Hacker News Reader

Comment by bencollver | Hacker News Reader

Wasn't QBasic the interpreter as opposed to QuickBasic the compiler?

It's been a long time, but my impression was that QuickBASIC had an interpreter and the ability to compile. Then later on, Microsoft bundled a more limited version called QBasic with later versions of MS DOS which lacked the compiler.

But all of them (QBasic, QuickBASIC, Microsoft PDS, and even Visual Basic for DOS which almost nobody remembers sadly) had the editor, interpretative execution, and built-in help.

I remember VB-DOS, and fondly too. It was magical. I think I used it even before VB3.

This matches my memory. When I got my dad's old work computer with QuickBASIC on it, and I discovered the compiler, and could write programs other people could "just run", I felt like a real programmer for the first time.

Yet you were even before that, the moment you made programs work at all.

Yes that was the case, by the time Visual Basic 5 came to be, its compiler was based on Visual C++ backend.

No, QuickBasic was both an interpreter and a compiler. QBasic was just an interpreter.

"Compiler". Even Visual Basic only compiled to p-code, which had to be interpreted at runtime. Not to fully native code.

That's why it always ran slower than Delphi.

Wrong, starting with with Visual Basic 5, a proper compiler was introduced based on Visual C++ backend, in addition to the P-Code interpreter.

Additionally VB devs no longer needed to rely on C++ for ActiveX controls, aka OCX, the VBX replacement.

VB6 (and IIRC 5 too) could compile to native, as seen in the compile options: https://imgur.com/a/v0QcbBU

P-code was still offered as an option because some wanted the smaller output binary sizes, and the build process was faster⁰.

Some incorrectly assume that the native option wasn't really fully compiled because the main supporting library (msvbvm60.dll) was still used¹, but this was for common library functions³ and the interpreter portion was not touched.

There were unofficial tools that would statically link your exe with the relevant VB runtime (and certain other libraries) but the use of those was rare.

----

[0] Though I don't think the build speed matter was actually significant for many, if any, workflows, even on really slow kit.

[1] Some didn't distribute it after a time, to reduce download sizes, as they were included with Windows so users already had them. Windows 7 (and maybe Vista?) included msvbvm60.dll and friends by default, and most XP and 98 installs² had it too as it came with Internet Explorer updates.

[2] though there was a compatibility break at one point that meant you needed to recompile with VB6sp6 if you hadn't included a local copy in your apps directory

[3] Much like many C programs don't have glibc statically linked into them, but work because it is practically ubiquitous on the systems they target.

Having to depend on msvbvm60.dll was hardly any different than msvcrt.dll, but try to explain that to most folks.

Both of these are incorrect.

Both QuickBASIC and the BASIC Professional Development System compiled to full native DOS code, and could make standalone EXE files.

VB finally gained this with VB 6 which could also make native EXE files.

QuickBasic produced a DOS .EXE file.

It didn't output p-code. You're confusing it with Visual Basic.

You are correct, and VB 6, the final version, at last included a native-code compiler as well.

All this brings back fond memories of my first programming foray, an ASCII game in QBASIC from Mars and Back: Computer Programming Handbook by Andrew J. Read. So much fun, so much frustration.

This is what I recall too. QuickBasic was perhaps BASIC's answer to Turbo Pascal, a relatively lightweight but usable text based IDE. I knew some happy users.

https://winworldpc.com/product/quickbasic/45 for a look at QuickBASIC 4.5 abandonware; they also had QuickC and QuickPascal.

No, the answer was Quick Pascal, however Microsoft didn't really cared that much about it.

Reply on news.ycombinator.com