I reply here to stop the tangent off the Mandlebrot topic that needs cleaning up.
NormanDunbar wrote:Windows? Pascal? Are you sure, because I was of the understanding that it was Macintosh no Windows. I've not found anything online about windows, but a few about Mac:
We have to be selective and investigate what the Net tells us or doesn't tell us. MS had to get Win out quickly so why use Pascal calling conventions if you were not going to interface MS DOS to Pascal? C was available but maybe the programmers were not and Pascal ones were and Fortran or Cobol were not a lot of use. Save a few bytes, really, when there was 640 K and we know an OS with a language can fit in 48 K. In those days the secrecy was paranoia about code theft but the Net is biased to the US and the last thing they want to admit is their computer industry is based on European developments. They rediscovered Borrelia and called it Lyme and broke the Enigma code according to parts of the NET. The most annoying one to me is calling 1000 million a billion but what will people 1000 years from now actually think and does any of it really matter?
I got my information from Tandy Trower who was the guy in charge of getting Windows "out the door" and while he mentioned Pascal, he was referring to Borland Pascal which was killing off Microsoft's Basic compiler both in terms of speed and cost - $60 for Borland Pascal, $400 for MS Basic compiler. He was given a choice of sorting out what to do about Borland (nothing he could!) or taking on the "ejector seat" of Windows manager. He chose the latter especially as most Computer Science graduates had been doing Pascal for years at university.
At no point did he mention Pascal as being what Windows was written in though -- which doesn't mean it wasn't Pascal of course!
NormanDunbar wrote:At no point did he mention Pascal as being what Windows was written in though -- which doesn't mean it wasn't Pascal of course!
Had a quick look and thought this quote fits the bill.
"While Microsoft had its own version of Pascal, it had been groomed as a professional developer’s tool, and in fact was the core language Microsoft wrote many of its own software products in before it was displaced by C."
Perhaps Windows wasn't considered a development.
By the way, he's been posting a new blog post every weekday since 2003. And I've been reading every single post of his since 2004... that makes roughly 4500 posts, wow.
By the way, he's been posting a new blog post every weekday since 2003. And I've been reading every single post of his since 2004... that makes roughly 4500 posts, wow.
Nice blog!
I guess 68k programmers forget about stack cleanup as its so simple on 68k given its massive number of spare address registers (compared to x86).
Indeed, I learned a ton there. It's just a pity that in one of the many server moves the comments were lost as the blog attracted a very interesting crowd and the comments were a treasure trove of information in themselves.
I guess 68k programmers forget about stack cleanup as its so simple on 68k given its massive number of spare address registers (compared to x86).
I think passing things on the stack is not really "a thing" on QDOS systems anyway, unless you write a compiler. I learned most about it when I started programming on other systems IIRC.
No need for conspiracy theories, API calling conventions are not a language and can be chosen independently. This is especially true if you own both the APIs and compilers, although I think major C compilers have been supporting multiple calling conventions for a long time.
Had the goal been to easily implement those old APIs in Pascal, they would probably have used Pascal-style strings (like MacOS has done for a long time) rather than C-style strings.
Last edited by M68008 on Sun Jan 30, 2022 10:04 pm, edited 1 time in total.
Last check-in 2 days ago Somebody is working on it, amazing. And all C... your argument that the Windows API at no point uses Pascal strings is a good one, too.
