I'm afraid that one even leads to a dead link, server can't be found. Besides, the example code from Eberhard has no relevance to DeviceMap or the W3C DDR Standard I could spot in those sources. Analysing UAAgent, sure, that is what a few other systems like 52Mobi do, but those are not W3C DDR compliant either.
For the talk on Wed I prefer to focus on those and DeviceMap. Any actual usage of its API (or where it's necessary the old OpenDDR precursors, the compatibility of these, at least where both cover the DDR standard is also an advantage I'll point out compared to proprietary and often expensive closed-source products like WURFL[?]) or benchmarks around it aare highly appreciated in addition to what I already presented before. Thanks, Werner On Sat, Mar 22, 2014 at 10:58 PM, Reza <[email protected]> wrote: > I just spawned this as an example. > > Human and 3rd party calls: > > http://www.rezsoft.org:9000/uadiag > > > > http://www.rezsoft.org:9000/uadiag?ua=Mozilla_5.0_iPhone_U_CPU_iPhone_OS_5_0_like_Mac_OS_X_en-us_AppleWebKit_528.18_KHTML_like_Gecko_Version/4.0_Mobile_7A341_Safari_528.16 > > > (note i dont have the url decode plugin available at the moment, so im > using _ ) > > Client calls: > > http://www.rezsoft.org:9000/uadiag.js > > > http://www.rezsoft.org:9000/uadiag.js?callback=jsonp > > > > ________________________________ > From: Reza <[email protected]> > To: "[email protected]" < > [email protected]> > Sent: Saturday, March 22, 2014 5:44 PM > Subject: Re: "Bad Request" > > > So there are several use cases for a DDR service. I think you kind of miss > both of them... :[ > > 1) A client side (javascript) call is made to the DDR service by the > actual client. In this case, the User-Agent field is correct. > > 2) A server side (code) or browser call (human) is made to the service to > get the user-agent. In this case, the user-agent should be passed as a URL > parameter, not a header. This is done for several reasons, most importantly > caching and for ease of use if using a browser. > > This is a very cryptic example, but if you look at the dClass varnish > config, it supports 3 methods: > > 1) > https://github.com/TheWeatherChannel/dClass/blob/master/servers/varnish/dclass.vcl#L106 > 2) > https://github.com/TheWeatherChannel/dClass/blob/master/servers/varnish/dclass.vcl#L70 > > For option 1 it can return raw JSON or Javascript/JSONP. > > > ________________________________ > > From: eberhard speer jr. <[email protected]> > To: [email protected] > Sent: Saturday, March 22, 2014 5:35 PM > Subject: "Bad Request" > > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi, > > yes, using the API URL in a browser would cause "Bad Request" because > the API requires the extra header : Ddr-User-Agent, as described in > the documentation. > The idea is that the API responds to a request to resolve "a" > user-agent string supplied in parameter [the Ddr-User-Agent http > header] and not the user-agent string of the device calling the API ;-) > > [on Skype now] > > esjr > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.8 (MingW32) > Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ > > iQEcBAEBAgAGBQJTLgIIAAoJEOxywXcFLKYcuscH/0YyJ6/hz/HWge+bVT6uPUmN > G+aYOXc9GKZqtMqMISOahVw024ZzMBjCP6XspdNycvQJ2UpxHCvU8oa2GGfvr14U > BW8vRxUpxP67aX5EzUdH972wPNDJriNfNiqd6w7BCjkuQpuCe7qpCpubm0ZzY5/0 > QpB1hbRbkLELhuPHaTz/x7Y0gaKGHwYo/E5oFi7giFW1ttivE7oqZ8HyMviGKANb > WPlOM+mve9PkfzQwc0frc9y9N2Wb7DHM/rTk7rlbGtpLqJbG/trrurcXQKCj9pGi > 4NlXuBqPjszjtU8HhAz/s7B2TfsLYEwmhsVz26S9QlGdlamg7ESWAYCtB28L7QI= > =RJSQ > -----END PGP SIGNATURE----- >
