On Jul 30, 2008, at 6:50 AM, Arkadiusz Miskiewicz wrote:

> On Wednesday 30 July 2008, Jeff Johnson wrote:
>> On Jul 30, 2008, at 5:50 AM, Ralf S. Engelschall wrote:
>
>> Instead of arguing abt <stdint.h> (which I will start ripping after
>> my 2nd cup of coffee
>> today, sigh), let's try the forward looking goal.
>
> Ok, so the agreement is that rpmcli.h should be self consistent  
> right? (which
> means stdint.h inclusion will end in this file) We don't care about  
> other rpm
> headers consistency then.
>
> Can these other headers be protected like this:
>
> #if !defined(H_RPMCLI) && !defined(RPM_BUILD_ITSELF...)
> #error Please include rpmcli.h instead of this file.
> #endif
>
> ?
>

I don't believe in NAG ware. It's not up to rpm to teach programming.

Nor are messages in include files the most effective training,  
example code,
or documentation, or man pages are far more effective.

rpm != glibc

> Or leave public headers (rpmcli.h) in /usr/include/rpm and move all  
> other "not
> to be directly included by user" header files to /usr/include/rpm/ 
> internal/ ?
>

I originally exposed all include files used by rpm in the hopes that
there were intelligent programmers around who might find the additional
includes useful.

I was wrong. All that has happened is that expectations have gotten way
out of line with reality.

I will start reducing the supported API in rpmlib in order to  
minimize the maintenance
accordingly.

E.g. there's literally no reason why rpm needs to carry <argv.h> out  
of tree.

Any routine that is currently using ARGV_t will be reverted to "const  
char **"
and I will make <argv.h> private thereafter.

73 de Jeff