Re: .NET Framwork and VB
- From: "Simon Hart" <srhartone@xxxxxxxxx>
- Date: Sat, 1 Jul 2006 15:46:06 +0100
Hi,
There is no difference between managed (.NET) and unmanaged (native) object
model design principles. Of course one would argue the main difference is
the fact that .NET code is intepreted, so the first time a method is called,
it is compiled by the CLR, where as unmanaged code (native) the dll is
loaded into memory.
Generally what you are doing in terms of distributed design methodology is
good, separating the system into tiers for good scalability and
maintanaility. Not quite sure why you divide the database operations into
separate classes though, but what we normally do is have a generic SQL
handler that parses SQL instructions (normally stored procedures) from the
business logic tier. The important thing here is that the data tier is
driven from the business layer.
Regards
Simon.
"Prateek" <prateekbaxi@xxxxxxxxx> wrote in message
news:1151674783.139398.118850@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Hi,
I have recently started coding and designing the ,.NET based system
using different technology involved in .NET and well from the subject
line of this post, it is obvious that I am coming from a VB background.
I have many years of expirance in designing distributed system based on
VB, C++ and SQL Server but when I have tried .NET it is a complete new
exerince for me and I am not hesitant in saying this is a damm cool
stuff from MS.
Now in this regards, I got a question but I am really not sure what is
the case in .NET.
In earlier days, when I used to design classes based on limited VB
object based model, we used to create 2 different callses for one
business component. One class which is used for all database read
operation and another for database write, update and delete opeation.
Now, why we were doing this so because of the locking in database via
VB programs. if all database select operation id seperated from
database write operation it palces less locks and almost comes out
immediatly by removing database locks (Offcorse when all the VB dlls
are regirsted in COM+). Now I wanted know is this still applicable in
.NET.
For example, if I want to write a Cusomter business component, do I
still have to seperate out the select and updates in seperate classes.
In earlier days, I used to designed a dll with following classes.
CustomerRead (Main class to read, exposed to world)
CustomerWrite (Main class to write, exposed to world)
CustomerSingleSelect - a calss which gets data from customer based on
primary key and always return me a sigle row
CustomerMultiSelect - a class which gets the data from customer where
select operation is not based on its primay key.
CustomerModify - a class to modify customer.
Belive me this approch was much faster then everything in one class.
But this was the case in VB time. Does this still a case in .NET?
Thanks & Regards,
pb
.
- Prev by Date: Re: Removing shell from startup and terminating it
- Next by Date: Re: Intercept a phone call and have custom module run instead of WM5 i
- Previous by thread: Re: images with title in listview
- Next by thread: Re: Intercept a phone call and have custom module run instead of WM5 i
- Index(es):
Relevant Pages
|
