On Mon, Aug 29, 2011 at 12:25 PM, Patrick Ohly <[email protected]> wrote:
> On Mo, 2011-08-29 at 11:50 +0200, Chris Kühl wrote:
>> 2011/8/26 Patrick Ohly <[email protected]>:
>> > On Mi, 2011-08-24 at 14:46 +0200, Chris Kühl wrote:
>> >> The bluetooth-device-id-merge-prep is ready to be reviewed now.
>> >
>> > I've merged the D-Bus server reorganization, dbus-test and
>> > bluetooth-device-id-merge-prep branches. I fixed some minor problems
>> > found with -Werror -Wall.
>>
>> Great!
>>
>> >
>> > Attached an updated revision of the bluetooth-device-id-inspector.py
>> > script. Should we maintain it in SyncEvolution?
>> >
>>
>> Makes sense. Should it go in test or should we create a
>> tools/utilities directory?
>
> At some point we should, but then we would also have to move some of the
> other scripts. Let's add it to "test" to keep it simple for the time
> being.
>
>> > Chris, I suspect that I gave you wrong guidance about the use use of
>> > deviceName/peerName, or we have to extend the semantic. According to the
>> > D-Bus API docs:
>> >
>> >      * deviceName: device template: the device name that the template
>> >        is for (copied verbatim from that device)
>> >      * templateName: device template: string identifying the class of
>> >        devices the templates works for, like "Nokia S40"; meant to be
>> >        shown to users; optional, fall back to first entry in
>> >        fingerPrint if not set
>> >
>> > Now the syncevo-dbus-server sends deviceName=Nokia, templateName=Nokia
>> > and peerName=Nokia N97 mini.
>> >
>> > I think it should send:
>> >      * deviceName: the vendor/model information if available, otherwise
>> >        the user-configurable device name (can't be empty, because
>> >        traditionally it wasn't describe as optional)
>>
>> Right, I can make this change. But when you say "vendor/model is
>> available" you mean when the product and not *just* the vendor is in
>> the lookup table, right? That would make sense to me.
>
> IMHO it would make sense also if just the vendor is available. That
> information is not provided to the UI otherwise.
>
> But see my later email: after some more thinking I came to the
> conclusion that "deviceName" should stay as it is (name chosen by the
> user) and the new information should go into "templateName".

Did you mean peerName instead of templateName?

>From the what you are saying here...

Patrick Ohly <[email protected]> wrote:
>> The UI should be adapted to use peerName instead of deviceName, if
>> peerName is available. That way the user would see his chosen name
>> instead of the vendor/model name, which will not be unique in the
>> (unlikely) case that the user has more than one. Not a big deal.
>>
>> We might have the inverse situation, too: multiple different devices all
>> called "My Phone". I think users should (and can) avoid this, so we
>> should continue to display only the chosen name instead of adding the
>> vendor/model information - right?
>
> Given that the UI should only display the chosen name, and expects it in
> "deviceName", perhaps we should keep the traditional D-Bus API semantic
> unchanged and put the Device ID profile information into the "peerName"?
>
> That is a bit backwards (peerName is used for user-configurable strings
> elsewhere), but has the advantage that no changes will be needed in the
> D-Bus clients to get the desired behavior.
>

... I understand the following.

1) In the case that we have no PnPInformation info we have...

deviceName = User-modifiable name
peerName = User-modifiable name

2) In the case that we have PnPInformation info but no product match
in the lookup table  we have...

deviceName = User-modifiable name
peerName = Vendor found in Lookup table

3) In the case that we have PnPInformation info and a product match in
the lookup table we have...

deviceName = User-modifiable name
peerName = Vendor + " " + Product found in Lookup table

Is this right?

Also, the templateName was and is simply the "templateName" field from
the template. The code is...

string TemplateConfig::getTemplateName() {
    return m_metaProps["templateName"];
}

Should this be different? We could create a different template for
each class of phone.

Cheers,
Chris
_______________________________________________
SyncEvolution mailing list
[email protected]
http://lists.syncevolution.org/listinfo/syncevolution

Reply via email to