10 project management lessons from the Titanic disaster

1: You need to know what you’re measuring
Frederick Brook says, milestones should be objectively measurable. If you do not have valid measurements for your project, you will run into problems.


2: Assumptions can kill you
Maybe we assumed that a particular system was using a newer software release than it actually was. 
Maybe we assumed that another department would take care of ordering cable.
Maybe we assumed that the vendor received our critical email message. 
Assumptions are important in your work, but if you proceed on the basis of them, make sure everyone is clear about what assumptions you are making.


3: Distractions are dangerous
If enough members of your team encounter enough distractions, your project will gradually fall behind.


4: Little things add up
The small delays are just as critical as the large ones, meaning that adherence to milestones is critical to the success of a project.


5: Stakeholders should be kept informed
Our stakeholders need to know about the status and progress of your project. Keeping them informed will keep them happier.


6: Other people’s perspectives matter
If a client asks a question, try to see beyond the question itself to the motivation behind the question. 
If a technical person is explaining a function of a system or program, make sure the explanation avoids jargon. 
Clear communication will lead to happier clients.


7: Moving targets can hurt you
Any change is rarely “small.” Rather, it typically involves changes to other parts of a system, results in greater complexity, and requires more testing.


8: Traceability is essential
How familiar are you with the strategic objectives of your company? 
Can you find a logical connection between the requirements of your project and those strategic objectives? 


9: Methodology is more important than technology
You might want to use sophisticated planning and tracking software and tools. 
More important, though, is that your plan be solid. 
The best software in the world will not save a poorly designed plan.


10: Documentation may have lasting benefits
Documentation is often the most important part of the project because it may exist long after the project team has disbanded.


http://www.techrepublic.com/blog/10things/10-project-management-lessons-from-the-titanic-disaster/3174

Becoming A Better Developer

• Motivation
• If you’re not growing, you’re dying…
• Confidence
• Good developers are usually quite confident in their abilities. Why? Because they know that they have put in the effort to truly understand what they are talking about. Be confident in what you know.
• Choose Concepts Over Implementations
• If you understand the concepts, that enables two things. 
• First, you can use those concepts in other languages with ease. But far more importantly it lets you think abstractly about different ways to solve a problem. 
• It also lets you learn from other communities. So rather than waiting for somebody else to solve a problem in your language, or looking for help in your language, you can look anywhere for help. 
• Read At Least One Book per Week (well~it's very difficult for me >_<)
• There are two main things that you’ll benefit from with this approach. 
• First, by reading actual books, you’ll focus on exploring a concept in depth.
• The second advantage to the one-book-per-week approach is that it will make you push your boundaries. 
• Share Your Knowledge
• The best way to learn is to teach.
• Become a Mentor
• The important thing is to look at the mistakes they are making and help them to understand why they are making the mistake. The obvious benefit here is that you are “paying back” the community and investing in the future of the industry.

For complete information : http://blog.ircmaxell.com/2011/11/becoming-better-developer.html

Apache Commons應用 - CollectionUtils

Requirement

報表輸出結果，應該依序輸出「營業事業所得稅」、「綜合所得稅」、「遺產稅」、「贈與稅」、「貨物稅」、「菸酒稅」、「證券交易稅」、「期貨交易稅」、「營業稅」、「特銷稅」。如果沒有資料，該稅別要顯示出來，數字要顯示0。結果如下，

根據user requirement要求的輸出結果，並不是按照tax code或者tax name的筆劃順序進行排序，此部份必須要根據query回來的的結果來做加工。

Problem

原本的作法是用SQL將統計資訊算出來，但是其只會傳回有存在的資料，若沒有符合條件，就不會回傳資料，故只有回傳兩筆資料。

從SQL回傳回來的結果會是一個List of value object，以目前的條件，只會有兩個value object，分別是「營利事業所得稅」與「期貨交易稅」。

Solution
由於資料固定只有10筆，放在SQL裡頭做的話，SQL會很複雜，日後維護不易，故此部份的資料加工放在AP端進行處理。

1. 建立一個新的List物件, List result = new ArrayList();
2. 根據回傳回來的list of value object來進行判斷
1. List中是否有「營業事業所得稅」的資料
• 若「有」，從List中找出該object，並add到新的List物件
• 若「無」，add一筆空的資料到新的List物件
2. List中是否有「 綜合所得稅 」的資料
• 若「有」，從List中找出該object，並add到新的List物件
• 若「無」，add一筆空的資料到新的List物件
3. List中是否有「 遺產稅 」的資料
• 若「有」，從List中找出該object，並add到新的List物件
• 若「無」，add一筆空的資料到新的List物件
4. 「 贈與稅 」、「 貨物稅 」、「 菸酒稅 」、「 證券交易稅 」、「 期貨交易稅 」、「 營業稅 」、「 特銷稅 」依此類推

Apache Commons有提供CollectionsUtils此API，其提供不少utility method來針對Collection物件進行一些處理，根據以上需求有兩個:

1. 從List中check是否有存在特定的object
2. 從List中取得特定的object

//List中是否有「營業事業所得稅」的資料

1:  private static boolean is35Exist(List list) {  
2:      return CollectionUtils.exists(list, new Predicate() {  
3:        @Override  
4:        public boolean evaluate(Object obj) {  
5:          NIG630ReportBean bean = (NIG630ReportBean) obj;  
6:          return "營利事業所得稅".equals(bean.getTaxNm());  
7:        }  
8:      });  
9:    }  

//從List中找出該object

1:  private static NIG630ReportBean find35(List list){  
2:      return (NIG630ReportBean) CollectionUtils.find(list, new Predicate() {  
3:        @Override  
4:        public boolean evaluate(Object obj) {  
5:          NIG630ReportBean bean = (NIG630ReportBean) obj;  
6:          return "營利事業所得稅".equals(bean.getTaxNm());  
7:        }  
8:      });  
9:    }  

Stop Redeploying in iReport Development

Context
以下是本案的project structure:


所有的jrxml放在jasperreports folder中，當jrxml有異動，處理程序如下：
1. 執行maven build
2.  將jrxml compile成jasper file會搬到nig-web\src\main\webapp\WEB-INF\classes\META-INF\jasperreport
3. 包成ear file
4. deploy到application server

Problem
在用iReport開發的時候，最浪費時間的部份，就是在我們無論做任何的細微的修改，如果要看到修改後的結果，都要將jrxml compile成jasper，然後redeploy到application server，此過程曠日廢時，浪費許多寶貴的開發時間。

Solution

Step1. 打開iReport工具-->工具--選項

Step2. iReport--> Compilation and execution--> unchecked "Use Report Directory to compile"-->將「nig-web\src\main\webapp\WEB-INF\classes\META-INF\jasperreports」貼進 "Compilation directory"-->click 「確定」按鈕

設定完成後，當我們在iReport Designer做任何異動，click "Compile Report" button，iReport會自動將jasper file搬移到指定目錄，application server會自動抓到最新的jasper file，不用再rebuild and redeploy了。

iReport print when expression 應用

Requirement
在報表呈現的內容，每筆資料之間，需要有間隔的虛線如下

Problem
原本iReport designer中的作法，是將line放在detail band中的fields下方，screenshot如下：

但是此舉，在最後一筆資料，也會出現分隔線，此結果不合理。

Solution
step1. 將分隔線放到detail band的fields上方，如此一來，列印順序是line-->data-->line-->date-->line-->data，以此類推

 step2. 由於列印順序是line-->data-->line-->data-->line-->data，為了滿足需求，只要讓第一次出現line的時候，不要讓顯示出來就好。此時，click line，並在attribute視窗中的print when expression輸入條件，當是第一筆的時候，不要印出line，從第二筆資料開始即可。

Project Manager and Subject Matter Expert

對於project manager這個角色，到底需不需要有技術能力，一直是常在爭論的議題。
這篇文章倒是給了一個不錯的答案：『有會更好』


他提到了，如果是technical project manager可以帶來幾個好處

1. 遇到問題的時候，他更能掌握問題的難度，更能掌握要花多少時間來解決這個問題，對於時程的掌握會更踏實
2. 你對於你的團隊更有同理心，因為你知道問題的難度。project manager總是想急著解決問題，常會脫口而出：『這個問題有這麼難嗎？有需要這麼久嗎？』
3. 由於你有技術的底子，對於某些問題，若有過往的經驗，可以更快找出解決方案，並帶了團隊走向正確的方向，避免花額外的時間去做trial and error

以下是摘錄全文中的部分：

• Scope and Schedule – If you have a background in the project you are managing it becomes much easier to understand what the project is trying to accomplish.  This helps you in defining the scope of the project and setting the schedule.  You will know roughly how long it takes to build a server, install a circuit, or design a network.
• Empathy – If you understand the work, it makes it much easier to empathize with the team.  I have built some strong relationships with the network engineers I work with, because they think of me as one of them and not just another project manager.  A project manager should never under estimate the value of relationships.
• Misinformation and Solutions – You will have a much better sense of when you are getting the wrong information and be able to contribute to solutions.
• Ramp Up – Most projects start life underfunded and behind schedule.  It is an unfortunate fact of life for project managers.  If you know the subject matter, you will spend more time focused on the project than trying to understand it.  This will save you significant precise time.

全文請看：http://www.cerebellumstrategies.com/project-manager-expert/

5 Reasons Why Software Quality Matters to your Business

Predictability
Do it once and do it right, and there will be less re-work, less variation in productivity and better performance overall.


Reputation
Some companies have a reputation for building quality software. 
A good, solid reputation is hard to establish and easy to lose, but when your company has it, it’s a powerful business driver. 


Employee Morale
The most productive and happy employees have pride in their work. Enabling employees to build quality software will drive a much higher level of morale and productivity.


Customer Satisfaction
A quality product satisfies the customer. A satisfied customer comes back for more and provides positive referrals. 


Bottom Line
Predictable and productive performance, a stellar reputation, happy employees, and satisfied customers are the formula for a successful software business.


http://blog.smartbear.com/post/12-04-03/5-reasons-why-software-quality-matters-to-your-business/

12 Ways to Become a Better Programmer

1. Read Other People’s Code
Reading other people’s code will allow you to understand different points of view in problem solving as well as introduce you to new techniques. It is important to have an objective to your reading, such as learning a new programming paradigm or how to work with a new API. 


2. Develop Personal Motivation
Being motivated to become better at your chosen profession is an excellent step in the right direction.


3. Get a Mentor
This should be a person that you aspire to be like professionally, that has a background that you can relate to and that you have a lot to learn from. 


4. Learn to Listen
Learn to listen to what the customer has to say, listen to your coworkers when they are discussing the project. Don’t assume that what you think is really what they want. 


5. Gain Confidence
Why are good programmers so confident? They know they have put in the time and effort necessary to truly understand what they are talking about. This knowledge enables them to make the best decisions and be able to discuss or defend them in a productive way when needed. 


6. Join a User Group
hese groups are usually structured around two concepts: teaching and networking. Not only will you be able to learn something new at each monthly presentation but you will gain valuable contacts in your industry that could later lead to internships, jobs or mentors.


7. Read at Least Two Books a Month
By reading books you will be focusing and exploring a concept in depth as opposed to “grazing” on information such as blogs, tweets and other online reading that rarely goes in-depth on a topic. This way, you will be better able to take these lessons and concepts and apply them to the real world.


8. Learn to Use Your Debugger
For some programmers, finding bugs means you did not spend enough time planning or that you have been sloppy during your code writing. Errors in your code should not be taken personally. Humans will continue to make mistakes and that is something we have to accept. Stop looking at your debugger as an evil you need to avoid and see it as a tool you can use to create better code.

9. Never Stop Learning
Over the years the continuous pattern of framework upgrades, version numbers and release dates start to run together but you must keep your skills up to date by evaluating new technology and trends as they are happening. 


10. Share Your Knowledge
The best way to learn something is to teach it to someone else.


11. Learn Keyboard Shortcuts
For programmer this means knowing how to type quickly and accurately as well as know your way around keyboard shortcuts that will make you a faster developer.


12. Learn When to Take a Break
When working on a task such as programming that requires concentration, you will often find that you are making more mistakes or creating more problems when you are tired. Stopping programming is often a better solution when tired than having to fix the mess you made.


http://www.learncomputer.com/12-ways-to-become-better-programmer/