Microlau's Blog

Oct 2007 #41 UG Meeting

This month, VSTS MVP Anthony Borton will be coming by from Australia to share tips on the Best Practices for Team Based Software Development. If you’re interested or intend to find out how VSTS can work for your project team, be sure to mark your calendars!

On the lighter side, Microsoft DE Chewy Chong will be meddling with phidgets and embedded devices. You can be guaranteed to catch a few glimpse of interesting gadgets.

Our UG meeting will be held on 4 October 2007 at Level 22, CF12 from 6.45pm – 9.00pm.

Best Practices for Team Based Software Development
Anthony Burton (MVP, VSTS)
A highly cohesive software development team is critical to a successful project. In this session, we examine all of the different ways teams members, from project managers to developers and QA, can work together to help enhance team collaboration, and ultimately process success. We then take a look at how these opportunities map into Microsoft Visual Studio Team System and other Microsoft products to show you how you can take advantage of these best practices on your teams today.

About the Speaker
Anthony Borton is a Microsoft MVP in Team System and has been working with VSTS/TFS since its early beta release. He has delivered much of the early VSTS training around Australia for the Microsoft Technical Readiness Team and is the convenor of the QLD VSTS Users Group.

Fun with Phidgets and Embedded Devices
Chewy Chong (Microsoft DE)
To be advised!

To register for the event, sign up here

XLS hell

Give a lucky guess for the most sought after application after VS.NET at my current project? Before I reveal the answer (which I suspect many folks can easily guess), here's the scenarios which requires us to use Software X.

1. Let's do up a leave forecast plan. Great idea! Let's use Software X.
2. How should we track our defects or fixes? Easy, let's use Software X. Import from Software Y and put the data into Software X. Drop the file into a network folder and all parties involved can read the file.
3. Let's have a inventory of integration points. Yeah, Software X just fits the bill!
4. ... ... ...
5. And the list goes on!

If you haven't guessed it by now, it's Excel - the top-down, one-size-fits-all software.

12 Sept : Lunchtime Talk with Steven Guggenheimer

12 Sept : Lunchtime Talk with Steven Guggenheimer, General Manager, Application Platform & Development Marketing Division, Microsoft

Steven Guggenheimer, a 12-year Microsoft veteran and GM in the Server and Tools Business (STB), leads the Application Platform and Development marketing organization. Guggenheimer focuses on Microsoft’s overall Application Platform strategy working in conjunction with the development, sales, and evangelism teams to drive more effective relationships with customer and partner communities. He is responsible for leading Microsoft’s integrated value proposition around the application platform business across all key constituents and specifically tying in the SQL Server, Visual Studio and BizTalk product lines.

Steven will speak about the current and future directions of SQL Server, Visual Studio and Biztalk Server. You should invite your CIO or CTO to attend this with you, if possible. If you have a specific question or topic you want Steven to address on that day – please include the question in your registration. Only questions submitted before 7th September will be forwarded to Steven.

Date:  12 Sep 2007 (WED) Lunch
Time: 12.30 – 2pm SHARP! ( Registration and Lunch from 12noon onwards).
Location: 22nd Floor, Microsoft Singapore, One Marina Boulevard.

To register, Pls email the following details to kcdewimaya@hotmail.com

Name:
Designation:
Company
Tel No:
Email :
Event : 12th Sep Lunch Chat with Steven Guggenheimer
Question for Steven (if any):

Passing tamper proof querystring parameters

Software Estimation

How often have you been asked to come up with estimates for a task package, change request or even bug fixes? I'm sure most of us in software delivery deal with estimates on a daily basis and we're probably used to coming up with estimates which a flick of our fingers. Statistics have shown that developers typically estimate 20% to 30% lower than their actual effort (van Genuchten 1991 "Why is software late? An Empirical Study of Reasons for Delay in Software Development"). It's truly a challenging art of mastering a proper estimate especially when you have little or zero knowledge about the task package, change request or business domain. Add to that, the pressure of giving narrow estimates or shorter time required, estimates tend to even err more. 

How to combat poor estimates? There are already a bunch of common techniques used by various IT folks, the most commonly being adding an extra 20% - 30% to whatever timeline/budget/manpower was estimated.

Is that truly sufficient? I believe that we should be given a chance to reflect on the whatever we do and this reflection exercise should be carried out meticulously and continuously. If you've grossly underestimated a build item, think about what lead you to the estimate in the first place? Learn from it and most importantly, apply it the next round. 

How do you and your organisation do your estimates? 

Raincheck for SGDN Sep UG Event