Hudson to Jenkins

After evil company acquired SUN, Hudson became a servant of them. So the good people thought that we have to survive in peace and be confident. And they re-branded our hardworking crew to Jenkins.
Long live new friend Jenkins!

OSYM sucks in Turkey

There exists a rumor about OSYM(Student Selection and Registration Center) in Turkey about the encrypted exam books. It is said that there exists a key for all answers can bu found with the right decryption algorithm. This means that until today, all the exams in Turkey are unreliable, untrusted and to make people believe that lie. Are we fooled by government?
So there should be an objective investigation about this exam for over all years. And the justice should be applied.

And a funny thing that the OSYM minister told that there is no encryption, than it is proven there is encryption. Then they told that the encrypted exam book is only for press release and never given to any of the exam participants, today we saw that the encrypted book is released before the exam has been taken. I hope they have a reasonable explanation about it.

Notes From Scandev2011-The Rules of SOA

SpokesPerson: Jeff Genender

-You have to use the right tool for the right problem.
-SOA is not only "Webservice", it is a way of simplifying the integration between different domains.
-You should be aware of the development patterns.
-You have to be aware of what is XML, SOAP and be aware of the cost,side effect, benefit of using them.

MapReduce?? Hmm should I know this!

If you are a software engineer than you should have the idea. You have Map and Reduction, (again we are dividing and going to conquer). We are trying to reduce the weight of problems by delegating the problems into sub nodes, called as worker nodes.

MapReduce framework is patented by Google.
-What patented?
-Don't wory my friend. You don't need to pay any money to Google. This is just for safety.

more on that

Notes From Scandev2011-Apache Hadoop

If you have chunks of data and have to do some manipulation or analysis on that, you probably need to use Hadoop. You don't have to, but life is easier with them. In SDC2011 Josh Devins from Nokia explained how to do log analysis with Hadoop and Pig.(If you are interested in working in Nokia on these cool stuff you may contact them) In some cases Log analysis is a really resource consuming process as you know. By using HDFS-Hadoop-MapReduce technologies you can save your time for your kids.
I don't want to explain every detail of these technologies but if you are in trouble these technologies may save your life. Before you need to use, you have to learn first. If you will be aware of the tools you have, than you can be aware of the problems arising in front of your eyes before they smack on your face. (Google, Facebook, Yahoo, Microsoft and more others are using... so they know something I guess ;) )
If you are interested-->

Notes From Scandev2011-Git Going with a Distributed Version Control System

Spokesperson:Matthew McCullough

The most interesting thing for me is you can make partial commits on the same file with git!
It is possible to search on commit comments.
It is possible to add a control before commit, you can say: Do the commit, if the test is successful.
And Matthew gave some real world examples how fast git is. Its power is impressive. I will try that.
And he mentioned that it is very easy to migrate projects from other source control system to git.

Notes From Scandev2011-NoSQL

Spokesperson:Tim Berglund

And this presentation is about a new approach to storing data in physical space in some other format, and keeping the relational information together to remove the relation information from tables.
He talked about the CAP theorem which can be rewritten as Consistency-Availability-Partition Tolerance, which our relational DB's are not capable of having three of them together.
He gave some example of NoSQL DB's and explained how they manage the most important assets DataModel-Query-Scale.
Some Examples:
Cassandra :DataModel:Big Table-column based(Consistent Hashing)
MangoDB :DataModel:JSON (No Schema needed, object is saved how it is!)
Redis :DataModel:Key/Value store(Redis is a DataStructure Server, wrting to disc optional)
Neo4J :DataModel:REST/JSON, DataMetaphor:Graph