So today it was Microsoft Research’s turn to crow, and RickRashid kicked off with a much more sober and academic presentation than the previous two days. He spent alot of time laying out why basic research is so important to Microsoft, and also showing how successful they have been. For example they now have over 800 PHD level staff, and at a typical academic tech conference, between 10 and 30 percent of the papers are authored by Microsoft researchers. Wow! It does clearly demonstrate the shift in recent years between pure academic sponsored reseach and corporate research – I’m not sure that entirely a good thing from an independance point of view, but you can’t argue it has accelerated the industry.

So the things that stuck in my mind from the keynote were:

SecondLight:

A system for providing a second layer of user interface above the primary surface of a multi-touch user interface.  So for example, by holding a translucent surface, such as a piece of tissue paper, above a surface displaying, for example, a map, a second set of information such as a satelite view is projected onto the translucent surface. Its pretty amazing to see. http://www.pcmag.com/article2/0,2817,2332945,00.asp

Mesh Sensors

A network of low power, self-networking sensors that report back data to a Cloud app. What was particularly neat about this was the ability for many different organisations to share and work on the sensor data. The demo showed sensor data over time from the conference venue, and also in the environment, with data displayed on a virtual earth map

Boku

A cool new way to teach programing to kids. This was a stunning display of how kids could learn basic programming principles in a fun way that allows them to build really complex and exciting games. The point MS were making is that programming should be a life skill, not just for geeks, and that its just the complexity of current programming environments that stop people learning it. This is there attempt to get around that

Virtual Telescope Release 2

Some stunning visuals in this presentation which now has twice the data and much better simulation tech to go from a planets surface to the farthest view of the universe. Already a huge hit with astronomers, this will make it even more popular, and also illustrates some great UI and graphics techniques.

Don’t forget all the PDC keynotes and other sessions are available the  next day from the PDC web-site

Todays keynote focussed on Microsoft vision for combining PC, Web and Phone using Windows 7,  WIndows Live and WIndows Live Services.

The morning was characterised by some very impressive demos, and alot of crowd appreciation from the assembled developers.

Here are some key features of each:

WIndows 7

  • Lots of UI usability enhancements including an improved start menu, taskbar and windows management.
  • New hardwares innovation – e.g. built in support for multi touch
  • The same kernel as Vista and Server 2008, so no breaking changes
  • Seamless software and services integration between windows 7 and windows live
  • Ribbon UI now in all windows apps
  • Jump lists – task bar or start menu right click for e.g. MRU
  • Libraries – unify data across locations
  • DirectX now not just for games – can use for 2d, animation and high performance text
  • Decrease in resources used
  • Increase in performance – boot speed, 256 processor support,  Task bar and start responsiveness
  • Built in bitlocker encryption on memory sticks, includes policy tools for managing
  • Natively mount and manage virtual hard drives – create a VHD natively from within windows; attach a VHD.
  • Boot from Windows 7 VHD, mount, boot and reset to help development
  • Can now easilty set iDPI and multi monitors
  • Improved magnifier
  • Single key select of monitors, dual monitor demo
  • Remote desktop connections can also be multi screen
  • Customise shutdown button
  • Better control of taskbar and icon positioning
  • Control over system messages e.g. firewall, malware
  • UAC – can now control how much of UAC you see on your machine
  • Seamless transition between work domain and home network, maintain security of domain join

Windows 7 Releases:

There will be 4dev builds m1,m2,m3m and beta.

The Pre-beta issued today is 1st stable build – M3.

It’s not feature complete until beta which will be early in 2009

Windows Live

There were some very impressive demos illustrating collaboration using standard office clients, live office, and Live Mesh to automatically sync documents and data whether users were on the web, mobile or pc. This combined with illustrations of using Live services to extend regular desktop aplications into the web illustrated very well the reality of Microsoft Software plus Services vision – its a very convincing model that Microsoft have executed very elegantly.

Windows Live Essentiais – for Rich applications

WIndows Live Services – mail, messenger, photos, blogging – to expand the PC experience to the web

Together, complete communications and web experience

Live Services

Key principles:

  • Access a users social graph
  • Easy integration and interoprability
  • Leverage familiar services

Services:

  • Identity
  • Directory
  • Communication and presence
  • Search and Geospatial

Live Mesh

This  bridges islands of data, apps and devices through synchronisation of Users – Devices – Applications – Data

Developers access live services through the live framework which is built-in to Visual Studio

Office Web Applications

Lightweight version of office apps in the browser – Word, Excel, Powerpoint and OneNote in Office 14

You can download windows live beta @ download.live.com

More tomorrow.

You may recall I tried to put a diagram together based on what Mary-Jo Foley and other Microsoft watchers had gleaned in the weeks prior to PDC on what the Cloud OS would look like. So now I have a few corrections to make, but thanks to my sources, it wasn’t too far removed from what has been revealed today. Here’s an update version of my diagram, as Ray Ozzie revealed today:

Tomorrow, I’ll drill down into some of these components.

A few more things.

You will be able to download all of the PDC sessions from channel 9 24 hours after the event – so the keynote will be available at 3:30pm on Tuesday 28th GMT from here.

Mary-Jo has also posted a transcript of the keynote here

OK, so its the end of day one, and time to sit back and reflect on the reality of todays announcements. After the hyperbole and back-clapping of the keynote, what really stands out about Azure?

Here’s the key diagram:

Here are my key takeaways:

It’s open platform – I was surprised and pleased that Azure will support non MS languages and environments including PHP, Java and Eclipse. In this respect it really does aim to be a utility computing platform.

Extending .NET – If you know how to develop in .NET and Visual Studio, then all your skills are transferable to Azure, the amount of new stuff you need to learn is fairly minimal. Its also great that you have a cloud environment locally to test your new Azure services.

Model driven – Defining how Azure services are run and deployed is done through “Oslo” i.e. is done through a modelling environment. This shows that Oslo is already part of cloud services, I had feared that Oslo was going to initially be limited to more traditional on-premise server-based development, but it appears it will be core to Azure as well.

Manageability is built-in from the ground-up. Microsoft seem to have spent considerable time ensuring that  IT-Pro requirements are addressed in terms of manageability. Providing simple management consoles, and integration with System Centre from the outset should help with supportability. The ability to automatically detect failures and replace failed or missing resources transparently and guarantee upgrades without downtime if true could significantly reduce total cost of ownership.

A true utility model – in alignment with Nicholas Carr’s writings, Microsoft will bill by unit of consumption and provide infinite scale, zero downtime, and back this with enterprise SLA’s backed with financial compensation (though they hinted that you will pay more for the more 9′s % of uptime you require).

But – its not there yet. Its all still early days, and many of the services required are not available in the CTP. They say this is because they are looking for our feedback, and will adjust accordingly, but also I think this is because alot of this is  still being built – the MS duck is calm and confidant at the surface – but pedalling like fury underneath.

I have to say, based on some other sessions today, Microsoft have got a really good balance between on-premise and in-cloud, and most imprtantly, appear to have a seamless model that will allows an enterprise to decide on the split between the two and optimise accordingly.

So why would you use Microsoft (or anybody elese for tha matter) cloud offering over on premise development? If you need to service a large, growing, but unpredictable number of customers, with global distribution, and 24×7 availability backed by SLA’s, then it appears this model could provide lower cost and lower complexity. Or, if you can’t afford a high capital outlay and need to grow fast, then this is the modal for you. Why choose Microsoft over Google or Amazon? Right now, I would say because if you are a Microsoft house you can leverage your exisitng skills, if not, then maybe their new openness would tempt you.

Take a look at www.azure.com to register for the CTP and downoad the SDK’s and add-ins for VS.

I would also recommend looking at Mary-Jo Foleys blog for the best analysis right here and now at PDC.

More tomorrow – keep watching, and send me your questions – I’ll do my best to get them answered.

Some highlights from Ray Ozzie keynote at PDC. This is raw, and I will update and make it prettier later.

Announcing Windows Azure – services in the cloud

This is the Windows offering at web tier offering ‘Windows in the Cloud’

Azure is the lowest level foundation for providing a scalable service

Automated service management, system provisioning and distribution

It’s a new service based operating environment

MS realised it is Important to use existing skills and code – existing rules and runtime

Fundamentally open environment

The Cloud design point is fundamentally new- needs to be different from other development

Not scale-up like past dev. New ground rules, patterns and practices, deployment – designed for paralle l computing and horizontal scale

Azure is a service running on MS data centres. Released as CTP with the first few services, it has a long roadmap, will incorporate your feedback

MS are betting on Azure as the highest scale, highest availability, most economical way to build services.

Azure services platform

Live services, .NET services, SQL services, Sharepoint (workflow, identity) services, Dynamiccs CRM

The platform underpins online services: windows live, office live, exchange online, Sharepoint online, CRM online

Windows Azure (project red dog)

Kernel of the cloud platform

What makes it special – scalable hosting environment  for you to deploy your apps in cloud

Multiple levels of security – ipfilter, vlans, hyper-visor enforces isolation, access control

Manages entire global data centre infrastructure distributed across the globe, manages all this complexity and provides layer of abstraction to ease programming burden

Automated service management – how to upgrade without bringing down – separates apps from OS and manages both separately

At its heart, the fabric controller manages lifecycle of services.

To change service, specify desired end-state, Fabric controller manages change to this. Manages services not just servers

You define service model:

  • Roles and groups
  • Channels and endpoints
  • Interfaces
  • Configuration setting

Using these can you compose complex service models

Service model are stored as XML files

Azure provides high availability – all components built to be HA. Highly redundant and fault tolerant

Not even double failure can bring these services down, replicated possibly across thousands of machines – no user intervention

Subsystems of Azure fabric controller:

Developers can use familiar desktop technologies, no need to deploy to cloud for test

Deep VS integration, managed and unmanaged

New Azure projects within Visual Studio

Developers can debug and test in normal way using breakpoints and ‘cloud on your desktop’ development server running offline on your laptop.

Apps are deployed to the cloud by choosing a publish option in VS. This then takes developer to Azure development portal. You can map any dns you own to your cloud  application.

2 files are published -  package file (output of bin folder), and an xml config file

Azure is an open platform:

  • Command-line interfaces
  • REST protocls
  • XML
  • Managed and native code

We welcome 3rd party apps – e.g eclipse and php

Scalable, available, rich developer exp, model driven service lifecycle, enterprise ready

Azure Developer Services -

5th generation of computing – 1970smonolithhic, 1980S-client-server, 1990S-web, todAY: SOA, 2009+services

Most SOA apps do not scale out well, as we move to cloud based env, create with app patterns designed to scale out from the beginning

Service requirments: Interop, BP, Identity & security, data mgmet and compliance, services mgmet

Reduce app fron capital cost, management cost,

24×7 uptime provided by underlying mechnaisms

.NET Services –

Service Bus – connect into cloud securely traversing firewalls, Access Control – federated identity providers both on premises and into cloud, Workflow Services – scales out across thousands of services WF workflow services and put in cloud workflows that span on premise and into cloud.

Pool of resources available to you to do things in you app very simply

Identity – users control their own identities. Single, federated identity platform, open and interoperable

Code name “Geneva”. Connect to AD on premise

SQL Services -

Database, Data Sync, Reporting,

Data mining, ETL, Reference Data

I’ll update this later and make it more coherent.

As we get closer to the Microsoft Premier Developer Conference on the 26thOctober, media (and blogosphere) hype and speculation is rising. It’s expected that there will be major announcement on a number of key technology areas from Microsoft at the conference:

  • Oslo (MS’s modelling strategy)
  • Windows 7 (MS’ future desktop O/S)
  • Biztalk 2009
  • Windows Strata (MS Cloud computing platform and strategy)

A key indicator of how important MS sees this year’s PDC is that they interrupted its normal two year cycle and didn’t hold it last year – so they could have it this year to coincide with these big announcements. Previous big strategies that MS announced at previous PDC’s have included Internet Explorer, .NET etc.
This last area has attracted the most speculation recently, but also has the sketchiest information. Based on what I have read, I thought I would have a go at divining what Microsoft cloud architecture might look like. This is based on what I know of other cloud architectures by Amazon and Google, and a trawl of rumour surrounding key code-words currently in use (lots collected by Microsoft watcher Mary-Jo Foley). Even ‘Windows Strata’ as the overall name is not certain, as this was associated with a number of presentations on the PDC agenda, but has recently disappeared again!

None of what Microsoft is announcing is pioneering – others started to offer cloud infrastructures offering similar services over a year ago, but what Microsoft is doing as a ’fast-follower’ in this space is ironing out the wrinkles and making all this work and deliver value to businesses. Think Netscape, think Word Perfect, think the Microsoft Elephant, panicked by a threat to its revenues, rampaging into the cloud space and squishing everything else in its path. In this post I’m going to cover Windows Strata.

OK so here goes. Windows Strata is a multi-layered approach to providing cloud services that will enable Microsoft Software + Services strategy – i.e. a hybrid of desktop applications and power mixed with in-the-cloud services to provide ‘information on demand’ to the enterprise, whether the internet is available or not.

Windows Strata could have three main layers:

  • Foundation Services
  • Infrastructure Services
  • Application Services

In addition on top of these will sit Microsoft’s  ‘Live Mesh’ applications which can themselves be utilised as services. End user applications will be written using services at any of these levels. This is what I think this may look like:

Now lets take a look at some of the code words and lower level components in each layer.

Foundations Services

There has been quite a bit of mention of ‘Red dog’.  Mary Jane Foley (eminent Microsoft watcher and blogger) believes that Red Dog will turn out to be the core platform services that form the ‘cloud OS’. This is what the basic hosted platform will be – running on top of Microsoft global-class data centres. To fulfil this function we can speculate that Red dog will as a minimum provide basic storage, networking and processing, and probably virtualisation and ops management as well.

But, hedging my bets, Red dog may actually refer to the Windows Strata platform as a whole, which is fine, within it you will still fine a foundation services layer of some kind.

Infrastructure Services

At the next level up, we can put a bit more flesh on the bones. Microsoft has already delivered beta or CTP’s of a number of services that fit into this layer, and more are likely to follow:

  • Biztalk Services.

This miss-named component (it doesn’t use Biztalk as we know it) will provide integration services and workflow services to your cloud applications. This is now what the illustrious Darren Jefford is actively blogging about (he has some sample apps that are just 5 lines of code ;) )

  • SQL Server Data Services.

This is, effectively an RDBMS in the cloud that provides on-demand database storage and query processing. No more worrying your database is going to run out of space or grind to a halt.

  • Windows Live ID.

Or more accurately Live ID Identity Services, is the identity gateway for all Microsoft online services, Live ID will provide federated identity and authentication services.

What is immediately setting these services apart from the competition is guaranteed scalability and reliability. As announced at Microsoft Architect Insight in April by Burt Kawasaki, Microsoft is going to be offering a business-ready SLA on these services which will include protection against data loss.

Application Services

This is the layer that is probably most to do with the initiatives code-named ‘Zurich’. Of course, Zurich could refer to all the layers, in the same way that Red dog could. Anyway whether it’s Zurich or not, Microsoft’s cloud offering will need an application services layer. Here’s what could be in it:

  • .NET Services will extend the .NET development platform into the cloud.  That could mean that  it will provide development services – model, code compile test and deploy as a cloud service. No need to be creating an expensive development environment in house – just use Microsofts. Or it could mean that they are extension .NET to better enable control of cloud computing resources for cloud based services. Hopefully both.
  • Mesh Framework.This is what was previously referred to as ‘Live Mesh Platform’. And its one of the keys to Software+Services. It provides a framework to build apps that ‘live’ in the Live Mesh, that is, that can be run both through a browser and off-line, and also has wrappers to utilise many of the infrastructure services. You will be able to build applications on the Mesh Framework in Silverlight 2 or Javascript. Mesh applications run in an Internet Exploer wrapper called Mesh Operating Environment (MOE). I’m wondering what the relationship will between Mesh Framework and .NET services
  • Dynamic xRM (code name Titan). This is Microsoft attempt at a Saleforce.com killer, and has already been released. Behold, Microsoft Dynamics applications will now be available in the cloud. The x stands for everything, i.e. going beyond the customer in CRM and managing any entity, internal or external. Now this in itself is nothing to do with the cloud – its about opening up Dynamics CRM as a platform so you can do stuff like (important for us) use it to roll-your-own Case Management system – which is interesting. But it goes beyond that. Titan was the codeword for Dynamics CRM 4, which for the first time has introduced Dynamics as a hosted solution in the cloud. But more than that, resellers can build applications (like a Case Management System) on top of Dynamics, and host it themselves. So not strictly part of the Microsoft cloud offering, but tangential and worth mentioning,

Cloud\Live Mesh Applications
So this is where applications utilising the Strata cloud infrastructure will sit. At the moment, this is largely the ‘Mesh’ enabled applications that will take advantage of the Mesh Framework to do peer to peer device independent data and always-available data through the online-offline software plus services model.  Microsoft already has a number of these applications that stand on their own, but also can be mashed-up as a service to create composite applications, for example live search and live earth. You can expect more in the way of Live Mesh applications, including I hear, Office applications (a demo of this is expected at PDC) but maybe different from the current Live Office (maybe  fuller functionality?).

So it will be interetsing to see how much of this guesswork turns out to be true. I loo forward to PDC!

Earlier in the year Jon Woodward and I referenced in a white paper some earlier work by Gartner that envisaged a world where service aggregators provided the major way for enterprises to benefit from cloud services. These aggregators would assemble whole or parts of business processes for an enterprise, choosing from highly competitive and specialised service design companies.

Well judging by Phil Wainrights post on Sequoia’s funding for Appirio, part of that vision has just got closer a bit closer to reality:

“Appirio is a leader in cloud integration in three core ways:

  • A strong focus on Google Apps projects alongside Salesforce.com
  • Specialization in ’serverless’ (ie cloud-to-cloud) projects (more on that below, plus a controversial prediction on Exchange)
  • Its combination of ‘trusted advisor’ professional services with productized on-demand services”

Phil goes on to say

“But Appirio has also carved out a fairly unique profile for itself in focusing on Google Apps projects as well as Salesforce.com. It already has over 30 enterprise projects built around Google Apps. More than most of its competitors, the company is truly a cloud integrator in that its core expertise (and indeed its product set) is around making cloud applications work together better. This makes it a leading enabler of the serverless enterprise, ie running businesses entirely on the cloud without any servers installed on-premise at all”

Which is pretty close to gartner original vision. It remains to be seen whether the emergence of service aggregators will lead to the proliferation of small specialist service providers. It looks like the Appirio model will fill both roles.

As a founder member of the new user group, I was somewhat nervous that we wouldn’t get a good attendance at our first meeting, especially as it was held in the evening. I needn’t have worried – we had 25 people turn up out of 30 that registered, both from interators and end users – a very promising start.

Thanks to Microsoft who provided the venue at their Cardinal Place offices in London and laid on food. We were able to agree the user group constitution, and we had an excellent keynote on the implications of OSLO from Black Marble, Solidsoft presented on modelling and I (SunGard) got to open up a discussion on what companies should be doing now in SOA, cloud computing and modelling to prepare for the future.

Of course, the best bit was the opportunity to network and chew over real world problems with a great community of people. All in all a great evening, and I’m confident the user group will go from strength to strength.

If your interested, come and register at our website: http://sbug.org.uk/

I’m just back from a TTI\Vangard conference discussing the implication of Digital Natives entering the workplace.  There has been alot of discussion recently on  how this new generation will have major disruptive impact on businesses mainly due to their atitude to workplace practices and technology.

The nice thing about the format of this conferences is that after a 10 minute head start, anyone on  the floor is free to ask questions (we each had a microphone), sort of professional heckling, which led to some excellent open discussions. As long as you don’t expect the presenter to complete their slide decks, its a great experience.

If your new to the subject I can recommend a couple of books – Generation Blend by Rob Salkowitz and Wikinomics by Don Tapscott.

Some interesting viewpoints were put forward on the subject that I hadn’t heard before -

There’s some evidence that the behaviours of digital natives are age-related rather than specifically to do with their upbringing in the digital age. Some firms are reporting that as they mature through the work force, this generations habits are changing and becoming more corporate. Maybe getting responsibilities (wife, kids, job) is actually modifying their behaviours, or maybe is pressure to conform to do well in a work environment. This might be good news for corporate IT department who are scared of the degree of openness and security naivety expressed by this generation, but is also bad if it means stifling the creative and collaborative working practices that are a defining characteristic of this group.

Some argues that the definition of the digital native was very much a middle-class, educated, north-American perspective on growing up, and that in the vast majority of the world, these behaviours were limited to an elite few with access to the technology and a level of education to match.Even amongst the privileged group, use of what we would consider core web2.0 technologies is patchy. For instance given a questionaire on what common web2.0 sites were used, many digital natives did not know about or use for example Wikipaedia or Flickr, or understand what an RSS feed is.

Another insight was that being a digital native did not necessarily mean a good understanding of the technology – quite the opposite in fact – from them technology is a commodity and they know no more about how it operates than about how their car works. This can mean that we technocrats who maybe have a similar level of use of technology, but a much deeper understanding of its working, can make assumptions about what digital natives understand that is completely wrong. For example, many digital natives did not understand the implications of the difference between http and https. This level of naivety has implications when these people enter the workforce.

A highlight of the conference was meeting many ‘digital natives’ who are active in creating and running web2.0 sites such as Reddit, Skribbit and Jooce. The dynamism and creativty of these young people is a great source of hope for the future of the web. But as one digital native said whay had we come to a conference run by adults to learn about what kids were doing online today?

All in all a very though provoking and enjoyable conference.

Next Page »

Follow

Get every new post delivered to your Inbox.