
• most of the use cases were for cloud versions of traditional solutions – aka the cloud was just a different deployment option
• traditional thinking means traditional requirements and hosted clouds do not always meet those requirements
• exploiting cloud computing to its fullest will occur as new programs are designed and fulfilled but there are starting points today

The common thread was that having flexibility was critical. Some commodity capability could go into a public cloud. Low risk capability could go into a hosted cloud. By far, the private cloud has the lower threshold of adoption (independent of cost or complexity) as it is easier to trust.

While “cloud” may be an over used term, it is clear there are both current use cases and new thinking for solution design to take better advantage of flexible options. The silver bullet of a “cloud” is it’s ability to smooth out load. The larger the cloud the more balancing that can take place. This starts the balancing act. A private cloud will be smaller than a public cloud. If the private cloud is only servicing one type of solution, then it is likely to surge and idle all at once. However, a service cloud is quicker to leverage than an infrastructure cloud.

• Software as a Service clouds benefit from large diverse user populations
• Platform as a Service clouds benefit from lots of diverse applications
• Infrastructure as a service clouds benefit from lots of diverse compute requirements

So, what do you need your cloud to do ?

Before I get into some of my comments on the 2010 edition of Lotusphere, let me first thank all of the Federal customers who took valuable time to attend and engage with us. I especially am thankful for everyone who made it to the IBM Federal User Group meeting on Wednesday. The Q&A was great and it was great to hear your feedback and questions.

To those whom I did not have the chance to connect, I and many from the team are available both here thru our blog as well as in person so don’t hesitate to “connect with us”!

So, how was Lotusphere 2010?

In a word, “extreme”! Even knowing much of what was announced at the conference, I still had a full time job getting the ins and outs from jump starts to deep dives.

What were the big take-aways?

I’d say that there were multiple big messages. On the environments side it was Cloud, first class mobile device support, and a biggy (pun warning) was Lotus Connections on Z-Linux is coming!

Back to the “cloud”, I’d say the message was blending on premise capabilities with capabilities in the cloud so you can balance the best of both. For many Federal customers, “private clouds” provide the higher security and governance needs.

Before I forget, mobile was everywhere I both is sessions and the hands of attendees. I saw more Blackberries, iPhones, and Androids than I did laptops. IBM gets it – “mobile is a primary interface now.” On a personal note, I wouldn’t say the iPhone out numbered Blackberries but they were definitely prevalent. I’ll following this closely.
Unified Communications was integrated into many sessions. It was all about controlling presence everywhere and to best suit the user, seamlessly integrating the office, VoIP, soft phone, cellphone and flowing between them without interruption – both at the user discretion and through user configurable routing rules. I’ve been using these capabilities for about 6 months and they make a big difference in both my flexibility and transparency.

Any other themes?

Another big message was modular components, capabilities and services. While a lot of sessions were in the scope of products, the content was about “capabilities” and how they are surfaced as portlets, widgets, an modular components. Further, those same user facing capabilities can be accessed with web 2.0 interfaces – REST, ATOM, XML, etc.

A great example of this emphasis on capabilities and modularity was a challenge a customer gave me – “we use Lotus Connections Activities for lots of small projects, can I get a view across all of them? Preferably as a timeline?”

We hit the “Ask the Developer Lab” and in about 15 minutes we had a Lotus Connections engineer, a Notes client developer and a Mac user bouncing ideas. They had two choices. The Activities to-dos can be accessed as an iCal feed and the Notes 8.5.1 client can display multiple calendars together. Alternately, the Activities assignments are available as a feed and IBM Mashup Center 2.0 was a new Timeline widget.

BTW, the three engineers were not together. The Lotus Connections engineer had Sametime Broadcasts to fire off the question.

Any surprises?

“Project Vulcan” was announced as was “LotusLive Lab”. When I get back to my laptop, I’ll write more about these.

What!?

Oh, I should explain. I came to Lotusphere and only carried my Blackberry (where I am writing this post) and occasionally had an iPod Touch. My goal was to test just how far these mobile devices could go.

I had access to email, calendar, Lotus Connections, my shared files, blogs, micro blogging (Twitter and my IBM Connections status), an a nice conference app built by business partners. I also could get to LotusLive where people were sharing emo materials, updated slides and more so could just Dogear (bookmarking) for later.

There is no way I could cover everything at Lotusphere, especially in one post but I hope this gives you a taste and feel free to comment on any points you’d like more info on!

There are lots of talk about netbooks and the cloud and web apps. If you believe all of it, you’d think all you need is a browser and an internet connection. You’d be mostly right. You would need a browser – a very modern one and soon an HTML5 based browser. You’d also need an internet connection and a fast one at that.

Web applications have the advantage of delivering both the user experience and the data from a server every time. If a bug fix needs to go out to everyone or an new feature, it happens the next time a user visits the application web address. No fuss. No foul. But, if the application has a rich user experience with drop downs, and graphics, and drag-n-drop, and worse, if the application was developed in a pristine modern lab, then that application makes assumptions about the quality of the network – it’s speed, it’s latency, and it’s resilience. While this describes the original lab environment and it may even describe what you will find in close proximity to the mega-cities, this does not describe most of real world.

Once again, in the third quarter of 2009, 19% of the Internet connections around the world were at speeds greater than 5 Mbps. This level was fairly consistent with the prior quarter, as well as compared to the same period a year ago. The United States, however, saw high broadband adoption rates decline slightly in the third quarter, losing 0.7%, while shedding a more concerning 8.8% year-over-year. The global level of high broadband penetration, shown in Figure 10, remained consistent quarter-over-quarter, remaining at 0.01 – that is, one high broadband IP per 100 people.

source: Akamai

When just looking at the United States, 91% of the people do *not* have the "high speed" internet.

What can be done ? There are two options:

1. deploy high speed internet everywhere
2. build applications that use the available internet connectivity better

The first will happen slowly and you as an application producer or consumer don’t have much control of it. What you do control is the applications. As a consumer – whether it be an individual or an organization can choose which web applications you use and standardize. You have two options "better applications".

Data centric applications can do a better job of getting the application "out of the way". A good example is this blog. It recently added a new format targeted at mobile devices which reduces load and display time by about 60%. This same interface could be used for any browser. The result is that all of the information of the blog is conveyed with some reduction in functionality but it may be those features were not needed in the first place. Personally, I think we can learn a lot by paying attention to the mobile device environment. It forces applications to leverage the network for what it provides – connectivity – but not assume it is ubiquitous, fast, or consistent.

User Interface (UI) intensive applications require a different solution. If the experience is important, then perhaps the UI should be persistent on the desktop / device and the internet is only used to transfer data as needed. These have an additional advantage in that different devices / user / conditions drive the application requirements. Many web applications have taken this direction. How you implement the desktop application is highly variable. It could be native code – rich experience with less training given it follows standards. It could be Java – common experience across multiple platforms; And soon, it could be persistent HTML5 with it’s localStorage. This last option has the advantage of being a web application so it can be updated quickly and has the advantage of local storage so it does not need to pass the entire user experience back and forth with the server each time.