Comments by Klaus Luedenscheidt (Top 49 by date)

Reason for my vote of 3 \n Nothing new and awesome. And the article shows only your personal poreferences.

The built in refactoring possibilities are very poor. You should use an addin like ReSharper if you frequently refactor your code.

Reason for my vote of 3 \n Nice point-and-click introduction, but i'm missing a sample for how to use the splash screen. Best will be to start the splash screen asynchronous to be able to init the application while the splash screen is showing

Reason for my vote of 3 \n i'm not sure if your tip is really useful. The whole purpose of a MDI form is to group similar forms together. If you want to display a form only once a time you can also display it as independent non-modal form.

Reason for my vote of 1 \n this

Reason for my vote of 5 \n keep it simple :-)))

Reason for my vote of 3 \n Your approach is very limited. First you don't be aware of localization issues (even in your last approach, which is limited to the current systems culture). Second the use of exceptions to handle input errors should be avoided because throwing and catching exceptions is very time consuming. Third it is always better to check the user input after he has finished his input. Checking each keystroke is time consuming and also the user experience is bad (you are typing something and nothing happens). Fourth if you really want a limited input a whole bunch of specialized user controls exists in the wild which are additionally capabale of handling different cultures and formatting of input. And a lot of them are free.

Regards
Klaus

Reason for my vote of 5 \n very useful tip...

Reason for my vote of 3 \n Simple and straight forward solution. But i think you can do much work with the forms designer. There is no need to completely create the form in code. Also you have duplicated the MessageBox enumerations. Why don't you use the existing ones?

Your naming conventions are a bit unusual. Main stream is to use CamelCase for class and enumeration names.

Do you know that there is already a MsgBox class in the VisualBasic.Compatibility namespace? You should avoud to use class names which are already definet in the .NET framework classes.

Reason for my vote of 4 \n you should provide a short example demonstrating the use of the API calls in practice

Reason for my vote of 2 \n Your article is too simple even for beginners. It shows not much more than you can read in the MSDN documentation. For this it is not very helpful.

If you use screenshots, you should pay attention that they are readble. The screenshots in par 2 are definetly too small to read without zooming.

At least you should embed code as text block in your article, Code Project has the capability for this as you can see in almost all articles.

Reason for my vote of 3 \n Seems to be a nice tool. But i am missing some implementation details (mainly the things you noted in the "Points of interest" section)

Reason for my vote of 5 \n The only disadvantage i see is that you are database dependent (you can't change the underlying database easily). But this, in my opinion, is academic becaus ein realit yyou don't change it often.

Reason for my vote of 5 \n That's why i always use braces ;-))

Reason for my vote of 5 \n I totally agree with you. I'm currently maintaining a huge legacy application and follow similar rules. This ensures my survival in the code jungle...

Reason for my vote of 4 \n Nice and easy to use icon editor

Reason for my vote of 3 \n your tip is quite good, but the title is misleading as you write about properties. At the end you write that one can collect the properties in an ObservableCollection<object>. First this is not optimal as it lacks type safety. Second i'm missing a sample which uses the ObservableCollection. Because in this case the binding is more complex it would be fine to add a XAML sample.

Regards
Klaus

Reason for my vote of 4 \n simple and straight forward

Reason for my vote of 2 \n The title of your article is "Understanding N-Tier

Reason for my vote of 3 \n You just write on the SOLID principles. The information you provide one can find on many places and even better described. If you provide some examples for the principles your article would be more helpful.

Reason for my vote of 2 \n Your explanation is too simple. If you will see a really good explanation of the Liskov Substitution Princple take a look at this article:

http://www.codeproject.com/Articles/597870/Liskov-Substitution

Reason for my vote of 1 \n this is waste

Reason for my vote of 3 \n This is basic knowledge for every SQL programmer, even absolute beginners.

Normally i would give you a 1 for it because such type of articles are messing Code project in the last time, but you tried hard to explain the difference and your example is quite good. Zherefore i could vote a 3.

Regards
Klaus

Reason for my vote of 2 \n i don't see a really sense for this. If i want to kill a process i use Process Explorer. If i have a virus or malware i either use a anti virus software to remove it or reinstall my PC.

On Windows 7 you may need to elevate your program to act as adnmin because otherwise you can't kill every process.

Reason for my vote of 4 \n Nice article. But one thing i don't understand. You first use a standard string to define the password and then copy it to the secure string. So the problem yiou described in your introduction isn't really solved. Because strings are immutable even if you clear the unsecure string the old value will remain in memory until the garbage collector removes it. If there is no other way to use secure strings than you describe the implementationm of MS is poor for me.

Reason for my vote of 2 \n My stomach says to me that the triple casting is a bad idea. I would prefer to write a specific method overload for each primary type in this case. This also prevents boxing and unboxing which may be a perfomance issue.

Reason for my vote of 5 \n simple and efficient

Reason for my vote of 4 \n Good introduction. You should check if the loglevel is enabled. This speeds up the logging. Here is a sample from my wrapper:

public void Debug(string pMsg)
{
if(null == mLog)
{
throw new Log4NetConfigurationException("Logger.InvalidConfiguration");
}
if(mLog.IsDebugEnabled)
{
mLog.Debug(pMsg);
}
}

Reason for my vote of 3 \n The tip is not so useful because it only works if you know the format of the date string you process.

On the other hand it shows that you always should try to avoid to work with date strings.

Reason for my vote of 2
You should take a look first, if similar tips where posted here before. See Vijay's link samples

Reason for my vote of 4
Good tip. Alternateively you can implement this as an extension method of EntityObject. But i wonder if this should work the the Entity Frameworks Code First approach because there are not necsessarily the attributes you query.

Reason for my vote of 3
See my comment on your other articles

Reason for my vote of 3
See my comment on your other articles.

Reason for my vote of 3
Why do you wrote 3 articles? All of them are on similar problems, so why the hell don't you handle them in one article?

Reason for my vote of 2
Never ever do this so. This is old fashioned pre VB.NET style and should be avoided. As emartinho states this style prevents clear readable code. I don't know why Microsoft continues supporting this.

Reason for my vote of 3
After taking a look at the source you made your own implementation of the standard events in .NET. It's a good example to show unexperienced users how events works in .NET but in an application i would prewfer to use the standard mechanism.

Regards
Klaus

Reason for my vote of 2
i totally agree with Markus

Reason for my vote of 1
This article is not more than a very simple quickstart with no additional advantages against the information you can get dircetly from Oracle. If you go to the OTN and look for information for example this article:

http://www.oracle.com/technetwork/articles/dotnet/vs2010-oracle-dev-410461.html

will provide much deeper information .

Reason for my vote of 3
Hope you will add more samples over time.

Think about your sample about "avoiding double calls". For me in a simple if / else it's absolutely legal to make the call the factory in both parts, because it's definetly called only once. If i have a more complex method where i need to make different calls to the factory object i would store an instance in a variable.

Reason for my vote of 2
Nice but useless. If you install the Productivity Power Tools you get much more functionality

Reason for my vote of 4
Nice tool, alternatively you can use the build in Firefox Sync to keep your bookmarks up to date on several machines.