Experimental Physics and
Industrial Control System

Andrew Johnson <[email protected]> · Fri, 10 Dec 2010 16:11:14 -0600

On Friday 10 December 2010 15:19:12 Davidsaver, Michael wrote:
> > My
> > worry is that the libCom/Makefile is unwieldy because it has to name
> > all of the files in the different sub-directories.
>
> As long as the EPICS build system forces the distinction between
> branch/leaf node we will have this problem.

It does, but I don't see why that causes the Makefile to be unwieldy.

> The Makefiles will be large, but there will be fewer of them.  Basically
> all of my dislike of the current build system come from the "evil"
> include statement, and having to hunt to find anything.  It may be a
> personal preference, but in the src/ directory at least I would rather
> have a few large Makefiles.

I've just pushed a partial solution to LP which splits the libCom/Makefile 
into a series of Makefile and RULES fragments in the various source sub-
directories.  See lp:~anj/epics-base/split-libCom-Makefile for the changes 
(which I will not merge without further discussion and agreement).  I suspect 
that you're not going to like it though because it uses include statements to 
pull those fragments into the libCom/Makefile.  Personally I think this makes 
it much easier to find the build instructions associated with a source file 
and vice versa.

If we used my approach with your structure, we could maybe end up with build 
directories (O.<target>) and "master" Makefiles in libCom, ioc, std and 
template (like you I'm not sure about ca).  Each source sub-directory would 
still have a Makefile in it, and some would have RULES files as well.  The 
disadvantage is that we would lose the ability to type 'make' in any source 
directory and have it rebuild just the related libraries and products.

- Andrew
-- 
If a man is offered a fact which goes against his instincts, he will
scrutinize it closely, and unless the evidence is overwhelming, he will
refuse to believe it.  If, on the other hand, he is offered something
which affords a reason for acting in accordance to his instincts, he
will accept it even on the slightest evidence.  -- Bertrand Russell

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System