Hacker News Reader

Comment by makapuf | Hacker News Reader

The readme says it should be compiled with borland c++. Do you know why wasn't turbo c/c++ used for these ?

As far I remember (I used these tools extensively at the time), Borland C++ was the bigger, more complete and more powerful package. It was basically the professional version while Turbo C++ was the hobbyist one.

I had the same recollection as you, but just looked through a Turbo C book I have here and couldn't find any reference to its big brother. A bit of Googling later, and it turns out Borland C++ was Turbo C/C++'s successor. The releases went like this: Turbo C 1.0, 1.5, 2.0, Turbo C++ 1.0, Borland C++ 2.0, then onwards. So Keen may require features that only appeared in 1991 onward.

Microsoft took more of the approach I think we're remembering. For example, BASIC PDS 7.x being the more professional version of QuickBasic 4.5.

So then where does Turbo C++ 3.0 fit in? My copy is (C) 1992, but it was still being sold in 1999, alongside what I'm pretty sure I remember as Turbo C++ 4.5 that targeted versions of Windows that were not out in 1991...

Boy, remembering that really digs up some nostalgia...

edit: I see the "Borland C++" wikipedia page present it as the successor, but the "Turbo C++" wikipedia page describes it as the hobbyist counterpart. The release dates and version numbers line up really well, so that makes sense. I was a hobbyist at the time.

Hmm, weird, I don't know. I only got as far as Turbo C 2.0 and then jumped ship to Turbo Pascal for the next several years (I know, I know.. not a great decision in hindsight!) ;-)

The Turbo/Borland distinction was present for both C and Pascal products from Borland all the way up to Turbo/Borland C++ 4.5 and Turbo/Borland Pascal 7.0.

For instance (different language but same company), Turbo Pascal 7.0 and Borland Pascal 7.0 co-existed at the same time. They were simply targeting different markets. I vaguely remember the same thing about Turbo C++ and Borland C++...

Or why not any compiler from the last 15 years

If it ain't broken, don't fix it. What's the point of updating an old game's source tree to compile with the latest and greatest compilers?

Even if one wanted to do that, it would probably be best done as a fork.

Well because GCC

Sometimes the code was written specifically for nuances of a given compiler, and compiling the same code on a different one would produce side effects.

For one, the inline assembly syntax wasn't exactly the same.

When I was a kid I had a DOS game programming book, the worst thing about it was that it came with the Microsoft C compiler.

EVERYONE used Borland back then. So most of the sample code and shit you would find on the AOL forums or BBS's wouldn't work.

Sure, if you knew assembly and your compiler it was probably trivial to get it working. Same goes for functions to call interrupts and things like that.

If you had two compilers for the same language from different companies back in the 80s or 90s, the chances of them both successfully compiling the same codebase without any changes was slim to none. A bit like sharing SQL between Postgres, SQLite, SQL Server and Oracle nowadays.

Reply on news.ycombinator.com