Choices
-
The big thing with .NET is the framework of 7,000+ functions.
There are a lot of things that .NET does with very little effort that required an good knowledge of the Windows API before.
If you look behind the scenes .NET apps make reference to MFC & as opposed to the previous incarnation of Visual Studio which references MFC4.2.
A lot of thought has gone into .NET and it does represent a paradigm shift for existing VB programmers.
I think of VB6 vs .NET as the difference between a car that is a 10 year old design and a brand new model. In the UK we had the last model Ford Escort which was automotive gruel vs its replacement, the Ford Focus which is light years ahead. Yes it costs you to replace the car and ultimately the new one is still going to get you from A to B but it is one of those costs you just have to bear.
-
To continue the analogy, over the 10 years you had the Ford Escort you traveled 300,000 miles/482803.2 kilometers...picking up untold tons of groceries, delivering thousands of suits for drycleaning and then picking them up, perhaps taking the missus to the hospital to delivery a baby or two every couple of years, maybe survived one or two traffic mishaps perhaps requiring some time in the hospital, went to half a dozen funerals, took the kiddies to hundreds of soccer games...
And then your new Ford Focus comes, and you have to do all of the above all over again.
Sometimes being able to do something "more elegantly" isn't the end game.
(By the way, I've used I have no clue how many direct calls into the Windows API over the years, but two things stick out: First, I NEVER used 7,000+ of them in the same program. Secondly, once I got an API call working to my satisfaction, I saved them off in a module for reuse without rewriting them ever again....)
-
I fully understand your point but from my limited experience with .NET the stories of M$ investing more in .NET than NASA did in the moon shot are entirely believable.
It is very obvious that a massive amount of thought and work has gone into the framework and also the languages that plug into that framework.
There is so much in ASP.NET that it beggars belief.
I cannot speak for traditional Windows development but certainly from the web development side of things .NET is a quantum leap forward.
Put it this way, the Open Source community are normally rabidly anti-M$ and yet, through the mono project they are in effect, endorsing an M$ development!
I never really understood why people were so hostile to the original VB as it is end results that matter but I can see that VB.NET is a major leap forward.
As an aside I was never afflicted with a Ford Escort. I had/have a Mitsubishi which my friends are keen to point out has an anagram of Shiti bumsi
-
I've been programming since 1968. Programmed in most of the popular languages of the day.
About 12 years ago three of wrote an app in VB3. It's been upgraded to VB4 and finally to VB6. In 2001 we added XMLHTTP from VB6 to access a centralized SQL Server database. We avoided the move to HTML/Javascript. We release two-three feature/bug releases per year.
The application has/is/and will provide a business need.
Why do we need to change to anything.NET? Do you think our clients will appreciate the rich opp technology?
The purpose of software is to fullfil a business need. A dramatic change in technology should be the catalyst for change.
Moving from IBM Selectric typewriters to PC based wordprocessors qualifies. Moving from journal ledgers to computers qualifies.
Changing because it's the geeky thing to do only makes Bill wealthier.
VB6 till I drop!
-
At the end of the day if the app fulfills the business need then it is mission accomplished whether it was written in VB3, 4, 5, 6 or JFDI.NET.
Yes migrating up through the versions is an expense but the points to consider, as I see it, are:
- Staff turnover - can you recruit easily for a (perceieved) obsolete skill
- OS Obsolesence - If your compiler doesn't work when you upgrad e OS then what do you do?
- Hardware obsolesence - I have had to ditch some useful utilities because they don't run anymore.
- Support obsolesence - A recent employer supported a Site Server 3.0 installation that required additional bespoke functionality.
Try finding Site Server resources on the web.
- Longevity - Is the proposed new environment likely to last or is it a "fashion" item
-
This is obsolete by M$:
- Longevity - Is the proposed new environment likely to last or is it a "fashion" item
Everything M$ is currently working on is proposed at a 2 year life cycle. This makes everything a fashion item. Even within the .Net IDE. Anyone upgrading from VS2003 to 2005, (as the solution and project files are not compatible), or .Net1.x to .2, (as M$ changed the name of common Objects like ConfigurationManager to ConfigurationSettings).
OOPs is a good term. It has made MCDonalds workers out of a profession. We wrote better code in assembly than we do in any OOPs language. Easier to create/write and RAD environments makes it faster to a mistake, not faster to get a quality product implemented. The key to a quality implementation is not in any language or development environment.
[font="Arial"]Clifton G. Collins III[/font]
-
I totally agree with Clifton. M$ is moving at too fast of a pace to produce quality work. They don't seem to be improving the results, only the product line cash flow.
Look at Office 2007 - why would they totally redesign it, if not to drive their sales? Especially since it does not use .net technology, so you need interop wrappers and such nonsense.
As far as David's mention of "migration" if you build a quality product that does what it is supposed to do, and continues to work without breaking, you don't need to migrate anything. I have a Fortune 500 customer that uses Oracle, SAP, and lot of other modern stuff but they also have a small dBase III-Plus system to process invoices - it worked when they moved to NT, still works under XP, so they're not changing a thing. I drive an 11-year-old Subaru with 108,000 miles and I'm not migrating that, either 😉
-
There is a light bulb in the garage at my parents' home that is AT LEAST 60 years old. (Big old clear thing - you can see ripples in the glass as if it were hand blown.) Amazingly, you flip the switch in the dark, and suddenly you can see.
The point is, how do you keep the vulture capitalists of the stock market and your board happy, if you make something that "just works" and so never needs replaced? Especially if those vulture capitalists had the foresight to ensure that the practice of compensating high level executives with stocks (and/or) options is pervasive throughout American business?
There is an awful lot of incentive to create "new programming paradigms" to move new product, just as there is to create new operating systems to move new applications and computer hardware.
(By the way...there was a duplicate of that light bulb in my parent's basement. I was screwing around one night 40 years ago and accidentally broke it. As more and more years go by and that bulb in the garage just refuses to quit working, I feel guiltier and guilter...sigh...
)
Viewing 8 posts - 16 through 22 (of 22 total)
You must be logged in to reply to this topic. Login to reply