P.s.: As Qualcomm MSM8960T Pro Snapdragon, Dual-core 1.7 GHz, Adreno 320 for the mentioned BlackBerry 10 model shows, this repository also maintains the chip architecture for many devices[?]
On Fri, Aug 1, 2014 at 4:10 PM, Werner Keil <[email protected]> wrote: > Hi, > > DeviceAtlas requires you to register/login for a UA probe (feel free to > test it if you want) but the list of properties is rather transparent: > https://deviceatlas.com/device-data/properties > > Note, somewhere along the line of Eberhard's proposal it calls such > boolean flag > isRobot > > "vendor", "model", etc. follow W3C DDR, while others mainly follow a > CamelCase pattern here. > It seems better to stay with some of the existing OpenDDR (OMA-inspired) > ones for extended properties, otherwise we might risk scrutiny from some of > the established providers. > > It's done regarding properties defined by OMA or see the underscore > "mania" of this Australian commercial player: > https://www.handsetdetection.com/properties/vendormodel/BlackBerry/Z30 > > Amazingly, HandsetDetection likely collaborated with OMA and others, if > you browse: > > https://www.handsetdetection.com/properties/vendormodel/BlackBerry/Z30#legacy > > You'll find all the "legacy" properties used by the likes of WURFL, there > at HandsetDetection they had them at least since 2008 or earlier, so when > WURFL was also still openly developed. > > This vendor probably changed its internal repository structure, but still > maintains a "v2_legacy" compatible version or does XSLT transformation to > those on the fly... > > I guess we are free to adjust the structure where applicable, but ideally > the core W3C DDR vobabulary should be maintained, OpenDDR backward > compatibility could also be a good thing, see the clients Reza mentioned > and other clients to OpenDDR like all those you find on GitHub, etc. > > Werner >
