[Top][All Lists]

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

Re: problem: unitialized +=

From: Akim Demaille
Subject: Re: problem: unitialized +=
Date: 18 Apr 2001 10:56:10 +0200
User-agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.1 (Cuyahoga Valley)

>>>>> "Alexandre" == Alexandre Oliva <address@hidden> writes:

Alexandre> On Apr 18, 2001, Akim Demaille <address@hidden> wrote:
>> Unfortunately it is quite unrealistic wrt the implementation:
>> Automake values are read _after_ was.  Supporting this
>> would be hard, and I am not sure it would bring really bonus.

Alexandre> How about this: when you find FOO += bar but FOO is not
Alexandre> initialized, replace it with FOO = @FOO@ bar.  It would
Alexandre> probably cover the most interesting cases, no?  Hmm...  Not
Alexandre> the case of DEFS :-(

Alexandre> How about adding a mark, such as @+FOO@, that you'd replace
Alexandre> with the contents automake would set by default, when it
Alexandre> gets to the point of setting the defaults?

But then, what will you do when a variable is defined conditionally by
the user, and not by Automake?  And the converse?

It also means we will have to keep some form of compatibility with
*possibly modified* variables!

Really, not only would this be hard to implement, but the semantics in
itself is not clear.  I personally agree very much with the current
status of Automake: either you don't play with its variables and it
does the whole job, or you take the responsibility of maintaining
something by yourself, but then you take over the whole maintenance.
Anything in between is hair.

reply via email to

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