Daniel Marjamäki comments

Results 389 comments of


                                            Daniel Marjamäki

pass a char buffer to simplecpp instead of a stream

> Please let's take that with a grain of salt. It's the instruction count not a time estimate. some kind of small benchmark that measures actual time would be interesting.

pass a char buffer to simplecpp instead of a stream

> That just moves the potential mistakes to a different layer. Sorry I feel we talk about different things. I talked about type safety and you about bounds safety. So...

pass a char buffer to simplecpp instead of a stream

I just don't like this. A istream parameter is better than 2 `const uint8_t*, int size` parameters imo. An istream is also more explicit than a string_view or a span.

Added check for deletion of this

> Perhaps a different warning message could be given, then... something like "deletion of this is highly unusual and can lead to problems if not handled correctly"? No. Looking at...

avoid some (unnecessary) default parameters in GUI class constructors / added TODOs

> So there should be a noticeable difference in window behavior when passing a parent. Are those classes ever used as child windows? Spontanously I don't think so. We either...

allow CLICOLOR_FORCE env var to force CLI color output

hmm I am not sure if a unit test would be possible. not if getenv is cached. maybe you want to uncache it it testing somehow. or another alternative is...

allow CLICOLOR_FORCE env var to force CLI color output

I think your code and testing looks good. Adding the NO_COLOR sounds easy technically, I am in favor of adding it also :-). But I am not sure how you...

fixed #12715 - made Visual Studio conditions work again

:+1:

Make simplecpp friendlier toward 64-bit builds.

some CI warning

Make simplecpp friendlier toward 64-bit builds.

I am not a fan about using std::size_t for various variables. Maybe we can shut off the MSVC warnings globally instead.