Pann's Thoughts

I think I might possibly, perchance, accidentally, unwittingly have been missing something the last couple of years.

I have generally struggled to bring a rich domain model into play when building web applications. The main reason for this is that a web application is for all practical purposes stateless. The problem I have (had?) is that I do not want to bring the whole "rich" domain model to life just for one "web" call. I.e. I do not want to load the whole model (I'm not talking about lazy loading concepts here) attach listeners (on to the model) where neccesary and then detach the listeners and "remove" the model just for one call. Also to me a rich domain (OO) model implies state, else I'm actually just back to normal procedural programming but with the benefit of "grouping".

So, to date, I've used the standard web patterns and tools e.g. front controller, session facade (and in the java world the standard BOs  and DAOs) with hibernate etc. However my domain model (or entities) were pretty dumb (and persisted with (n)Hibernate) and would probably have been the poster child for the anemic domain model. (PS. I still think the anemic domain model has its place)

Well, today I again paged through my Domain Driven Design book (Eric Evans) and skimmed through repositories (since I've been using it WAAAY before the book was even a gleam in Eric's eye) and something struck me. What if, instead of creating a new repository everytime (like it seems a lot of people do) I keep it around.

What if the repository was implemented with the aid of that other (bad child) pattern named  Singleton. It would mean that my Repository would be in memory for the duration of the web application, and not just the one web call. All the CRUD operation can then take place through the Repository and since the Repistory is not destroyed everytime, it means I can even attach event handlers etc to my objects.

Admittadly it's not as easy as it sounds. One of the possible problems is that you'd end up with the whole database in memory and that is definately a BAD idea. Still, I think I'll persue this line of thought further and see where it leads me.

If you have any thoughts or experience in this regard, I'm eager to hear from you.

I'm not a fan of averages. Truth be told I think averages are evil.

I normally go into an "average" rant whenever a new developer starts doing performance testing. Said developer will then give me "average" times, thinking it indicatesa program/process' performance. When I explain to said developer why averages (or they way he calculated) are useless, he (normally they) still don't believe me. (Most of the times though it's probably because their ego can't handle it)

Recently I experienced it again. (Although, to this developers credit, he understood the problem as soon as I explained it.)

The problem again had to do with performance testing. In our application, we need to do calculations on various products and, understandably, the various calculations were different. The developer then came back with the following results (timings given are ficticious but indicative)

Prod1: 0.01 ms

Prod2: 0.01 ms

Prod3: 0.05 ms

Prod4: 0.70 ms

Average time 0.19 ms

Now this average looks acceptable (as a response time), except that most of our users will actually choose Prod4. I.e. their "average" is actually 0.7 ms and NOT 0.17 ms. Almost half a second difference.

So, what am I trying to get accross. Make sure your average is really the average you think it is. ;)

Greets

Roaan

A new IBM/Lenovo Thinkpad Z61p

- 2 Gb RAM

- 100 Gb HD (7200 RPM)

- Core 2 Duo

- 15.4" Screen

- ThinkVantage software suite

- and.........keyboard drainage holes

What more would one want?

This is mainly to see if the photo plugin works.

The picture to the left is a image of windows live writer

So found the link via dotnet.org.

Now I'm also gonna post a code example.

Does ne1 know of a way to post something from live writer without it displaying on the main dotnet.org feed ?

I want to derive some control of the base .net WinForm controls (e.g. TextBox) in order to set the look&feel of the control once, and then reuse this derived control throughout my application. This way, if I want to change, lets say, the colour, I only need to go to the derived control, change the colour and all the textboxes should have the new colour. (Unless the forms themselves specify a different colour)

That's the theory.

In practice I'm struggling to make this work (in VS2005).

If I drag this control onto the form once, everything works fine and according to theory.

However, as soon as I drag the control onto the form a second time, somehow the form detects the color of the control and hardcodes it into the form designer code. The effect of this is that if I then go and change the derived control's colour, it is not reflected in the form when I run the application.

A workaround is to use composition instead of inheritance but that is more work.

Can anyone shed some light on the subject?

I want to derive some control of the base .net WinForm controls (e.g. TextBox) in order to set the look&feel of the control once, and then reuse this derived control throughout my application. This way, if I want to change, lets say, the colour, I only need to go to the derived control, change the colour and all the textboxes should have the new colour. (Unless the forms themselves specify a different colour)

That's the theory.

In practice I'm struggling to make this work (in VS2005).

If I drag this control onto the form once, everything works fine and according to theory.

However, as soon as I drag the control onto the form a second time, somehow the form detects the color of the control and hardcodes it into the form designer code. The effect of this is that if I then go and change the derived control's colour, it is not reflected in the form when I run the application.

A workaround is to use composition instead of inheritance but that is more work.

Can anyone shed some light on the subject?

Does anyone know how to use the new (VS2005) MSDN ?

Or more specifically how to search for content that contain two or more words (not phrases).

I tried doing a search by specifying the two words, but then it does an "OR" search.

Can anyone help on how to do an "AND" search ?

PS. Can anyone point me to info on setting up Infragistics UltraTree with recursive custom object (i.e. not datatables)

A collegue of mine introduced me to Charles. It's an HTTP proxy that intercepts your web requests. Using this, you can inspect the requests a web site makes as well as the responses it receives. Not only does it show you the content of the HTTP requests/responses, but also the time the request/response took as well as the size of the response. But that's not all....if you phone now...(I digress) it also has throttling capability built in. This allows you to emulate a modem or various other connection lines.

We found it very usefull to determine the usability of our website and using it was definately an eye opener.

Very Cool :D

When casting (in C#) you generally have two options.

Typicall, when I’m going to cast only one once and want to call a method on the casted object, I use

((Foo)myObj).Bar()

The other day, I saw someone write

(myObj as Foo).Bar()

Which do you use (Vote here)

A small list of opensource windows programs can be found here.

(At least the icons look nice)

Let’s say I have a factory method that creates an instance of an object (e.g. I wrap Activator.CreateInstance).

I have 2 options

1) public object Create (string className) {return Activator.Create…….;}

2) public T Create<T> (string className) {return (T) Activator.Create……;}

The client code will then be one of

1) MyObj obj= (MyObj)Create (“myObjKey”);

2) MyObj obj= Create<MyObj> (“myObjKey”);

Which would you prefer ?

In VS.Net 2003 you could close the current open document tab by right clicking on it and selecting “Close”.

VS.Net 2005 takes this one step further and allows you to select “Close all but this”.

So my question. Why did they not add a “Close All” option? Quite often I’ve got a number of documents opened that is related to one “idea” (e.g. task or whatever) and when I want to start working on the next “idea” I want to close all the open documents and start opening the documents I need for the new “idea”.

Now I need 2 clicks. One to “Close all but this” and another to “Close”.

If I’m missing something I’d be more than happy to learn a new trick

At a client site today, we received new PCs and today I had to setup my screen (19” Dell). This proved extremely difficult, and as yet I do not have it setup properly yet. Why do screen manufacturers make it so difficult to setup a 19” (and even 17”) monitor so that the whole display area can be used. The problem I have currently is that my horizontal borders (top and bottom) are not parallel. One would expect the monitor setup menu to have an option for adjusting this… One would expect wrong…

 I can change almost anything, including the ability to adjust the vertical borders (i.e. left and right) to make them parallel but I can’t find a way to adjust the horizontal borders. The best of all is that I didn’t need and don’t want to adjust the vertical borders since they are displayed properly.

What gives?

PS. Thanks Armand for the upgrade to CS 2. It looks great! It’s really appreciated that somehow you sqeeze in time between development and comrades  to keep the site running as smooth as it does.