gnustep-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: stripping makefiles


From: Dennis Leeuw
Subject: Re: stripping makefiles
Date: Mon, 13 Oct 2003 16:56:33 +0200
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.0) Gecko/20020623 Debian/1.0.0-0.woody.1

Two things:
1) Why should you also install the source. You wouldn't do that for every other "binary". One expects the compiled version to work correctly, so I would expect the gnustep makefile system to work.

2) However if the source to the makefile system is THAT important to a developer is the win in speed justified by stripping the comments?

Just my 2 cents.

Dennis

Nicola Pero wrote:
Between other new features and things I'm implementing in gnustep-make,
I've experimented with stripping the makefiles after installation.

'Stripping' means you delete all comments and empty lines - the resulting
makefiles execute faster because make has less data to read and parse.
How much faster depends on the hardware and caching systems available I
suppose, on my machine I get approximately a 5% faster gnustep-make, which
means around 1 second when running through my company's source tree (when
already built), which takes around 18 seconds at the moment.

Stripping does make makefiles difficult to read of course - my idea was to
install two copies of the makefiles - one stripped for execution, and the
other one unused but unstripped so that anyone can still read the source
code and hopefully understand how it works.

So my plan would be to make stripped makefiles the default, but still
include the whole unstripped makefiles in a separate directory (say,
Library/Makefiles/Source). But I don't have a strong opinion, as the performance gain is big. Comments ?



_______________________________________________
Gnustep-dev mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/gnustep-dev







reply via email to

[Prev in Thread] Current Thread [Next in Thread]