Bert Frees
Bert Frees
> Edit: obviously for the tests it is not evil. Right, was just about to say that. Maybe not statically link the whole library, but maybe we could have a...
I don't expect they do. brl_checks has a similar role but only for test related utility functions (and had to be moved to tools later because of lou_checkyaml). The new...
But how is it done on *nix? I'd like to understand too.
Before you go and add more functions to the API, I'd like to know exactly which functions are we talking about?
> Maybe not statically link the whole library, but maybe we could have a small common utility library that we could statically link with the tools. I guess in the...
> - showString > - findOpcodeNumber These are not used in the tools as far as I can tell. I wouldn't redefine any functions (e.g. charHash) in the tools (no...
I thought so :) For symmetry is fine.
No I was saying the whole effort of separating the functions in their own file would be in vain because they will still be visible to the outside on Unix....
> So what is the Unix criterion for making a function visible? That is, is everything visible, or only those things declared in headers? I'm no expert, but as far...
> I'll give it a try though. OK but please don't put too much of your time in this. As Christian said all this tweaking has little gain. > If...