On 5/29/2026 2:52 AM, Janis Papanagnou wrote:If you're doing your own personal stuff that doesn't surprise me.On 2026-05-28 11:57, BGB wrote:Possibly.On 5/28/2026 2:18 AM, Janis Papanagnou wrote:[...]On 2026-05-28 01:49, BGB wrote:>I'm a big fan of abstractions. - So many things beyond "C" are fine!>
I am not saying that abstractions are bad, but I haven't usually found them to be worth the costs IME.
Wow! - That's completely different from my experience and practice.
>
It's what makes usage simple, fast, reliable. Not wasting time for
details, or fixing technical bugs that should be prevented by the
language.
But, it is also possible I approach programming in a different way.
Well, you mentioned gcc now repeatedly. Myself I had been using gcc[...]Possibly, a lot could depend on how one is counting things as well.
In a lot of cases when using GCC, I end up using:
-ffunction-sections -fdata-sections -Wl,-gc-sections
mainly in private context only. Professionally we (mostly) used the
tools that came with the vendors on commercial (often Unix) systems.
I really cannot tell about Windows environments, Cygwin, or using
(maybe older?) gcc versions for large scale software development.
For newer systems and versions I'd certainly not expect problems
as you seem to have encountered.
[...]Thanks for explaining your view on "syntactic sugar".[...]To be excluded from being syntactic sugar, it needs to be something that is not generally possible to express within the base language.
So, for example:
Things like operator overloading or classes are syntactic sugar IMO, as what they do can be expressed in C, even if a lot less pretty (or far from an idiomatic style).
I would not consider exceptions or RTTI as syntactic sugar, because these involve things that do not map to native C.
Using longjmp, pointer-tagging, etc, could be considered as analogous, but not functionally equivalent, to what C++ is doing in these cases.
[...]We've been talking about C++; for constants I regularly use constants.
We avoided macros if possible.They are de-facto for constants and similar, but for longer stuff is better avoided.
In "C", frankly, it appears to me less important; I used to use Cpp
names for constant items or expressions, mainly because that's "how
it is done in C". - OTOH, if I can give an item a type, why should
I use a primitive mechanism. Because it's faster? No memory wasted?
In our C++ projects we used Cpp features certainly not for constant
literals. We used them for tags (prevent double inclusion of h-files),
conditionals, and some such. Rarely for more tricky things involving
tasks to circumvent limitations of the "C" language in cases where we
wrote code that should work with both of these languages.
Janis
[...]
Les messages affichés proviennent d'usenet.