On 14/01/2024 01:36, bart wrote:
People have suggested using make for everything, from hello.c up to
JP's and SL's massive applications.
While it's literally not wrong when you formulate it as "massive",
that's not the central point. The insight (for me) is another...
My take on that is based on the observation, now a bit abstract,
that we have processes that transform some input to some output...
input --> [ process ] --> output
...and the Makefile (syntax, and operational semantics) reflects
this relation and operation (formally quite straightforward) as
output: input
process
(And as noted elsethread already, this is even a generic view,
not restricted to compiling programming languages.)
There may be more or less input and output entities, and more or
less complex processes, and a lot of interconnected processes. But
it scales, both, the process view and the Makefile organization.
Whether it's trivial in one case, or "complex" ("massive") in other
cases doesn't change the coherent view on the topic, and also the
systematic approach (here [for example]: Makefiles) to tackle it.
A coherent systematic approach has (in my experience) advantages
(per se!), and more so if compared to isolated ad hoc handling of
such transformations. It's independent of and with respect to the
scaling.
(Note: There are of course also other things to consider that are
not covered in this post; like, e.g. whether it is more sensible
to use an available standard tool or method, or to cobble together
ones own solutions. The latter decision lies in the responsibility
of the person who is responsible for the project.)
Janis
PS: Bart, note that I read any responses from you only if others
find them valuable and answer them with quotes of your text. So
don't expect answers from me any more.
It would also be desirable if you'd open an own thread (also with
an appropriate subject) to discuss your themes. - Thanks.
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)