There is a cost in adopting new ‘mind-set change’ pattern

I am not the old school guy that do not like in new technologies, as well as in adopting new technologies /techniques /patterns – then I call it Patterns in this article – in software development. But sometimes people don’t realize what are the impacts on their projects, especially when they try to adopt it – because of its better performance promises – in to their running projects.

If we draw back from couple years ago up-till now, some samples of mind-set change caused by implementing new technologies, techniques or pattern we can see as below:

-OOP (mind-set change from Procedural programming model)
-Server Side Scripting (mind-set change from static to dynamic to web app)
-Web Class (mind-set change from plain server side script to compiled components)
-Multi Language Programming (mind-set change from single language programming)
-Compiled Web app (mind-set change from script interpreter process)
-WebForm adopting WinForm /desktop development model (mind-set change from embedded script in html to drag-drop desktop like web form designer)
-MVC with various implementation (mind-set change from code behind to controllable views)
-AJAX with various techniques (mind-set change from post back)
-XAML (mind-set change from designer generated code to xml extension)
-MVVM with various techniques (mind-set change from tight couple UI to separation of concern and easier to test UI)
-Multi Touch UI (mind-set change from single touch /press input)
-etc.

Adopting the Patterns that will causing mind-set change during running project took cost in various amount of time, but no doubt because it needs time to learn, do some researches, test in current project while still maintaining current Pattern that need to be replaced.
If you have managing projects running in more than 1 year, it is likely you will face new Patterns coming, as efficiency of the tools to solve known problems.

Some of Patterns do not have significant affect in mind-set change, therefore it doesn’t matter or at least have absorb-able cost because of its small impact in the project. Some of them for example:
-Enterprise Library (from using self made utilities and DataAccess layer)
-New All-in-one Library compiled (JQuery or other Toolkits from using scattered client side javascript files)
-LinQ (from implementing DB query base into middle layer)
-WCF (from implementing WebServices)
-SQL Azure (from using local SQL Server)
-etc.

Unless we have planned to adopt these Patterns in ‘start from scratch’ projects – which would be predictable in cost and time frame – since design phase, we have to be carefully and wise in decide new Patterns to adopt.
But thankful to these Patterns and all the ideas creator as we are moving fast to be more efficient in creating softwares.

Error on The service System.Windows.Forms.Design.IEventHandlerService already exists in the service container.

When you have an inherited .NET WinForm from its base Form, some times you’ve got this error message when trying to open the Form in designer view, then the designer view doesn’t want to load.

The service System.Windows.Forms.Design.IEventHandlerService already exists in the service container. Parameter name: serviceType

What happened is when the designer view try to load the Form, its actually instantiate the base Form, and then linked it to the Designer through event handler service registration, then it will trigger the Load event, and also other event caused by visual change like VisibilityChange, etc.
The error is thrown by the codes that been triggered in that event handlers block that failed to be executed, it might be the codes for initialization, load setting, handle control changes, or anything. And when the codes which you have put in the event handlers have error or failed to execute, its causing the Visual Studio could not load the designer.

To solve it, either you have to fix the codes, or you just need to put this line on the very top of your designer view triggered event handlers.

if(DesignMode) return;

I hope this will help.

Its not all about building the team and use a modern methods

When I’ve been questioned about how do you think on building the productive team for software development. I said I always dreaming to have a small and solid team, where there is a similarity of level, no hierarchical in rules and positions so the chemistry works well within their body language; skills are complementing, open discussions are conducted when needed, accountable performance between them are transparent. Off course this team is not easy to build. Not so easy. But possible. In fact I do have one, once.

And when the next question is what methods you will use then, I said there is no perfect method to adopt as is but you will find a perfect method to play with if you let your team form it as they know exactly what they want. I would say yes this would be base or mostly inspired by modern agile methods but even if you are not literary refer to them, you will deserve that it refer to the similar aspects.

This kind of team, will be very solid for all projects they are working on, because they are playing with their heart, knows strong characters in a good chemistry.

Posted from WordPress for Windows Phone

WebBrowser control and Crystal Report problems on Windows 64 bit

I have an application that using WebBrowser control from System.Windows.Form for some reasons. My desktop client application has an embedded WebBrowser control to access my ASP.NET application from the web server. It will access its reports and other pages using Crystal Report, Silverlight controls. The client also take all benefits of windows form controls drag-drops, drawing designer controls and also printing.
Below are the problems related when I run my application on either Windows 2008 Server 64-bit and also Windows 7 64-bit.

Symptom:

1. When I run my apps on 32-bit Windows it have no problem at all. But when I run it on Windows 64-bit, all plug-ins have problems and crash without error shown.
2. I have more problems on the ActiveX Crystal Report Printer Control which unable to do automatic install even if I use the 32-bit version of Internet Explorer.
3. More problem also in the Crystal Report page, all Crystal Report toolbar icons are missing.

Causes:

1. The WebBrowser control will pickup 64-bit IE automatically by the process. It is indeed because the application has been built with target of Any CPU. The current version of Internet Explorer 64-bit (8.0.7600.16385 64-bit) has some problems on performing plug-ins: Silverlight, ActiveX (Crystal Reports Printer Control) and Flash. As the 64-bit IE browser have issues on the plug-ins, it crashed while its trying to print the Crystal Report pages using ActiveX Print mode and also when trying to load my Silverlight control contained pages.
2. When I check on Downloaded Program folder in Windows folder, there is no ActiveX installed, seems the IE doesn’t want to download the cab and install the ActiveX automatically.
3. When I check on Framework64 folder, there is no ASP.NETClientFiles folder as it is required by the Crystal Report to redirect its icons requests.

Solution:

1. There is no way to switch and force the browser from 64-bit running app to load 32-bit browser. It is by design.
   To force the process to automatically pickup 32-bit version of IE by the WebBrowser control, I have to build my application to specific x86 target instead of Any CPU, so it will force to pickup the 32-bit version of IE.
2. I need to manually register the PrinterControl.dll from Crystal Report to be able to run the ActiveX printing on Crystal Report.
3. I have copied the ASP.NETClientFiles folder from C:\Windows\Microsoft.NET\Framework\v2.0.50727\  to C:\Windows\Microsoft.NET\Framework64\v2.0.50727\ . This is because I need my ASP.NET application keep running on 64-bit so it requires the Crystal Report icon files from Framework64 folder.

These are what I have done to make my desktop client application with WebBrowser control embedded running all plug-ins correctly on Windows 64-bit.

I hope this will help anyone encountering the same issues. Or any other different tips to solve the same problems would be wonderful if anyone can share it.

Changing and adding behaviors on DevExpress Grid control

I have searched to most of the licensed Silverlight based grid controls, but no one meet our requirements. Even the most expensive one. Our customers request to the grid control behavior a bit unique to the standard grid control in the market. Especially on how the group behave, eg. Group header summary, mouse over the group header, color code changes, keep expand the group while navigating, etc. Its because we need to apply it into rich UI and need fast interaction to the control. Our customer’s users are not geek or office admin types, but very busy operators like in the stock floor. Therefore they need to be able to operates the grids quick and fast with a lot of listed data and be able to grouping-tabbing from different category. They also need to have a fast edit.

Then I found the free grid control from DevExpress AgDataGrid 8.2.8. Fortunately this free royalty version come with the full source code. And the looks and feels are pretty good. So i play a bit to enhance this control to meet the requirements.

By mostly by inherited the original classes from AgDataGrid (Grid, Row) and doing changes on the xaml looks and feel I finally came up with sort of result as below.

AgDatagrid grid group summary enhancements

AgDataGrid controls in group header

AgDataGrid group auto collapse expands

Actual Error Message can’t displayed in .NET CF deployment

When you are developing SmartDevices apps, and have experiencing error message like “an error message cannot be displayed because an optional resource assembly containing it cannot be found..”, possibly because you are intend to run the apps and assume it will work – because you did not found the error when run in IDE /emulator –  but sort of conditionals causing the errors which you didn’t predict before.
This generic error displayed because we did not deploy the completed component /assemblies into device. This is good as long as the assemblies not used, but in this case you need to know what happened.
The actual error should comes up after you copy the System_SR_[Lang].CAB file and install it into the device. The cab file you can get from C:\Program Files\Microsoft.NET\SDK\CompactFramework\[Version]\WindowsCE\Diagnostics\ and there you can choose what Language you want to read from the actual error message.

You can uninstall the cab later if you already confidence the app will works properly using Control Panel – Remove Program and choose the installed cab.

Cheers.

Why you need Software Project Manager?

I have read a lot of discussions about the differences between Software Architect, Software Analyst, Software Project Manager, Software Product Manager, etc. and some of us I believe still have ambiguity about what exactly are they doing.
In this brief note, I just want to share a bit differences between them, and try to answer the question on the title.
In the medium – large software development organisation, there are these positions used in order to follow the good process, but it’s obviously depends on what methodology are adopted and also the management style.

Basically, what is the Software Architect does is responsible to design the whole system, integration between the systems, make decision on technology challenge, responsible for the risk on technical side. They have to know what the best suit to solve the domain problem that have been identified by the Software Analyst. They work with Project Manager, Software Analyst, Developer Leaders and Developers.

The Software Analyst is responsible to understand the business flow and domain problem, then create the business requirements which are their responsibility to make all parties understand especially the Developers. They also have a lot of discussions with the Software Architect in term to gain feasibility of the domain problem into logical software. This is then has been sub-positioned by System Analyst and Business Analyst that will make clear different responsibility area but still in the same communication channel. They will work with Software Architect, Project Manager and Customer.

Software Project Manager is the person who have responsibility on the project and scheduling management, resources, resources share load balance, cost, risk in term of project budget and delivery time. Standard management process are still in the main concern: Initiation, Planning, Executing, Monitoring and Controlling, Closing. All of these should compliance to the software development process and the methodology being adopted. Software Project Manager will work with Software Analyst and Customer to define the requirements. They will report to the Stakeholder in term of project budget, process improvement and customer relationship.

Software Product Manager is a bit not popular compare to the Software Project Manager. This is because Software Product Manager is more focus on the single product or some products that they will always responsible to. They have a bit wider in the responsibility areas. Their area responsibility are: Strategy, Concepts, Market Entry, Development, Evolution. And Of course the Development in term of Software Development it self contains mostly from the Software Project Manager area. The differences are because of Software Product Manager always working for their specific products, but Software Project Manager working for several projects. Software Project Manager do not even know what product they will work for the next upcoming prospects.

Now, to answer that questions. What in your opinion?
Mine is; If the organisation is quite small – Small-Medium Software development organisation – some of those positions are intended to be removed or merged. You can – normally they are also did – merge the Software Architect with Software Analyst into a single person, since this responsibility need similar skill and most of them have capability to do both. Then if this person also has capability to do most of Software Project Manager works, you can also merge three into one person, especially on the small organisation. Some and maybe more than half of Software Project Manager area can be done using easier the tools, and there are a lot of popular tools in the market. Some popular methodologies support those tools.
So, my first answer is: You need Software Project Manager because of your organization is quite large and difficult to get Software Architect /Software Analyst who has capability and time to manage the Projects.

In the most of ‘in house’ software development who only develop specific products, they do not need a Software Project Manager, because Software Product Manager will cover all the responsibilities, and also plus some areas that required to manage developing products. So, the second answer will be: You need Software Project Manager because you do not manage specific products but manage various and unpredicted projects to develop instead.

Hope this help to make more clear,

Cheers.

My another 100 projects

Today after me and my team have synchronizing and merging some projects – some from the prev development stage and some from the next development stage – our solution comes up with 101 projects
Finally we hit 100 number of project! cheers…

My Q1 2010 project schedule finally has finished

I have used the Microsoft Project to design my Q1 – 2010 projects schedule, because its simple and we can see the resources shared load and arrange the Gantt chart properly and see what is the impact in the time frame. This is commonly used by the project managers or someone who has related job.

Then I have exported it into the Axosoft OnTime – using Agile template, so I can run the projects, assigning tasks to the resources, adjust the tasks into Releases then everybody can see when the releases will be ready. Viewable measurements like Burndown chart, features tree map, pie chart count, and all could be consolidated into a Dashboard to easy control the running projects.

Ideally, I should be able to synchronize in two ways between Axosoft and Microsoft Project. I have been looking the solution for this, then I found the idea here, as David Higgins said:

To synchronize the two Projects (OnTime and the MS Project) you would have to store relative information in both about each-other, the most likely thing to store is the OnTime Item ID and the Project Task.ID properties — so when you add the MS Project Task to OnTime as an Item, simply update the MS Project Task and store the returned OnTime Item ID (TaskId, FeatureId, DefectId, IncidentId — depending on your use).  You can then also store the MS Project Task.ID Property in a Custom Field in OnTime — for Tasks, this is sort of simplified since there is a legacy property exposed by the SDK called “TaskNumber” which is simply a string that you can store an alphanumeric value in — this property is not exposed through the OnTime UI however (where a Custom Field would be) – David Higgins.

but didn’t have time to apply it. I will in a short time, maybe next month.

Cheers,

Have a good day

New blog about my activities in the work & hobbies

Hi, this is my new blog – created for this new year 2010 – about my activities in the work and hobbies. Well, actually my hobbies here means my horrible hobby related with my jobs…

I am a Software Engineer, working in some projects personally and also at my employer’s company.
This blog actually is use to continue my writing and sharing from my old blog at  here.
Hopefully been moved here, I can share thoughts and experiences more often and independently.

Knowledge is free, if you find something useful.. just re-share it to the others… if you find more useless, please apologize..
So, let’s start it…

Cheers…

Arif Budimartoyo,
Sydney, NSW