> if c++/cfront didn't ride on the tails of c, I'm skeptical it would've seen widespread use
What launched C++ into success was Zortech C++. At the time, 90% of programming was done on MS-DOS. Cfront was nearly unusable on DOS, because:
1. agonizingly slow to compile
2. no support for near/far pointers, which was essential for non-trivial apps
Zortech C++ fixed those problems, and sold like wildfire. This provided critical mass for C++ to succeed. The traffic on comp.lang.c++ angled sharply upward. Borland saw our sales, and abandoned their OOP language product and did Turbo C++ instead. Microsoft saw Borland's success and then did their own C++.
We sold a lot of Zortech C++ compilers to Microsoft. They used it to develop COM.
I heard rumors that Microsoft was developing their own OOP C, called C*. I've never been able to confirm it, though.
Your perspective on this may be distorted due to your personal involvement. Do you believe MSVC++ or Turbo C++ would still have existed without Zortech C++ arriving first? Because if so then I don't think you can really take credit for C++ popularity on the PC.
> Do you believe MSVC++ or Turbo C++ would still have existed without Zortech C++ arriving first?
Nope. As I mentioned, Borland was working on their own OOP C language. ZTC++'s success caused them to abandon it and do Turbo C++. I know this because of my conversations with Eugene Wang. And Turbo C++'s success caused Microsoft to do MSVC++, which was quite late to the game.
Before C++, the newsgroup traffic on C++ and ObjC was about the same. The C++ traffic took off after ZTC++ was released.
You can verify some of this by:
1. looking at newsgroup programming language traffic over time
2. looking at the release dates of Zortech C++, Turbo C++, and MSVC++
3. looking at the programming magazines of the time and what language/compiler they used
4. looking at the programming magazines at the time and seeing they were all about programming on the PC, not mainframes, and not Unix.