help-make
[Top][All Lists]
Advanced

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

Re: Deferred expansion and computed rules and rant about make bug


From: DaleWKing
Subject: Re: Deferred expansion and computed rules and rant about make bug
Date: Fri, 05 Aug 2005 17:49:27 +0000

> %% Dale King <address@hidden> writes:
> 
>   dk> So I am looking for another workaround to the issue. I thought
>   dk> perhaps target-specific variables might do it, but haven't gotten
>   dk> it to work.
> 
>  $(target-$(foo)): FLAGS := $(flags-$(foo))
>  $(target-$(foo)):
>      @echo $(FLAGS)

Thank you. Yes, I later figured it out myself and got such a scheme to work for
binding the variables. I couldn't reply and say that right away.

However the problem with eval is still a problem for me. Since I am doing a
non-recursive make I want to do the following:

  $(foreach DIR,$(DIRS),$(eval include $(SUBDIR_MAKEFILE)))

Which of course gives me the virtual memory exhausted bug.

>   dk> Sure its fixed in CVS, but how does that solve the problem. Do you
>   dk> really expect everyone to go download the latest unreleased
>   dk> source, and rebuild it for their system?! Note in my case this is
>   dk> on a team of about 20 developers and management would not look
>   dk> favorably on using a patched make tool for doing builds.
> 
> ftp://alpha.gnu.org/gnu/make/make-3.81beta3.tar.gz

Which is meaningless. Where is the released version? We are using make for
building real-world commercial software. You expect us to get management buy-in
on using a beta to build the software? You expect each developer to have to
install the patched make to build the software?

This is a problem that is widespread in the Linux and open-source community.
Most people are only going to use a released version. Expecting end users to
install patched versions and hack their kernel, etc. is just not realistic.

I don't want to knock you personally, but not doing a release to fix bugs for
almost 3 years is inexcusable.

Imagine if someone like Micro$oft did something like that. Sure, they have bugs
but they have an update mechanism where you can get updates to fix the bugs
easily and automatically. The only somewhat automated update process in the free
software world is doing releases. Then people like RedHat or in our case Cygwin
will eventually see the changes and update their distros which users can then
get when they do an update.







reply via email to

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