Why did Microsoft abandon ,killed VB6?

[Elroy]

I've just got one thing to say to all the people trashing VB6 programmers: If what you say is true, then why isn't there a strong VB5 or VB4 or even VBDOS community around, like there is for VB6?

Were we all willing to learn new things when VB6 came out, and then something happened to us and we were no longer willing to learn anything new? I don't think so. There may be nuances for each of us, but you can't escape that there's something that keeps this VB6 community alive, and to blame it on personality traits is just wrong (bordering on disgusting), IMO. You're pointing the blame arrow in the wrong direction, and for what reason, IDK. (I could speculate, but that would be counter-productive.)

[xiaoyao]

Before VB 6 came out, there were few computers in use in many countries.So there were very few forums at that time, and even if there was one, it could not last for 20 years.The forum of Vb6 may also disappear in 10 years.

Delphi 7, C + + builder, power builder, and many other competitive products at that time.

Sometimes they are killed by their competitors, and they lose badly. Sometimes your own company's decisions are wrong, or another development team takes your place.
F# , I feel that there is absolutely no need for him to exist.
It is better to renew the VB. Net for a long time. Develop some functional components to implement scientific computing. Similar to MATLAB
But he really shouldn't have developed it as a separate programming language.

If you turn PowerShell into a separate programming language. Or even cross-platform, the goal is to kill pythn and become the world's first scripting language.

Microsoft has fallen behind Google and Apple in the mobile space. In September 2013, Ballmer proposed buying Nokia's handset division with the primary goal of accelerating the growth of Windows Phone and helping to improve Microsoft's overall device ecosystem. During the acquisition, Ballmer asked his senior management for a public vote, and Nadella voted against it. His reason: I really don't understand why the world needs a third operating system.Unless we can change the rules of the game. Because Nadella's strategy is not to build a new closed operating system, but to use Microsoft's unique technology, so that Microsoft's services can run on a variety of devices, Android, IOS, Linux can run, thus expanding the ecosystem of Microsoft technology, not only mobile phone business, but also cloud business. He also changed Windows Azure to Microsoft Azure to inform people that Microsoft's cloud is cross-ecological, trying to make cloud services across any device and any platform. He was also the first Microsoft CEO to publicly show off an IOS phone at a press conference.

[Shaggy Hiker]

If you had a degree, you had C.

Heck, it's the first line of the US national anthem.

[Shaggy Hiker]

I've just got one thing to say to all the people trashing VB6 programmers: If what you say is true, then why isn't there a strong VB5 or VB4 or even VBDOS community around, like there is for VB6?

Because there was an upgrade path out of VB4 and VB5. With no upgrade path out of VB6, it became a dead end...that wasn't dead...so, an undead end. With no simple way out, people either had to make a laborious change or stay where they were. For some, such as me, the change wasn't all that laborious, as I could pretty much always just cut away everything done before and start over, but that wasn't an option for lots of people.

Still, of one thing we can be certain: VB programmers of all stripes will perpetually have a chip on their shoulders.

[xiaoyao]

The following just proves that f # is actually a glue language, a scripting language.
But Microsoft has enough products, and no one likes him to invent new things. Because a lot of things in Microsoft will become a kind of thing that will be abandoned after three to eight years after trying.

In fact, it is correct not to think of VB. Net as a separate programming language.
Develop it as a scripting language, a glue language. And python is powerful enough that we can't kill it.

Microsoft employees can at least write scripts in VB. Net, and they don't even need to install idE, just write code in Notepad. Or open a web page and debug it directly. Then a simple compilation tool is built into the system. Like vbsedit IDE

VB. Net calling JS, Java, python, and various other languages package class libraries, no problem.
Microsoft Research has also explained the F # language, which integrates many advantages such as security, performance, scripting and
Modern Runtime systems (Java Virtual Machine and Microsoft Universal Runtime), and supports interactive scripting languages such as Python, strong type inference, and ML security. In addition, the F # language has access. Net library and other database tool software.


Functional compilers on the market can already translate standard meta-languages into network architectures, and it seems unnecessary to develop F , but Microsoft claims that it hopes F # will be seamless. "I guess anyone in the IT industry knows that this means' the language of the future 'and that Microsoft has the power to make dictatorial decisions about it," an anonymous source said in a comment.

[The twinbasic, 100 megabyte development tool can do this.? F # requires 5000 MB and is a complete failure.]

[Elroy]

Good points all, Shaggy.

Because there was an upgrade path out of VB4 and VB5. With no upgrade path out of VB6, it became a dead end ... and start over, but that wasn't an option for lots of people.

Precisely, and it bewilders me why some people can't just accept that, and have to resort to attacking personalities.

Still, of one thing we can be certain: VB programmers of all stripes will perpetually have a chip on their shoulders.

That's certainly true for me. I'll die feeling betrayed and abandoned by Microsoft because they didn't provide a workable upgrade path, IMO (regardless of what others say).

[xiaoyao]

Microsoft's ambition was too great, which was also a major reason for its failure. They want to monopolize everything, so they will invest huge amounts of money and can only be abandoned if they fail.


There is no problem with this. Like investing in stocks, it is impossible to succeed every time.
But the excellent works and a lot of good software are mercilessly abandoned. A lot of wrong decisions are really wrong.
If we can buy less of these useless Nokia companies or others. Maybe VB6, VB. Net, will last 200 years on its development budget.

Sometimes Microsoft's management has too much power and can make a lot of wrong decisions.
There are many development teams in the same company. f#,vb.net,vb6.foxpro,java++
These team funds compete for resources and get more bonuses.They will fight each other to the death.

[xiaoyao]

If we can develop a plug-in for twinbasic, we can convert VB code directly to python code.
Or pull together all of python's class libraries. It's like controlling the excel. Applicatiom,Then we have all the features of python, the most popular language in the world.

[Shaggy Hiker]

Microsoft's ambition was too great, which was also a major reason for its failure.

Are you sure about that? I do think they've made some questionable choices over the years, but the bottom line is that they are currently the most valuable company in the world, so their ambition doesn't seem to have resulted in failure.

[Niya]

In 40 years of programming, I've never seen the need to define a lower limit different from zero.

Funny thing. By default I used to start all my arrays from 1, a habit I had all the way from QuickBasic to VB6. After all a child learns to count from 1 so it seemed natural to me as a child writing BASIC code. It wasn't until far later when I started interfacing VB6 and C++ that I started to reconsider that habit since C/C++ demands a lower bound of 0.

At first I found it annoying but eventually I couldn't do it any other way. I found offsetting calculations to be far easier with 0 based indexing. For example if you had an array of pixels for a 10x10 surface laid out in row-major format and you wanted to get the entry for a specific pixel based on an X,Y value. For 1 based indexing the math looks like this:-

Dim pixelIdx = (X + (Y * 10)) + 1

With zero based indexing, it just comes to this:-

Dim pixelIdx = X + Y * 10

A lot of offsetting calculations demand you add the lower bound as the final step when you use non-zero based indexing schemes. I just got sick of those annoy +1s and eventually abandoned the practice of defaulting my arrays to a lower bound of 1.

[Elroy]

Well, I suppose I'll address this array bound thing.

Yeah, I too started a while back always using 0 as the lower bound. In fact, I often don't even specify the lower bound in VB6.

But that's not what the comment was about. The comment was about the massive neglect that Microsoft had regarding backwards-compatibility. It was a HUGE snub to all the VB6 programmers and VB6 code (the most prevalent programming language at the time) that was out there. And I can't help but believe that it was somewhat deliberate, or, at least massively disrespectful.

[Niya]

Yeah, I too started a while back always using 0 as the lower bound. In fact, I often don't even specify the lower bound in VB6.

Welcome to the civilized world

But that's not what the comment was about. The comment was about the massive neglect that Microsoft had regarding backwards-compatibility. It was a HUGE snub to all the VB6 programmers and VB6 code (the most prevalent programming language at the time) that was out there. And I can't help but believe that it was somewhat deliberate, or, at least massively disrespectful.

I can perhaps add some context to this. I can't speak for all .Net programmers but after using .Net for years, it's very hard for me to not come to the conclusion that .Net was designed for C# and Visual Basic was just an afterthought. There are many little tiny things that make me feel this way, and this array bounds thing is just one of those little things. C# follows a lot of the conventions established by it's namesake, C and as we all know, non-zero lower bound arrays are not first class citizens in C, therefore it's not so in C# either. In the context of first class citizens, VB.Net could be very loosely thought of as a subset of C#. So if something is not a first class citizen in C# you can be sure it won't be in VB.Net.

However, all isn't lost. You can still create non-zero lower bound arrays in .Net with a little know-how:-

        'Creates and Integer array of 10 elements with a lower bound of 1
        Dim numbers As Array = Array.CreateInstance(GetType(Integer),
                                                    New Integer() {10},
                                                    New Integer() {1})


        Debug.WriteLine(numbers.GetLowerBound(0))
        'Prints 1

This can be done in any .Net language but it isn't ideal and I certainly won't recommend anybody write this kind of .Net code. For one thing, that code is extremely slow compared to using first class constructs and when I say slow, I mean slow. It's approximately 28 times slower. The other problem is that it's not strongly typed as you have to declare the type as Array. This doesn't give you the protection of the compiler so bugs will have to be caught at runtime.

[Elroy]

...it's very hard for me to not come to the conclusion that .Net was designed for C# and Visual Basic was just an afterthought.

Now that's a statement I couldn't agree with more. IMHO, Microsoft just tossed out a wonderful cash-cow when they effectively tossed out VB. As I've stated previously, it's a wonderful mid-level RAD language, with a second-to-none GUI designer.

[Niya]

Now that's a statement I couldn't agree with more. IMHO, Microsoft just tossed out a wonderful cash-cow when they effectively tossed out VB. As I've stated previously, it's a wonderful mid-level RAD language, with a second-to-none GUI designer.

I think an objective argument can be made that C# and .Net were a resounding success for Microsoft. My guess is Microsoft was also expecting VB.Net to be a hit with VB6 programmers even as an afterthought. Interestingly enough, I actually think it was.

Now before you flay me for that statement, understand what I'm saying here. Yes, there are a lot of talented VB6 programmers here that were devastated by MS's decision to abandon VB6 and start from scratch with VB.Net but I get the distinct sense that this is a minority of the VB6 community. The thriving community of VB6 was gutted by the arrival of .Net because many left VB6 for C# and quite a few like me left for VB.Net.

The majority of VB6 holdouts can be found on VBForums but when I stroll by other communities to see what folks are doing in the world of programming, I feel as if VB6 doesn't even exist anymore. No one is talking about it. Hell, it doesn't even feel like VB.Net exists outside of this site. Rust, Python, Go, C# and even C/C++ dominate a lot of conversations these days.

I say all of that to say this: From Microsoft's perspective, it was a good decision. They were able to create a massively successful platform at the expense of a few very disgruntled VB6 programmers. I suspect the reason MS never gave a second thought to VB6 was because enough of us left VB6 to justify it and C# attracted a LOT of new talent to the world of Windows programming in .Net. There simply aren't enough of super passionate VB6 programmers left to justify a reaction from Microsoft.

Just to give a little context to what I'm saying here so you know I'm not making this up to provoke anyone, here are some stats I pulled off StackOverflow:-


As you can see there, C# is very popular getting 121 questions per day while VB6 gets 6 a month. On the Tabs page of StackOverflow, C# is the 4th most popular programming language currently(Python is number 1) so Microsoft's decision to invest in C# instead of Visual Basic paid off big time.

[xiaoyao]

At first I found it annoying but eventually I couldn't do it any other way. I found offsetting calculations to be far easier with 0 based indexing. For example if you had an array of pixels for a 10x10 surface laid out in row-major format and you wanted to get the entry for a specific pixel based on an X,Y value. For 1 based indexing the math looks like this:-

Dim pixelIdx = (X + (Y * 10)) + 1

With zero based indexing, it just comes to this:-

Dim pixelIdx = X + Y * 10

A lot of offsetting calculations demand you add the lower bound as the final step when you use non-zero based indexing schemes. I just got sick of those annoy +1s and eventually abandoned the practice of defaulting my arrays to a lower bound of 1.

If you want to bind an array of pixels from an image control or memory DC. In fact, you can also define subscripts starting at 1.He doesn't have a rule that says you have to start from scratch.

[Niya]

If you want to bind an array of pixels from an image control or memory DC. In fact, you can also define subscripts starting at 1.He doesn't have a rule that says you have to start from scratch.

This thought came up because I recalled an image class I wrote that stored the image's pixels as an array of 32 bit Integers. One of the methods of the class allowed you to select a pixel by using an X and Y coordinate. Image surfaces are typically based on a kind of Y-inverted cartesian coordinate system which starts at 0,0, not 1,1 so you will always have to do +1s and -1s if your backing array is 1 based and this leaves a lot of room for bugs. On occasion I would mix up +1 and -1 back when I used 1 based arrays as my default.

[xiaoyao]

As you can see there, C# is very popular getting 121 questions per day while VB6 gets 6 a month. On the Tabs page of StackOverflow, C# is the 4th most popular programming language currently(Python is number 1) so Microsoft's decision to invest in C# instead of Visual Basic paid off big time.

It is because Microsoft killed its own child, VB6, VB. Net, so it leads to fewer and fewer users.
I saw the opposite result. VB. Net, around 2020, the number of users has exceeded that of C#

Because.net core is cross-platform. Added a lot of VB users.
But at this time, Microsoft said that we no longer regard VB. Net as a language to develop. That is to say, there is no such language.He just said that we were still allowed to use it, but it would never be updated.
Instead, Microsoft employees were banned from using python from the start. Instead, the founders of py were invited to the company to develop Microsoft cloud plug-ins. Support for the python language has even been added to the vs development tools.
vs=c#+vc++,python,f#

Microsoft has been adding f # to vs since 2002.
Just like Bill Gates married a new wife, and the original wife divorced.He even married several other wives.
Actually, it's like an acquisition. Nokia mobile phone company. And developing windows phone. These investments are hundreds of billions of dollars.

After all, Microsoft's main mission is to make operating systems and vs development tools. Ms SQL server database, office software。

On the one hand, he wants to do the mobile phone market, but. They also want to make hundreds of thousands of dollars by selling licenses to each mobile phone.
They want these costs to be passed on to the mobile phone manufacturers.
To put it simply, if you open a hospital and then invest in a modified version of Nokia's mobile operating system.
Operate across industries. Bill Gates has so many wives, and it's normal to divorce a few.

It's just that he didn't think I could marry many wives, but why should I divorce?Like venture capital, many companies get money on the premise of signing. Vam requires companies to withdraw all investments and compensate for losses if they fail to reach a certain user or profit scale within a few years.

Capital is cruel, they will never say that I want to be with you for a lifetime, even if I lose more, I will support you.
So Microsoft turned its core assets into language, and threw away all the core technologies that were the first to start a business.Apple, Google or ChatGPT, they can become the richest people in the world in a few years. The largest listed company by market capitalization.

The most valuable company has never been Microsoft. He's just an old lion, not dead yet.
Bill Gates went to start a company and took a lot of orders before he graduated from university.Earned the first pot of gold with So, until about 1998, in fact, it also developed for nearly 20 years.

The most successful moment of VB6 was killed by Microsoft. Maybe it was Ballmer who made the decision, maybe not Bill Gates.
A lot of individual developers, they invented some programming idEs, like powerbasic, vbsedit, but they didn't give up until they died.
Bill Gates has thousands of software products, so he sells them crazily and kills them.
In fact, sometimes we don't necessarily need to follow up all his functions every year. c#

We only need to add a small amount of functionality like office's Vba. Fix some bugs. It may be enough to invest $10 million. a year. For Microsoft, it's just a cup of milk tea.
But his decision is like this child is not mine, and my wife is with some man.

What Microsoft needs is. Think of him as your own son.
Originally, ie and IIS all support VBS, and his goal is to run on the browser like JS, but he failed 100%.The main reason is that he is not open source, he is not cross-platform, and others dare not use it.

[xiaoyao]

This thought came up because I recalled an image class I wrote that stored the image's pixels as an array of 32 bit Integers. One of the methods of the class allowed you to select a pixel by using an X and Y coordinate. Image surfaces are typically based on a kind of Y-inverted cartesian coordinate system which starts at 0,0, not 1,1 so you will always have to do +1s and -1s if your backing array is 1 based and this leaves a lot of room for bugs. On occasion I would mix up +1 and -1 back when I used 1 based arrays as my default.

I also often use the pixel array of the image for editing. For example, RGB three bytes. Or rgba 4 bytes.
I can bind it to an array of integers to determine if the pixel has data.
With this method, its speed is increased by 50% to 200%.
In fact, its subscript can. From 1, no one stipulates that he must use the beginning of 0. It's just, it's just a lot of people are used to doing this. This is the code you can search on the Internet.

b GetPictureBytes()
  Dim PicBits() As Byte, PicInfo As BITMAP

  GetObject Picture1.Picture, Len(PicInfo), PicInfo

  ReDim PicBits((PicInfo.bmWidth * PicInfo.bmHeight * 3) - 1) As Byte
  GetBitmapBits Picture1.Picture, UBound(PicBits), PicBits(0)

  ReDim PicBits(1 to (PicInfo.bmWidth * PicInfo.bmHeight * 3) ) As Byte
  GetBitmapBits Picture1.Picture, UBound(PicBits), PicBits(1)

for x=1 to width
for y= 1 to height
 color1=rgb(PicBits(i+1)  ,PicBits(i+2), PicBits(i+3))
i=i+3
next
next

In fact, the subscript starts from zero, and the biggest advantage is that the running speed can be improved a little bit.
3048 * 2156 Width and height. In the assembly process, the addition calculation can be reduced so many times.

i=1
for x=1 to width
for y= 1 to height
 color1=rgb(PicBits(i)  ,PicBits(i+1), PicBits(i+2))
i=i+3
next
next

[Niya]

In fact, its subscript can. From 1, no one stipulates that he must use the beginning of 0. It's just, it's just a lot of people are used to doing this.

There's actually a solid reason for this convention. At the machine code level you never think of array access as getting element number n, instead, you think in terms of base + index * size or base + index * size + displacement. In fact, x86 processors support these as addressing modes so they're built right into the processor. These addressing modes are meant to handle various scenarios like arrays of multi-byte data like integers or structures and they all resolve the first element from an index value of 0.

When you use a non-zero based index in a higher level language, you're actually giving your compiler and processor a bit of extra work because it will always have to add an extra displacement to the calculations to match how the hardware actually works. It could be in the form of an extra addition or subtraction instruction to change the base pointer or the displacement could be encoded into the instruction itself I think.

Perhaps in ancient times, when performance was at a premium, this was a serious concern. However, the impact of the extra displacement today is negligible enough to be completely ignored, so you can safely continue using 1-based indexing if you prefer.

[dz32]

Isn’t c# / .net basically a reworked j++ after they got sued? That would put c# as the primary implementation language as well.

[Niya]

Isn’t c# / .net basically a reworked j++ after they got sued? That would put c# as the primary implementation language as well.

Yes. C#/.Net is well known to have started out as a Java clone. It has since diverged though. Java as far as I know runs on a VM while .Net applications execute as native code.

[dz32]

Is the .net default compile still IL that gets jitted to native code? Seem to remember an optional manual post processing compiler that could convert to full native code. It’s been a while details fuzzy

[Niya]

Is the .net default compile still IL that gets jitted to native code? Seem to remember an optional manual post processing compiler that could convert to full native code. It’s been a while details fuzzy

.Net typically compiles to IL but the CLR has never had a bytecode VM that executes IL. Under certain conditions like a Linux build, .Net will produce a self-contained native binary. There are also other tools and options that could coerce a full native code build on Windows.

IL was never made to run on a virtual machine like Java's bytecode, rather it allowed the JIT to be developed completely independently source languages like C# or VB.Net. You or I could write our own .Net language if we wanted without ever having to worry about if it would compile and execute because all we have to do is compile to IL. That is the purpose of IL, to decouple the programming language from the machine code that eventually gets executed by the processor.

[xiaoyao]

Now click on the cross-platform language C# , f #, and Microsoft will continue to waste tens of billions of dollars in cost.
f#，from 2002，The development of this language led to the bankruptcy of VB6, the VB. Net.
It may have wasted tens of billions of research and development costs.
Perhaps Microsoft's idea is to kill MATLAB, a huge piece of high-tech software. Use the f # to beat him.
I am an ordinary amateur individual developer, I understand this truth, do not understand why Microsoft do not understand it?
I have always felt that using VB dinner or C shop to develop a complex formula calling method as a plug-in is enough to replace f#.

Perhaps 20 to 40% of scientific calculations can be done with VB. Net.
Some slightly simple point of the formula, I think with VBS,vba,python as a simple script VBS, VB. Net dynamic compilation can be, there is no need to install two GB of matelab runtime. After unzipping and installing, it may be more than 10 GB.
When you think about the programming languages used by businesses, names like Java, C, and C + + come to mind first. There will be absolutely no F , the open source functional programming language created and supported by Microsoft The lack of code, documentation and work is the biggest problem with F. When asked what was the most frustrating thing about learning F , respondents' highest responses were not having enough real-world code for learning (53%), not having a good tutorial/documentation (35%), and they didn't understand compiler errors (20%). Similarly, the top three saddest things about developing applications in F # are: reducing the complexity of refactoring tools (49%), having to use C # -centric tools (47%), and having trouble troubleshooting (21%).

The positive aspect of F # is that it has given a big boost to the new version of.NET (many things are optional for C # and http://VB.NET, but necessary for F , and the F # team has been following the CLR team to add those things), and C # has benefited from those things that were prepared for F.
This object is a bottomless pit ah, Microsoft needs to give him a lot of blood, but can not get any return.His only advantage is that he can make science and technology progress continuously.To put it simply, it is not for us ordinary people to use, it is only for those scientists or senior engineers of Microsoft's own company to use.
They can use other development languages, such as MATLAB, in order to save money, do not have to spend extra money to buy the software, They want to do it in dot net's own way.，Maybe the development cost is many times more than the cost of buying the software.

[xiaoyao]

Microsoft explicitly prohibits employees from using ChatGPT, and all internal devices block ChatGPT websites.
gpt doesn't steal the code on your computer。If Microsoft is afraid that these codes of employees will develop on the Internet, it is better to prohibit employees from publishing codes on the Internet to seek help, which can only be solved within the company.

[PlausiblyDamp]

Now click on the cross-platform language C# , f #, and Microsoft will continue to waste tens of billions of dollars in cost.

So where is your proof that it wastes billions of dollars?

f#，from 2002，The development of this language led to the bankruptcy of VB6, the VB. Net.

F# was created around 2005, VB6 had a final release in 1998. Your numbers don't add up.

It may have wasted tens of billions of research and development costs.

And then again it might not, MS are still a large company worth billions - perhaps this wasn't a waste after all.

Perhaps Microsoft's idea is to kill MATLAB, a huge piece of high-tech software. Use the f # to beat him.
I am an ordinary amateur individual developer, I understand this truth, do not understand why Microsoft do not understand it?
I have always felt that using VB dinner or C shop to develop a complex formula calling method as a plug-in is enough to replace f#.

Different languages can have different strengths - F# being a functional language lends itself to engineering and statistical work, features such as units of measure make it easier to avoid simple mistakes. Nothing stopping you doing the same in another language, but it might involve more code that is harder to read.

Perhaps 20 to 40% of scientific calculations can be done with VB. Net.
Some slightly simple point of the formula, I think with VBS,vba,python as a simple script VBS, VB. Net dynamic compilation can be, there is no need to install two GB of matelab runtime. After unzipping and installing, it may be more than 10 GB.
When you think about the programming languages used by businesses, names like Java, C, and C + + come to mind first. There will be absolutely no F , the open source functional programming language created and supported by Microsoft The lack of code, documentation and work is the biggest problem with F. When asked what was the most frustrating thing about learning F , respondents' highest responses were not having enough real-world code for learning (53%), not having a good tutorial/documentation (35%), and they didn't understand compiler errors (20%). Similarly, the top three saddest things about developing applications in F # are: reducing the complexity of refactoring tools (49%), having to use C # -centric tools (47%), and having trouble troubleshooting (21%).

Why only 40%??? You could do 100% of scientific calculations in any language, including assembler. Why is F# suddenly the focus of your anger?

[2kaud]

Bring back pascal/Delphi! You can specify the bounds of arrays as a type:

TYPE SUIT = (CLUB, DIAMOND, HEART, SPADE);
    VALUE = (ACE, TWO, THREE, FOUR, FIVE, SIX, SEVEN, EIGHT, NINE, TEN, JACK, QUEEN, KING);
VAR CARD : ARRAY[SUIT] OF VALUE;
....
CARD[HEART] = KING;

or:

TYPE COORD = (-10..10);
    BOOL = (0..1);
VAR MAT = ARRAY[COORD, COORD] OF BOOL;
...
MAT[-5, 7] = 1;

I miss stuff like that in c/c++ ....

[yereverluvinuncleber]

40% of the time it works EVERY time.

[Shaggy Hiker]

IL was never made to run on a virtual machine like Java's bytecode, rather it allowed the JIT to be developed completely independently source languages like C# or VB.Net. You or I could write our own .Net language if we wanted without ever having to worry about if it would compile and execute because all we have to do is compile to IL. That is the purpose of IL, to decouple the programming language from the machine code that eventually gets executed by the processor.

We need a Rockstar.Net.

I really like that language, though it isn't useful for more than some math types of things. The idea of writing source code to serve two purposes has a real appeal to me. Poetry that does something useful would be designing on two different levels.

[xiaoyao]

Microsoft continues to develop C # and f # across platforms. Abandoned the VB. Net.

C #/F # Performance Comparison The best thing about pure math is what PerpetualCoder says, F # looks more like a math problem, so it will be easier for mathematicians to write. When I checked it out, it reminded me a lot of MATLAB.

If you're talking from a performance perspective nothing beats C. If you're talking from a library I believe in functional languages (F # and like), Fortan (yes it's not dead yet), Python has good math libraries.
Perhaps this market will eventually be replaced by python。
The f # has been in development for 22 years, and it seems that Microsoft takes it very seriously.
The code is simple and easy to understand, should be the most basic office workers, data analysts, financial personnel quickly, very simple can be called.
In fact, this is the best aspect of VB6, VBA, VBS, and python

Although F # is a modern programming language, the language design is far less elegant than Scala. F #'s standard library is incompatible with the.NET standard library, and F #'s algebraic data type (ADT) -based type system is incompatible with the.NET native class library style. The F # ecosystem is even weaker than Scala. In engineering practice, Scala has a large number of frameworks and third-party libraries for both object-oriented and functional programming.Not to mention that the Scala standard library itself blows up the F # standard library. I think Microsoft, under the leadership of Ballmer, who sold DVDs, was a bit obsessed with the understanding of technology, designed APIs that were bloated and stuck to backward compatibility. So the language designed by Microsoft is the same. It has not thought about how to integrate functional language into object-oriented language at all. It is simply compatible with OCaml syntax and adds a lot of irrelevant functions. By contrast, Martin Odersky has worked on Pizza, Generic Java, and has tried to implement functional programming in the JVM for many years. Finally, Scala abandoned the source code compatibility with Java, and finally smoothly integrated functional programming features with object-oriented.

Haxe is developed in OCaml, as was Rust originally, and F # is a Microsoft upgrade to OCaml that seamlessly calls C # and other.net resources (similar to Scala, which is a functional upgrade to Java). Given that the.net core can support Windows/Linux/Mac, it's a good prospect.)


Knowing so much, it is finally determined that this thing is not something that ordinary people can play at all. It is used by university professors and some high-end people. The same way you use a $5,000 iPhone.
Doomed not to have many users.

What the majority of users need is an Apple mobile phone of about $800 to $2000.
So Microsoft's development language is getting farther and farther away from the public.
The python could crush him with a slap.

[PlausiblyDamp]

Microsoft continues to develop C # and f # across platforms. Abandoned the VB. Net.

C #/F # Performance Comparison The best thing about pure math is what PerpetualCoder says, F # looks more like a math problem, so it will be easier for mathematicians to write. When I checked it out, it reminded me a lot of MATLAB.

If you're talking from a performance perspective nothing beats C. If you're talking from a library I believe in functional languages (F # and like), Fortan (yes it's not dead yet), Python has good math libraries.
Perhaps this market will eventually be replaced by python。
The f # has been in development for 22 years, and it seems that Microsoft takes it very seriously.
The code is simple and easy to understand, should be the most basic office workers, data analysts, financial personnel quickly, very simple can be called.
In fact, this is the best aspect of VB6, VBA, VBS, and python

Although F # is a modern programming language, the language design is far less elegant than Scala. F #'s standard library is incompatible with the.NET standard library, and F #'s algebraic data type (ADT) -based type system is incompatible with the.NET native class library style. The F # ecosystem is even weaker than Scala. In engineering practice, Scala has a large number of frameworks and third-party libraries for both object-oriented and functional programming.Not to mention that the Scala standard library itself blows up the F # standard library. I think Microsoft, under the leadership of Ballmer, who sold DVDs, was a bit obsessed with the understanding of technology, designed APIs that were bloated and stuck to backward compatibility. So the language designed by Microsoft is the same. It has not thought about how to integrate functional language into object-oriented language at all. It is simply compatible with OCaml syntax and adds a lot of irrelevant functions. By contrast, Martin Odersky has worked on Pizza, Generic Java, and has tried to implement functional programming in the JVM for many years. Finally, Scala abandoned the source code compatibility with Java, and finally smoothly integrated functional programming features with object-oriented.

Haxe is developed in OCaml, as was Rust originally, and F # is a Microsoft upgrade to OCaml that seamlessly calls C # and other.net resources (similar to Scala, which is a functional upgrade to Java). Given that the.net core can support Windows/Linux/Mac, it's a good prospect.)


Knowing so much, it is finally determined that this thing is not something that ordinary people can play at all. It is used by university professors and some high-end people. The same way you use a $5,000 iPhone.
Doomed not to have many users.

What the majority of users need is an Apple mobile phone of about $800 to $2000.
So Microsoft's development language is getting farther and farther away from the public.
The python could crush him with a slap.

Seriously, what is this obsession with F# destroying Microsoft and People developing apps on an Android phone is your previous posts, but now an iPhone, using Python all about?

[xiaoyao]

Microsoft hates all simple programming languages. Including python, VB6, VB. Net,
But f # it is professionally calculated, so it will remain.I just don't know how many people buy vs development tools because they need this scientific and technological calculation.

In fact, if it's good enough, most universities or companies don't need to buy MATLAB, which must be a great invention.But he doesn't have the ability.
As I understand it, f # should not be developed exclusively as a programming language.The next step is to see if Microsoft can continue to develop for 20-30 years, maybe 5 years.

Take the price of mobile phones as an example. That is to say, for a product, we need high-end products with high prices and very cheap products.
In China, it is a cloud computing service company like Amazon cloud and Microsoft cloud. If Amazon's price is $2000 per year.
Some of the low-end prices in China are. Twenty dollars. For those college students who study in school, there are many companies and individuals who need huge traffic to start a business.

Cheap products and the products that sell the most. Free product. The most expensive products, in fact, these things need to be considered in a balanced way.For example, ms sqlserver also has a community version, the free version MSDE.He can't say he's stuck in the 2010 version forever. This is the difference between stopping updating and completely abandoning it.

python,It can also do scientific calculations. If he can still run in the Android phone. Or even desktop icons. It's like an app. If repackaged, it can even be compiled into apk.
Does not show windows Linux, Android, Apple. Py runs stably on all operating systems. Compile to exe, compile to apk.
In fact, he can do it. A simple language. Omnipotent. The only difference is whether he wants to do it.

https://www.b4x.com/android/forum/th...y.19332/#posts

b4a.ide,Android apps have been developed with VB syntax for more than 10 years.If Microsoft buys it and makes it cross-platform. This must be another great invention, which can attract more users to join Microsoft products.
I don't know how many people now use VB. Net or other Microsoft languages to develop Android apps.It must be a huge development tool.
Sometimes, like python, it only needs 100 MB of hard disk capacity, b4a, and the core function may also be about 20 MB. And twinbasic, he only has 30 megabytes.

Android app ,ios app,can be written with VB code, and cross-platform Java programs can be developed. These are all great inventions.

Of course, Microsoft certainly rejects such a small development tool.
He likes to use huge vs tools to accomplish everything. In the end, no one used them at all.

[xiaoyao]

https://tieba.baidu.com/p/8582003868#/
https://learn.microsoft.com/en-us/xa.../xamarin-forms

It's very complicated to develop Android apps in VB. NeT, and it seems that someone has found a way to hack the configuration.
Microsoft has never been sincere about cross-platform app development.
Otherwise, he should provide a large number of various templates, even some relatively complete apps. For example, develop some small video apps like YouTube and Tik Tok.
You can download a large number of complete examples to get users started quickly.

[Niya]

Here is something I just came across based on something that showed up in my YouTube feed which is relevant to the discussion about how Microsoft thinks about programming languages.

Just for a bit of context, a small panic was started on Reddit because Microsoft apparently wants to rewrite some Office 365 services from C# to Rust. My feeling is that it scared some folks into thinking MS was going to abandon C#. A guy claiming to work for MS chimed in. There is no way for me to verify whether this guys is who he says he is but I can't imagine why someone would pretend to be an MS employee just to make a basic Reddit post so judge it as you see fit.

This is what he said:-

Hey there!

I work at Microsoft. I can shed a bit of light here.

We use .NET for TONS of things. Absolutely tons of different products and services. I'm on the Office 365 side of the business, currently managing Deployment for all of the hundreds of services that roll out across the world... And we use .NET extensively.

I'm starting my new position managing a team that does routing next week. They have some things that are extremely performance critical. As others have pointed out, we're talking about supporting services and traffic literally across the planet. When it comes to optimizing, they will find ways to squeeze out what they can.

There are languages like C and C++ that get used for some extreme use cases like I mentioned. Reducing as much overhead as possible in certain situations even leads to .NET apps with unmanaged pieces included with them.

There's been a lot of hype around Rust, and for good reason. But it's a system language. It's not like Microsoft is about to go rewrite millions and millions of lines of code and toss out C# (for anyone getting nervous &#128517. They're just being pragmatic and using an effective tool for the job.

Hope that offers some clarity.

He responded with this post later in response to someone suggesting that MS is sending mixed messages by always boasting about improvements in .Net performance yet NOT writing their own tools in .Net.

I think it's important to consider scale here.

We're talking about hundreds and hundreds of services spanning many teams tackling many different things. DotNet performance has been increasing at incredible rates. There is a ton of work put into that and many teams adopting patterns and practices to enable it.

But a programming language is just a tool. DotNet is incredible, but like all tools, there is no tool perfect for every job. There will always be scenarios where something might be a better fit, given the constraints or requirements.

I think the message is pretty clear, personally:

DotNet continues to get faster

MSFT is very invested in DotNet (it's the overwhelming majority of services using it. I don't have stats to back this, just anecdotal from supporting service owners)

MSFT is focused on delivering the best possible experience, which means we'll use appropriate tools in different situations.

Again, I think that there's a misrepresentation in this thread... Having some things in Rust (many of which are already not. C#) should hopefully not suggest to others that MSFT is just departing from C#. We're talking about a tiny fraction of what's in Office 365.

I hope that helps!

I'm not trying to make any particular point here but I thought it gave some decent insight into how MS devs think about programming languages. This of course assumes you believe he is who he says he is which personally I do believe.

Here's the actual thread for reference:-
https://www.reddit.com/r/dotnet/comm...rosoft_career/

[Niya]

Here's another post from that same thread I found quite interesting. Again, it's from a guy claiming to be a former MS Dev that I can't verify so judge as you see fit:-

Former Microsoft dev..

There are tons of MSFT code written in c/c++/rust and similar unmanaged languages. Much of core apps (windows , office apps and even mobile ones) will be written in unmanaged languages because ..

Performance, especially on desktop/devices.

Microsoft can afford it ! (Read below)

The internal ballpark estimate is that it takes 10x the devs , 10x time to get any software written in unmanaged languages versus managed ones (c#).

So, internally c# is the language of choice of everything that would not run into performance issues when the load on the software runs into 100s of millions or even billions.

Heck a whole bunch of software is actually first developed in C# to see if it gets adopted and gets popular enough, before 10x the money is spent on developing it in c++/rust.

Lots and lots of internal apps (even cloud services ) within Microsoft is written in C#, sometimes usually for the first iteration. Unmanaged development is so expensive that very high level approvals and RoI studies have to be provided before any team is allowed to move away from C#.

C# , especially with .net core is at an amazing sweet spot and one should not hesitate to invest their time on it..

I find it interesting because he mentions some stats about development time that seem to be correlated with whether you're writing managed or unmanaged code. I don't know if it's true or just anecdotal but I do find it an interesting thing to ponder.

https://www.reddit.com/r/dotnet/comm...t=share_button

[xiaoyao]

Maybe now it can only be forced to be transferred to C # development.
VB. Net, used as an aid in learning C# , It's like an upgrade project, a slow transition.
As a scripting tool, VB6 also develops some small functions.
The main problem is that because we have learned such a simple VB programming language, we are resistant to change and do not want to learn those complex languages.

If speed is not important, then C# is perfectly adequate.Where high performance is required, such as a virtual machine kernel, or VC + + development.There are some nuget packages that may be able to make some of the functions into VC + + DLLs to achieve the goal of high performance.

[xiaoyao]

So far in 2020, F # has not yet become popular, but as Microsoft's only FP language, F # has a special position and can be paid attention to. Ecosystem. The F # development community is small and there is no library equivalent to Elixir If you want to develop a script in a Microsoft language, what should you use? *.bat command line, PowerShell, f # parse run, excel VBA, VBS, VB6, VB.Dynamic Compile.net runs the test. VB (net code), and it feels like everything can be done, nothing works well, and nothing is done well. The python node. JS can perfectly solve these problems, mainly by adding the functions of directly calling Java and.net class libraries, Windows API calling, com technology, and shell calling other programs (including running the console and returning text).

[xiaoyao]

Python can call Java class libraries.net class libraries,
Perhaps the reason why Microsoft's VB family has all gone bankrupt and been abandoned is that python also occupies an important position.

import clr
clr.AddReference("System.Windows.Forms")
from System.Windows.Forms import Form

import jpype
import os
jar_path = os.path.abspath('.') + '/AddTwoNumbers.jar' # get the absolute path to the jar package jpype.startJVM('usr/local/java/jdk... /jre/lib/amd64/server/libjvm. so', '-ea', '-Djava.class.path=%s' % jar_path) # load java virtual machine, the first parameter is the Java jdk installation location, can be viewed by env | prep JAVA_HOME; the second parameter is written as the example; the third parameter is the absolute path of the jar package;
 java_class = jpype.JClass('AddTwoNumsClass') # get the specified Java class by inputting the class name
 result = java_class.addTwoNumsFunc(1, 2) # call the specified function of the class by the function name
 jpype.shutdownJVM() # shut down the Java virtual machine, if not written will be shut down automatically when the Python program exits

[Niya]

Again, Python and VB6 don't occupy the same niche. They are not competitors. VB6 is a Windows-based RAD environment geared towards data-centric thick client application development. Python on the other hand is typically used as a "glue language" to bridge the gap between fundamentally incompatible systems.

[Schmidt]

Python can call Java class libraries.net class libraries,

import clr
clr.AddReference("System.Windows.Forms")
from System.Windows.Forms import Form

import jpype
import os
jar_path = os.path.abspath('.') + '/AddTwoNumbers.jar' # get the absolute path to the jar package jpype.startJVM('usr/local/java/jdk... /jre/lib/amd64/server/libjvm. so', '-ea', '-Djava.class.path=%s' % jar_path) # load java virtual machine, the first parameter is the Java jdk installation location, can be viewed by env | prep JAVA_HOME; the second parameter is written as the example; the third parameter is the absolute path of the jar package;
 java_class = jpype.JClass('AddTwoNumsClass') # get the specified Java class by inputting the class name
 result = java_class.addTwoNumsFunc(1, 2) # call the specified function of the class by the function name
 jpype.shutdownJVM() # shut down the Java virtual machine, if not written will be shut down automatically when the Python program exits

Python has no problems, calling into the win32/64-COM-"layer"
(which - on Win-OSes - still plays a more important role than the .NET-runtime)

import win32com.client
D = win32com.client.Dispatch("Scripting.Dictionary")

D.Add("abc", 123)

print(D("abc"))

Olaf