Song Way

Creative Commons License
The blog posts here by Song Wei is licensed under a
Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.

November 29, 2006

11.29

Filed under: Uncategorized

SE project

stayed over night with Srikant in the library. we committed our version at 4 a.m., and found that Brandon commited one hour before that.

in the afternoon Brandon told me about the concurrency problem. went ahead with him to fix it. it’s only 1 hour before the demo when we finished it. we’ve been greatly sucked in this 3 hours.

got to demo. i think our project sucked but i had to fight for the team. so did Srikant and Brandon.

Asked Venkat what he really thought about our project. He said :"you don’t have a good team" Well, Venkat the allmighty knows all. But it sucked me so much that we do have good boys in the team and still we’re dragged down. I felt so sucked up.

 Nice:)

Come on, rtss project.

11.28

Filed under: Uncategorized

wrote a paragraph full of shitty ideas to Rong.

se project 

November 27, 2006

11.27

Filed under: Uncategorized

rtss…

WIL

SE project 

11.26

Filed under: Uncategorized

rtss project

measurement

wil reading 

automata assignment

 

November 26, 2006

11.25

Filed under: Uncategorized

wireless test bed stress test.

se project

dinner with Guo Zijing and  Chen Minmin

11.24

Filed under: Uncategorized

reading

11.23

Filed under: Uncategorized

reading

November 22, 2006

11.22

Filed under: Uncategorized

still haven’t got my tax forms!!

ip application form

se project. css. we’ll meet again on the weekend.

rtss programming. 

11.21

Filed under: Uncategorized

Talk with Rong.

programming for RTSS.

Rong is not satisfied with my work. She said there was no initiative for investigation.

I know that. And I know why. I’ve been adapting this whole semester, learning about the adaptive approach, but, ultimately for me it’s a binary choice: to do it or not, whatever I do, wherever I am.

OK. I need to "do" 3 projects in thanks-giving: software engineering project, RTSS project, WIL project.

Called Sara that I’m not going to Dallas for the thanks giving.

11.20

Filed under: Uncategorized

sat in vain almost the whole day in the dorm.

spent 5 mins preparing before the software engineering mid-term 2. 

stayed over-night to write the program for Rong. 

this damn inertia struct me again.

must break it. 

11.19

Filed under: Uncategorized

sat in vain…

it sucks. 

11.18

Filed under: Uncategorized

SE project.

November 18, 2006

11.17

Filed under: Uncategorized

WIL

wireless test bed 

software engineering project 

11.16

Filed under: Uncategorized

reply email 

wireless mesh test bed

WIL 

11.15

Filed under: Uncategorized

Unit Test

11.14

Filed under: Uncategorized

SE assignment — unit test

11.13

Filed under: Uncategorized

WIL

11.12

Filed under: Uncategorized

os assignment

WIL 

11.11

Filed under: Uncategorized

played basketball with Daniel and his friends

went to HEB. 

November 11, 2006

11.10

Filed under: Uncategorized

peer programming and Daniel’s birthday.

11.9

Filed under: Uncategorized

unoffical fedora 5.

 

11.8

Filed under: Uncategorized

os assignment.

11.7

Filed under: Uncategorized

discuss the rtss solution.

11.6

Filed under: Uncategorized

algorithm and coding of WIL

 

November 6, 2006

11.5

Filed under: Uncategorized

WIL paper

prepare the presentation with Greg

WIL programming

software engineering report 

software engineering programming. 

November 5, 2006

11.4

Filed under: Uncategorized

planning and misc stuff in the morning.

peer programming with Anil about the validation.

Helped Soji to install the server.

Skipped through Contour’s dilema.

Any thoughts after reading it?

Yes. Rong’s in fact really nice to me. 

Anymore?

They’re not free. I’ll try my way. 

November 4, 2006

11.3

Filed under: Uncategorized

We’re all dreamers.

Let’s give it a shot! 

November 3, 2006

lessons learnt from the WIL simulation project

Filed under: Uncategorized

I need test driven methodology, since I’m careless. MUnit seems good.

It’s extremely harmful to follow others without establish a framework of my own when dealing with legacy code, especially when the style of my thinking is in such a contrast with others.

I need simplicity. I need to separate the core logic/ algorithm from other messy code dealing with various input and output and formatting.

Although matlab has a profiler, it’s not desireable to spend too much time optimizing the speed in matlab. matlab code is meant for the clarity of the algorithm.

The first task should be identify and solve really important problems when writing simulation. That’s what simulation is all about. 

11.1

Filed under: Uncategorized

It’s extremely important to try out something independently before you’re going to do some real work on it.

November 2, 2006

11.02

Filed under: Uncategorized

Spent too much time with the servers in the lab this morning. Got to reminder myself to focus on using my own brain not computers.

There is a very cool concept from XP: Refactor without mercy.

I need to refactor myself without mercy. 

 

found a cool site:

http://www.markshuttleworth.com/ 

When once you have tasted flight, you will forever walk the earth with your eyes turned skyward, for there you have been and there you will always long to return. — Leonardo da Vinci

November 1, 2006

studying notes on some agile practices

Filed under: Uncategorized

Crystal:

http://alistair.cockburn.us/index.php/Crystal_methodologies_main_foyer

http://proquest.safaribooksonline.com.ezproxy.lib.uh.edu/0201699478/pref01#X2ludGVybmFsX1RvYz94bWxpZD0wMjAxNjk5NDc4L2NoMDE=

Plus

It’s simple. I like the idea of keeping crystal "has the least methodology that could possibly work." In doing it reveals some crucial element of an agile process. And it’s very extensible since it leaves a lot of space for its practitioners. In fact crystal encourages its practitioners to tune it from time to time by adding or removing extra features to or from the process.

       The philosophy of crystal is a Zen abstraction from the details of agile software development. By comprehending such an abstraction, one is expected to be able to absorb all kinds of useful new techniques according to the context on the fly.

Minus

I talked about Zen philosophy of crystal. Just like all eastern philosophies, it’s somewhat obscure and uncertain for beginners. In fact Alistair says in "crystal clear" that crystal requires very experienced programmer as leader, and can not tolerate more than 1 beginner in the team. Also, since most of its elements are based on the ease of communication and simplicity of a small team, it’s unable to be scaled to be used in large or distributed teams. And it’s not suitable for life critical applications since it does not encourage a rigid process.

XP:

http://www.extremeprogramming.org

http://www.extremeprogramming.org/lessons.html

 

Plus

“Going extreme” is an effective way of motivating people who are depressed for long. There is a Chinese saying of “In order to correct a wrong behavior, you have to over adjust yourself.” In fact extreme appears at many cases as a cure of an unpromising process. XP includes a rich bundle of techniques. It’s very technical compared with crystal. I think the essence of XP is to leave all the compromises for uncertainties at the planning game stage, and be extreme in the implementation, with an extensive set of feedback mechanisms to ensure the healthy circulation between the two. The many techniques of XP, like peering programming, backlog, planning game and test driven could be an introductory course for teams on the way to agile. I like Alistair’s idea of classifying developers into level 1, 2, 3. For level 3 developers you don’t need to tell him anything specific. A general description will be enough, and it’s best to let them create their own strategy along the way, which is the very essence of agile. But for beginners, the family of agile practices seems to be not that easy to master. XP provides a comprehensive way for the beginners to start with in agile path.

Minus

Being extreme compromises flexibility, so I think XP’s application potential is some what limited by its extreme standard. However, XP developers do encourage partially applying XP in a project. Also it takes a lot commitment and courage for developers, especially none-agile developers to try XP practices.

Context Driven Testing:

Plus

The context driven testing is an agile revolution for testers, not developers. But I think it’s also beneficial for testers to learn something about it, so they can improve their test driven development and better their relationship with testersJ. Context driven testing rethinks testing as an evolving complex process, which has to be dealt according to the context of it. It avoids following procedures, focuses on solving problems. It encourages a positive relationship between testers and developers. For developers I think their test for their own software is kind of context driven in the first place, for not many of them do the testing often, and even few of them will stick to ceremonial “best” testing methodologies. But still developers can utilize some techniques here to gain understanding of others’ system, which is important in agile practices, and try to treat testers as consultants, or promote this relationship. Also, I think it gives us a clear idea that acceptance test or functional test in some agile practices is an evolving process just like coding, where good enough solution is also needed.

Minus

Context driven testing requires the testers to figure out what to do based on the context. This not only requires constant thinking, but also a lot of experience, for you have to know the assumptions and consequences of your options.

Scrum:

Plus

Scrum seems like to be the opposite of CMMI. Yet it proves to be a successful management style in projects with a high degree of uncertainty, like most software projects. It maximizes communication, collaboration in the team, and minimizes distraction.

Minus

Scrum encourages “all at once”, which limits the scope of the projects and the number of team members as well. Also in interface clean projects, like OO development, we can partition the projects to smaller pieces and have a separate scrum team for every piece of work. It’s not going to work if you have people who is not committed or not experienced on the team. But when everyone is highly motivated and committed, they may keep pushing some real limits since in scrum team the management tends to impose as little restrain on team members as possible.

10.31

Filed under: Uncategorized

setting up the wiki for group. spent a lot of time hacking with httpd.conf file.

writing software engineering report… 

Get free blog up and running in minutes with Blogsome
Theme designed by Jay of onefinejay.com