[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Proposals for new features.
From: |
R. Bernstein |
Subject: |
Re: Proposals for new features. |
Date: |
Tue, 4 Sep 2007 22:56:49 -0400 |
Paul Smith writes:
...
>
> Note that some of the changes you describe above (such as changing the
> code to require ISO C89 compilers, adding prototypes, etc.) have already
> been done in the code base. I declared a while ago that 3.81 was the
> last release that would compile with K&R C compilers. However, I still
> won't use ISO C99 constructs in GNU make. Maybe in a release or two.
Here are a couple purely structural things that could be done even in
K&R Compilers:
* extern declarations for foo.c go in foo.h, not in the global make.h
* interface documentation such as via Doxygen.
Feel free to cull from remake for this purely mechanical/structural
(and modularity) kind of change.
> The other question is how much size will this add to the package, etc.
> I don't want to greatly increase the size of the code, the amount of
> operating system prerequisites, the size of the documentation, etc.
> This can easily be solved by making the debugger build something you
> enable at configure time and keeping the debugger manual in a separate
> doc (this is probably a good idea anyway).
Sure. remake's autoconf/configure already checks for GNU Readline and
disables the debugger portion if that's not around.
> Or, a third alternative is to just ensure the right hooks/etc. are added
> to GNU make and keep the debugger as a separate package.
I've long been a fan of this approach. IBM Mainframe architecture had
what was called a subcommand interface which allowed for external
editors and programming languages. I believe Jeff Korn wrote a Ph.d
thesis on a general debugger interface. What I did was more for
expedience than elegance.
Re: Proposals for new features., Paul Smith, 2007/09/04
Message not available
- Message not available
- Message not available
- Re: Proposals for new features.,
R. Bernstein <=