-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

El Tue, 18 Aug 2009 08:33:21 +0000
Eric MSP Veith <[email protected]> escribió:
> On Tue, Aug 18, 2009 at 05:18:03AM -0300, Ismael Luceno wrote:
> > El Mon, 17 Aug 2009 14:33:42 +0400
> > I don't care much about that because we can't support those
> > restricted environments anyway.
> > <...> 
> > But there might be problems related to message queues that I'm not
> > taking into account...
> 
> It's obviously an elegant feature, that's true. Using two message
> queues will give us bi-directional communication. We can do
> multiplexing, if we want to, using IDs or something like that.
> 
+1, we can use transaction IDs.

> Whether you want to support embedded devices or not is mostly not a
> decision based on technology, but on our/your preferences. I,
> personnally, think we shouldn't knowingly place obstacles in the path
> of those who want to use InitNG on an embedded devices, but that's
> 100% up to you. :-)

I don't see POSIX mq as an obstacle. Probably a slimmer implementation
cloud be made if needed, and we can even ignore priorities, but it's
just a few kb...

> The only thing that I fear we loose is some bit of backwards
> compatibility with the old init, i.e. the possiblity to control
> InitNG through the old "shutdown" tool. That's all.

We already are incompatible. In fact, this can be an excuse to
implement a better emulation layer ;).

- -- 
Ismael Luceno
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkqKdWEACgkQ/mxY0+yOXJolPACdGapSlMlxwSj5l+wOOmfQw1LD
cUkAoNr3YEgzj9G7x2PNCc1ugbIy53+4
=HfYw
-----END PGP SIGNATURE-----
--
_______________________________________________
Initng mailing list
[email protected]
http://jw.dyndns.org/mailman/listinfo/initng

Reply via email to