[rrd-developers] RRDCacheD - Client rewriting path

[Tobias Oetiker]

Today Florian Forster wrote:

> Hi Tobi,
>
> On Tue, Aug 11, 2009 at 07:48:30AM +0200, Tobias Oetiker wrote:
> > >   * Some magical version of the functions will check some sort of
> > >     environment and call the appropriate function.
> >
> > this is to maintain the present interface ?
>
> well, I would much rather see the end of that argv/argc style interface
> and have all functions use an interface in the style of the `*_r'
> functions. But as I said: That's my ideal dreamworld.
>
> Maybe a possible transition would be to break ABI compatibility but
> provide an optional legacy API. I. e. rename `rrd_update' and friends to
> `rrd_update_legacy' (this brakes ABI compatibility). Then provide
> something like this:
>
>   /* declaration of rrd_update_local and rrd_update_remote */
>   #if RRD_LEGACY_INTERFACE
>   /* declaration of rrd_update_legacy */
>   # define rrd_update   rrd_update_legacy
>   # define rrd_update_r rrd_update_local
>   # define rrdc_update  rrd_update_remote
>   ...
>   #else
>   /* declaration of rrd_update */
>   #endif

that does sound reasonable ... I think it is important to allow for
a backward compatible interface ...

> > I guess its time to finally push out 1.4 so that the code can move
> > again ...
>
> I'd be glad if you could consider my per-socket authorization patch for
> the 1.4 release. I think it's a big improvement over the ``restricted
> sockets'' (the `-L' command line option). I'll send an updates patch
> with docs in a minute..

to me this patch looks good ... as you know the whole inet socket
interface without authentication is a thorn in my side. your patch
makes this a bit better ...

I'll wait to hear what kevin has to say on the patch.

cheers
tobi
>
> Regards,
> -octo
>

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
http://it.oetiker.ch tobi at oetiker.ch ++41 62 775 9902 / sb: -9900