Michael Viking said:
The compiler in VC6 is fine. It's not optimal, it's not best, but it's
fine.
That depends on what you do. It's almost ten years old
and disregards standard C++ in many critical areas. IME,
as soon as the keyword 'template' appeared somewhere,
VC6 caused trouble.
The proof is all the applications that were developed with it.
LOL! This is "proof" that (ancient) BASIC was a fine
language.
[...]
(You avoided adressing the other two points.)
I'm not sure which points I missed, so I'll try to address each one. We
don't use Boost. Some do, but I'd guess that most don't?
I have no idea why one wouldn't want to use a box
full of tools as boost is. But even if -- it didn't
take boost to make VC6 choke. It couldn't even keep
up with /my/ template code (and that was long before
I got my nose into template meta stuff).
It prevented us from doing things in better, safer,
and more efficient ways.
We use very
little of the std lib, and it's not hard to repair the issues we come
across, since they're just header files. [...]
Oh boy.
It's my understanding that quite a few people got burned here trying to use
C++/CLI or whatever it was.
<shrug>
I wouldn't know.
Still, this has nothing to do with any of my points.
[...]
Right. And where the point inbetween is, that is a
personal preference. So yours isn't better or worse
just because you feel it's right.
For me using libraries like boost and applying some
modern template techniques (that shift work to compile-
time, thus making the code faster at run-time and
making more fail at compile-time instead of run-time)
is a necessary abstraction, and dealing with raw
pointers etc. is too close to the metal and a last
century's way of doing things.
These are valid. For the size of our dev team and what we need to do, we
don't feel moving to the compiler, making these changes, etc. is a tradeoff
we want to make. If you have the time and resources, that's great.
There's about half a dozen developers working on
the projects I work on. Not all of them work on
Windows, though. As a small company that ships
software to some very big companies, we know how
brutal deadlines can be.
We see it the other way around: We can't afford
to not to use technologies that save time and
make our code more stable.
[...]
And from what I've seen of people we've
interviewed, the ones who could write in C and assembler and understand that
Win SDK and know what they're doing are a heck of a lot better than the
people who can put together some sort of UI with the latest .NET and that's
all they know. Reality is these people don't know much of anything, and if
they do have some kind of problem, it's abstracted away to where they can't
understand or debug it.
In my youth[TM] I heard the same about people not doing
assembler.
I honestly don't know what you're saying here. [...]
When I was young, I was told that those who only
write code in high-level languages and don't do
assembler don't know much of anything. Since most
applications today are not written in assembler,
but in some higher-level language, and since that
wouldn't be possible if the overwhelming majority
of those making them where morons, I have to
conclude that this statement was wrong.
Now you say those who only write .NET stuff and
don't know the much about the Win32 API don't know
much of anything. I compare that statement with my
experience with similar statements and don't see
much value in it.
(BTW, while I write most of my code on Windows, 95%
of the code I write nowadays runs on quite a few
platforms. Therefor I write plain and portable std
C++ and almost never use the Win32 API at all --
and thus know very little about it. Your above
statement paints me as not knowing much of anything.)
Schobi
--
(e-mail address removed) is never read
I'm HSchober at gmx dot de
"If there were some arcane way to remove the heads of every
newsgroup troll on the planet, I think it would elevate
humans to a whole new level of intelligence."
Rocky Frisco
