All of lore.kernel.org
 help / color / mirror / Atom feed
* gpsd re-write
@ 2010-01-24 13:22 Michael Lippautz
  2010-01-24 15:09 ` Koen Kooi
  0 siblings, 1 reply; 2+ messages in thread
From: Michael Lippautz @ 2010-01-24 13:22 UTC (permalink / raw)
  To: openembedded-devel

Hej,

As I am re-writing the gpsd package, a few questions did come up.

1) gpsd can deal with multiple different gps devices (being served as
different nodes). Is it sufficient to start gpsd with a control socket
and add recipe for gps devices that just add init script that are
executed after gpsd startup. These init scripts could add the device
over the control socket. I.e. a recipe gpsd-device-... could add the
/dev/ttyACM0 node, using the control socket.

2) Since gps devices can also be integrated into platforms (machines)
a second approach using RDEPENDS_${PN}_machine can be used to build a
machine specific package. This machine specific package could init the
device and also attach it using the socket.

Is it "ok" to create such machine specific packages, i.e.
gpsd-hardware-om-gta01 ? How about "device" specific recipes (should
they be done in an other way)?

Additionally I can also add an /etc/default/gps-hardware file that
uses update-alternative to add the device at program startup.

Are these approaches correct? Don't want to sit and write these if
they are rejected anyways ;)

regards,
Michael



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: gpsd re-write
  2010-01-24 13:22 gpsd re-write Michael Lippautz
@ 2010-01-24 15:09 ` Koen Kooi
  0 siblings, 0 replies; 2+ messages in thread
From: Koen Kooi @ 2010-01-24 15:09 UTC (permalink / raw)
  To: openembedded-devel

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

On 24-01-10 14:22, Michael Lippautz wrote:
> Hej,
> 
> As I am re-writing the gpsd package, a few questions did come up.
> 
> 1) gpsd can deal with multiple different gps devices (being served as
> different nodes). Is it sufficient to start gpsd with a control socket
> and add recipe for gps devices that just add init script that are
> executed after gpsd startup. These init scripts could add the device
> over the control socket. I.e. a recipe gpsd-device-... could add the
> /dev/ttyACM0 node, using the control socket.
> 
> 2) Since gps devices can also be integrated into platforms (machines)
> a second approach using RDEPENDS_${PN}_machine can be used to build a
> machine specific package. This machine specific package could init the
> device and also attach it using the socket.
> 
> Is it "ok" to create such machine specific packages, i.e.
> gpsd-hardware-om-gta01 ? How about "device" specific recipes (should
> they be done in an other way)?
> 
> Additionally I can also add an /etc/default/gps-hardware file that
> uses update-alternative to add the device at program startup.
> 
> Are these approaches correct? Don't want to sit and write these if
> they are rejected anyways ;)

Your approaches are a lot better than the current 'gpsd only starts if
it was built for your machine *AND* OE has a config for it'

regards,

Koen
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFLXGK4MkyGM64RGpERAk7CAJ4kbdekGFkioFbpHZ/28fsmqcOqtACfQgx3
i5aHG9ULaUksACfO759u7sc=
=VAwM
-----END PGP SIGNATURE-----




^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2010-01-24 15:12 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-01-24 13:22 gpsd re-write Michael Lippautz
2010-01-24 15:09 ` Koen Kooi

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.