• From pplication to technology platform
• The strategy is to
  • Increase selection
  • low prices
  • Confidence
• Books are a big selection thats why it was selected by the beginning
• the feed back is important
• Invite 3d Parties into the platform
• Amazon itself is a retailer and has to compete like others on the platform
• With more retailers the platform had to develope for new data types (are different sizes of a Jean deffirent items?) 
• Open the data for any developer
• You should not be surprised of the unprofitable use of your services, you ll be rewarded somehow
• Started exctracting functionality code for big third parties (target)

• 2001 : Cant build anymore
• 7 years to kill the previous architecture
• Turned to be a lot of smal standalone teams (technology diverese) each responsible for its services
• Unexpected popularity can be a problem
• Getting Real >> Priorities: It is too hard to scale

• Two Phases Transaction doesnt work and fails (avalability)
• Transaction should be in a one box
• There is no technology that can do it

     

• Turning all conferences into a coffee break !
• Principles
  • the right people show up (who want to be there)
  • It starts when it starts
  • It is not over till it is over (if it is not over , you should find a way to continue it)
• It is driven by both passion and responsiblity (one cant stand without the other)
• Host has an addition role :
  • to make sure that sessions start and continue (no matter what the subject is)
  • Makes sure that people participate, or assign someone to do that
  • Make sure the that the notes get recorded
• The one law of open space is the law of two feet: If you are not participating or learning then you take your two feets and you go where you think you can
  • ButterFly: People moving arround spending 5 min in each session, or maybe create there sessions
  • Bumblebees      : People move ideas between sessions
• Self organized
•  

If Architect doesnt run code then there is a problem.

Architecture doesnt live in  documents

Architecture embodies the critecal design that typify a system

• Good Architecture is a light weight upfront one that enables you to take decisions in the future.
• It helps to see the big picture
• It is Stable but NOT static!
• Dont play dumb, if you know something and sure about it then take a decision…
• Documentation is not a skill that developers have or like.
• Architecture of seperation of concerns not a divorce!
• Defering decisions to the last RESPONSIVE moment!
• TDD is not about design shape, it is about getting code right!
• Architect is not a verb, the verb is Design, and one of its productions is Architecture
• TDD works well when you have a right upfront architecture that shapes out the system. Otherwise it will turn the system into procedures wrapped in objects!
• we need a balance of details in an agile architecture, just enough details to help us taking decisions…
• Architecture deffinition (not a finished one, not done but done right!) in a scrum process can be the first sprint, but you deliver code!structure classes that build. We call it Inception

 

• Without it , softwares fail in advanced sprints because of lack of it, after being maybe released in the market! 
• Think of time software decomposition as  well as the classical one.

• Projects are late, developers are blamed!
• scrum is to release people from their chains!
• Scrum is based on one of japaneese best practices  (Toyota)
  • We place the highest value on actual implementation and taking action.
  • There are many things one doesnt understand th therefore, we ask them why dont you just go and  do them.
• Developers Start to feel responsible
• Dicipline should be brought by creative leadership…
• Weekly meeting of all stakeholders
• Companies like to “start light”, by choosing features from a methodolgy! Then they blame it for any fail.
  • Behind the ABC of scrum there are certain needs such like collaboration,
  • google in its evoleving towards agility observed what problems they got
    • got problems with deadlines: Introduce a Release BurndownChart
    • got problems with testability because of dependency with other developers work : they used Scrum daily meetings
    • Google then failed on delivering on date
• Introducing full Scrum implementation
  • Product/Release Backlog
  • Iteration Based Development (2 weeks iteration)
  • Testing tasks for iteration’s feature s in the same Iteration
• Sprint Burndown  to guide actual development efforts
• Release Burndown to guide the overall release planning

Practices: Transperancy >> Daily scrum taskboard

 

Scrum of scrums for communication teams

The scrum check list

When Scrum is will implemented by a good scrum leader, Scrum starts running itself.

Customer Responsibilities Board designed for assuring that customers dont get 50% of their needs, thats a 100% project failure!

Scrum makes presonal team individuals presonal problems as important as tachtical problems , up on the top, for improving communication and productivity.

Scrum introduces something like a viruse that activates the communication and productivity cycle.

Scrum is introducing a management roles that responsible to make a strategy that works… it was suggested by CMMI and introduced by  Jeff Sutherland and Ken Schwaber and reviewed by Scrum Alliance.

• Metadata Driven?
  • meta data is dangerous (stopping the world to describe it)
  • The format you get your message in is quite unboundable because of meta-meta-Data
• It’s all about messaging, messages systems are natural way of thinking
• cool URIs Instead!
• Wsdl is a big horrible documant
• Rest is browser friendly (using more safe protocols like get  rather than post )
• So why all the WS-* hell?

Anne Thomas-Manes, Burton Group Analyst

SOa is (or should be) a business initiative

Making IT more responsive and cost effective

Don’t Expect Business people to see value in SOA technology

You must represent the value in terms they understand

Dont try to sell SOA to an unreceptive aifience

• sell business projects based in business value
• Follow SOA principle when implementing

SOA concepts

• from a technical perspective
  • An architecture for designing systems
    • a “service” is a core unit of design, a service implements a discrete piece of functionality, exposes its functionality via a well defines interface
    • SOA is NOT made for interoperability , but the good in having standard service protocols and specifications that it can be consumed by multiple apllications
    • can be consumed by multiple applications
    • If functionality is required by multiple applications, it should be implemented as a service
      • if it wont be reused, it shouldnt be implemented as a service
  • SOA is not about design, not technology
    • is something you do not something you build or bu
    • ESB is an integreation technology , it even helps you to duplicate services!
  • Requires new design approaches
    • service-centric rather than application-cuntric perspective
    • “Proven” design patterns may not apply (we are changing the context  that the pattern where done for)
  • Soa is a lifestyle
    • it is more about culture than about technology
      • technology provides the tools and raw materials, but it is up to to use them effectively
      • SOA impacts planning,finding,development,operations,utilization, and managent of software
    • It’s like physical fitness
      • it requires a lifestyle change
      • The goal is to make IT healthy ( doesnt matter what the technology you use, what matters that u do it, consistently!)
  • IT Challenges
    • IT is not healthy
      • IT budget is enormous
      • IT viewed as unresponsive that results in missed oppotunity
      • IT Budgets share is 20% for new projects and 80% for maintenence!
    • Redundency is the problem! no budget for new projects!!!
  • Strategic business goals
    • Align IT and business, IT should support the business strategy
    • Increase flexibility and agility , to increase responsivity
  • benefits:
    • Give users better access to the information and capabilities they need
    • Increase productivity, efficiency, effectiveness, satisfaction
  • Reduce time-to-market
    • Deliver new features, new capabilities faster
  • Increase ease of doing business, make it easier to integrate with customers, partners, and suppliers
• SOA should be part of a largereffort to reengineer IT
  • Enterprise architecture
    • provide an inderstanding of the IT envoirement: Bussiness and technical strategies >> applications, information and infrastructure architectures
• SOA is a long term strategic investment
  • SOA is an emerging discipline
    • requires new design approaches and collaboration
  • Plan for a 15-20 year maturity cycle
  • SOA will impact planning, fundingm development
• Where are we now?
  • Stage 1: Using SOA technologies for integration (ESB doesnt bring SOA automatically, a minimal governance needed)
    • Generates spagetti integration and little reuse
    • Not SOA
  • Stage 2: Pulling in the reins
    • Focus on defining a standard runtime infrastructure (Do not think in terms of technology)
    • Recongnition of the need for governance
  • Stage 3: Enterprise-level planning
    • adopting an enterprise architecture respective
• Where are we going?
  • Stage 4: Delivering strategic value
    • Building a portfolio of reusable services
    • Decomissioning redundent systems
    • Reduce time to delevirey…

Making the Business Case

• Understand the motivatios of the people holding the purse
• How respective are they to SOA
  • If they are then tie SOA investment to high priority business goals, use it as an opportunity to reengineer IT (reduce redundency)
  • You have to spend money to save money
  • Think globally act locally
    • SOA must be accomplished incrementally - project-by-project
  • If they are lukewarm to SOA
    • Tie SOA investements to important business projects
      • focus on the hot buttons
  • If they are not reseptive to SOA
    • Practice Stealth SOA
    • Sell Business-driven projects
      • Implement SOA principles in the projects
    • Sell infrastructure projects based on hot buttons
      • Cost reduction
      • Compliance
      • Find evidence that u can use next year
  • Gather metrics

Soa can deliver enormous business benefits

• Alignement between IT and business
• Flexibilty and Agility

Erik Meijer, Creator, LINQ 

• The quest is to do web programming easier 1997
• using strings as a glue, gives weired results!
• 2007 :the problem is still there, it became a nightmare Making distributed web programming easy for everyone…
• Solving the Impeance Mismatch..
• Uneversal data model (xml , Uml) is a deadend!
• solution: Linq 1.0:
  • define strandard query operators for each model .
  • support external explicit relationships.

 

•  
  • to have differenct relationships view that is not absolutely inside the objects. 
• Linq 2.0
  • solving the multytier provlem?
  • Make Irreversible Decisions at the very last moment?
    • answer: Stretch current client applications to cover the clouds
    • Use only materials available in the room
    • deploy the application without downloading any extra components
    • we cant program without IDEs anymore
    • pickup your choice of language : Cross language inheritance, the power of dotnet
    • run everywhere
    • Common intermediat language, IL

 

•  
  •  
    • Organic Software :
      • Ajax= Single Tire Application + Refactoring
      • Late Binding into the client program, write one program then bindingit later into a client technology .
      • uselesspickels.com
      • Recursive Tier-splitting through refactoring.
      • Run exactly the same complex validation logic both on the client and on the server.

•  
  •  
    •  
      • Javascript: no typing system that stops you on the way!
      • Tier Splitting
      •    Expansion Teorem, take on the synchronization anf concurrency after refactoring by the compiler. Refactoring will use algorithm to split to tires to assures safety

 

•  
  •  
    •  
      • Compile LINQ quries to gigantic cluster of sql server
      • doing quries   is repartitioning and joining, linq is good at it, typed queries, run quries in different datasources, run the same code on any tools without requiering to alter! Use the materials available!

 

•  
  •  
    •  
      • Linq 1.0 is just the beginning!
      • No paradigm shift needed! take what we have and extent it !
      • (Relationships, contracts, concurrency)
      • tier-splitting refactoring!  

 

Joe Walker, Creator, DWR 

Closures ( sytax sugar )

{int,int => string} c;

{string => } stringUser={

string m =>

System.out.println(m)

}

Closures are usefull for extending the language (FP)

JVM is changing : Special Dynamic Bytecode

Sun Now suppoert 3 languages:

Java,Ruby,JavaScript

The JVM in 2007

• Clustered objects
• Define any object to be clustered from a hashMap to a string Bean
• Innocating use of AOP to keep separate JVM’s in sync

The Libraries are changing:

• JPA replacing Entity Beans
  • Seam joins JPA to Jsf
• Ajax
  • DWR
  • GWT
• Swing starting to look nice
• Virtually all the libraries are open source

Chalanges to Sun

• SCA
• Eclipse / OSGi / SWT
• Adope:Flex / Apollo