Re: [HACKERS] Configure problem (and fix).

Поиск
Список
Период
Сортировка
От Billy G. Allie
Тема Re: [HACKERS] Configure problem (and fix).
Дата
Msg-id 199810250032.UAA16592@bgalli.mug.org
обсуждение исходный текст
Список pgsql-hackers
Brook Milligan wrote:
> 
> The normal usage is fine for things like Makefiles that know about
> environment variables and do their own expansion (i.e., the vast
> majority of the usage of these variables).  In this one instance we
> are constructing *.sql files that can't expand environment variables
> and hence need to do that in advance during configuration.
> 
> It seems much cleaner to do the following:
> 
> - keep the unexpanded versions in the Makefiles (since it reduces
>   cluttered duplication and uses variables as they should be used for
>   indirect references), and
> 
> - introduce a new expanded variable for this purpose

Is it cleaner?  It makes sense to me that if I want to use 'libdir' or any of 
the other defined variables, it should have a value that can be anywhere it is 
needed without having to depend on the expansion of variables in the place 
where it is used.  Configure knows the information needed to expand the 
variables and should do so before making subsitutions.  If this is not done, 
it leave the door open for problems to occur.  For instance, if after running 
configure, the global makefile is modified so that 'prefix' is changed (an 
assumption can be made that since the definitions have unexpanded variables, 
then this can work without having to re-run configure), then the location of 
'libdir' will be differenct from 'expanded_libdir' and the *.sql files will no 
longer be correct.  The assumption that a variable, such as 'prefix', can be 
changed in the global makefile (an assumption encouraged by the way configure 
works without my patch) and still have the system build correctly is incorrect.

> 
> The patches below replace Billie's and accomplish his intent of having
BTW: That's Billy's ----------^

My intent is to be able to use the subsitution feature of configure any where 
it is needed without haveing to depend on expansion of embeded variable 
occuring where they are used.

-- 
____       | Billy G. Allie    | Domain....: Bill.Allie@mug.org
|  /|      | 7436 Hartwell     | Compuserve: 76337,2061
|-/-|----- | Dearborn, MI 48126| MSN.......: B_G_Allie@email.msn.com
|/  |LLIE  | (313) 582-1540    | 




В списке pgsql-hackers по дате отправления:

Предыдущее
От: Bruce Momjian
Дата:
Сообщение: Re: [HACKERS] Features list
Следующее
От: Bruce Momjian
Дата:
Сообщение: Re: [HACKERS] Features list