_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Pierre-Yves,
>> this indeed looks like a bug. Seems like the new name is not written
>> to disk. Can you re-produce this with bluez-utils-3.29?
>
> Ok, I'll upgrade this box to 3.29 and try to reproduce it. FYI this
> output comes from a box running v3.28.
my assumption is that it still is broken then. I don't recall changes
in that area.
Regards
Marcel
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel,
> this indeed looks like a bug. Seems like the new name is not written
> to disk. Can you re-produce this with bluez-utils-3.29?
Ok, I'll upgrade this box to 3.29 and try to reproduce it. FYI this
output comes from a box running v3.28.
Regards,
Pierre-Yves
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Pierre-Yves,
>>> It seems that once a friendly name has been cached to /var/lib/
>>> bluetooth/*/names, RemoteNameUpdated signals contain this cached
>>> name forever, and user is never informed of a name change unless
>>> those "names" files are deleted.
>>
>> the names are never updated once retrieved. At least not via an
>> inquiry run. If you connect to a remote device, it will update the
>> name. Another option is Extended Inquiry support if you have the
>> hardware for it.
>>
>> The reason behind this is that the remote name operation is too
>> expensive if you don't have an ACL link between two devices. And a
>> lot
>> of times the name doesn't change so often or it is unimportant
>> anyway.
>
> It makes a lot of sense, at least for the python example I submitted,
> which was only performing calls to "DiscoverDevices".
>
> But I've encountered another situation, where bluez' caching behaviour
> produces some nonsense. I have no simple test application to provide,
> but it is quite simple to explain and understand.
>
> I've an application which performs a continuous inquiry (eg when it
> gets
> DiscoveryCompleted, it calls DiscoverDevices() and so on). In the
> meantime, this application does also perform sdp extraction and rfcomm
> connections to the detected devices. Here is a record of all
> RemoteNameUpdated signals from a particular device:
>
> I delete /var/lib/bluetooth/*/names, set my device's name to "Name 1",
> and start my application
>
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
>
> So far so good. At this point I set my device's name to "Name2": I
> start
> to get mixed values (I understand that "Name1" are sent out from the
> discovery, and thus use the cached value as you explainded above, and
> "Name2" are from the other operations, which are actually getting the
> current name, because they do connect to the device)
>
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name1
>
> At this point I delete /var/lib/bluetooth/*/names. As cache is
> recreated
> with the current value (Name2) everything looks good again
>
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
>
> At this point, I change my device's name to "Name3". Things get
> mixed up
> again.
>
> RemoteNameUpdated signal received: 006057BB4D92 - Name3
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name3
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name3
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
> RemoteNameUpdated signal received: 006057BB4D92 - Name3
> RemoteNameUpdated signal received: 006057BB4D92 - Name2
>
> So, shouldn't the cached value get updated when bluez gets a different
> name later on?
this indeed looks like a bug. Seems like the new name is not written
to disk. Can you re-produce this with bluez-utils-3.29?
Regards
Marcel
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Marcel,
Thanks for your fast answer.
>> It seems that once a friendly name has been cached to /var/lib/
>> bluetooth/*/names, RemoteNameUpdated signals contain this cached
>> name forever, and user is never informed of a name change unless
>> those "names" files are deleted.
>
> the names are never updated once retrieved. At least not via an
> inquiry run. If you connect to a remote device, it will update the
> name. Another option is Extended Inquiry support if you have the
> hardware for it.
>
> The reason behind this is that the remote name operation is too
> expensive if you don't have an ACL link between two devices. And a lot
> of times the name doesn't change so often or it is unimportant anyway.
It makes a lot of sense, at least for the python example I submitted,
which was only performing calls to "DiscoverDevices".
But I've encountered another situation, where bluez' caching behaviour
produces some nonsense. I have no simple test application to provide,
but it is quite simple to explain and understand.
I've an application which performs a continuous inquiry (eg when it gets
DiscoveryCompleted, it calls DiscoverDevices() and so on). In the
meantime, this application does also perform sdp extraction and rfcomm
connections to the detected devices. Here is a record of all
RemoteNameUpdated signals from a particular device:
I delete /var/lib/bluetooth/*/names, set my device's name to "Name 1",
and start my application
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
So far so good. At this point I set my device's name to "Name2": I start
to get mixed values (I understand that "Name1" are sent out from the
discovery, and thus use the cached value as you explainded above, and
"Name2" are from the other operations, which are actually getting the
current name, because they do connect to the device)
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name1
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name1
At this point I delete /var/lib/bluetooth/*/names. As cache is recreated
with the current value (Name2) everything looks good again
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
At this point, I change my device's name to "Name3". Things get mixed up
again.
RemoteNameUpdated signal received: 006057BB4D92 - Name3
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name3
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name3
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name2
RemoteNameUpdated signal received: 006057BB4D92 - Name3
RemoteNameUpdated signal received: 006057BB4D92 - Name2
So, shouldn't the cached value get updated when bluez gets a different
name later on?
Best Regards,
Pierre-Yves
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel
Hi Pierre-Yves,
> It seems that once a friendly name has been cached to /var/lib/
> bluetooth/*/names, RemoteNameUpdated signals contain this cached
> name forever, and user is never informed of a name change unless
> those "names" files are deleted.
the names are never updated once retrieved. At least not via an
inquiry run. If you connect to a remote device, it will update the
name. Another option is Extended Inquiry support if you have the
hardware for it.
The reason behind this is that the remote name operation is too
expensive if you don't have an ACL link between two devices. And a lot
of times the name doesn't change so often or it is unimportant anyway.
Regards
Marcel
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Bluez-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-devel