News for hackers. Stuff that matters.

A comment on Reflection for C++26. Parent comment

Re:Reflection for C++26 by npoc on Tuesday July 02, 2024 @03:34AM (#40854358)
Just because it's newer doesn't make it better. There are good reasons for avoiding iostream
[ Parent | on HN ]
Re:Reflection for C++26 by tsimionescu on Tuesday July 02, 2024 @04:52AM (#40854750)
That's my point - iostream is a really bad piece of code, and if you're anyway going to use modern C++, it's really recommended to stop using it.
[ Parent | on HN ]
Re:Reflection for C++26 by npoc on Tuesday July 02, 2024 @05:37PM (#40860669)
Sorry, I got your comment completely backwards
[ Parent | on HN ]
Re:Reflection for C++26 by pjmlp on Wednesday July 03, 2024 @05:53AM (#40864401)
iostream is good enough for most jobs, unless one is writing high performace IO code battling for each ms.
[ Parent | on HN ]
Re:Reflection for C++26 by tsimionescu on Wednesday July 03, 2024 @10:42AM (#40866381)
Or unless one wants to write formatted output, or unless one wants to handle IO errors with RAII...
[ Parent | on HN ]
Re:Reflection for C++26 by pjmlp on Wednesday July 03, 2024 @10:47AM (#40866441)
Perfectly fine with existing operators and handle classes.

Happily using iostreams since Turbo C++ 1.0 for MS-DOS in 1993, and will keep doing so, unless chasing ms optimizations.

[ Parent | on HN ]
Re:Reflection for C++26 by spacechild1 on Tuesday July 02, 2024 @05:03AM (#40854788)
> There are good reasons for avoiding iostream

I guess that's not what you wanted to say, but I fully agree :)

[ Parent | on HN ]