- From: Luca Passani <luca.passani@openwave.com>
- Date: Wed, 27 Jul 2005 12:31:49 +0200
- To: <public-ddwg@w3.org>, <www-mobile@w3.org>, <www-mobile-request@w3.org>
> You may not be aware of situations where laptops and PDAs use > wireless data but we are aware of plenty of developers > who worry about such things. I am sure there are cases in vertical applications, but hardly in horizontal applications for the wireless market. Connection speeds are typically an implicit parameter of the application. For vertical applications that may require this knowledge, the problem can typically be solved with a tiny page that looks like: Choose connection speed: 1 (GSM) 9.6 kbps 2 (GPRS)less than 30 kbps 3 High (ISDN, ADSL, LAN) 4 Not sure, please test > You mention WAP markups. WAP is not restricted to WML as you >are probably aware but includes XHTML profile support. I am perfectly aware of that. In fact I also invented a very popular solution to the problem: http://wurfl.sourceforge.net/java/tutorial.php >. Laptops and PDAs also use such >profiles and developers may wish to provide content targetted at these >markup languages and devices and device capabilities knowledge is >important to have optimal matching of content to device. exactly. Have you checked out WURFL? we have addressed and solved all of this. >To me HTTP headers giving an indication of device capabilities, e.g. >providing device type, URLs to appropriate profiles such as UAPROF etc. is >a much more reliable way to achieve device capability It's not! this model has been failing egregiously along the way. I lost track of how many developers have been complaining of how bad UAProf is, yet I have *bever* heard a developer praise UAProf. Even one of the main UAProf supporters criticizes UAProf and CC/PP's shortcomings: http://www.hpl.hp.com/personal/marbut/someQuestionsOnCCPP.htm face reality, guys. Device profiling is too important to be left in the hands of device manufacturers. Luca -----Original Message----- From: Alastair Angwin [mailto:alastair_angwin@uk.ibm.com] Sent: 27 July 2005 10:56 To: Luca Passani Cc: public-ddwg@w3.org; www-mobile@w3.org; www-mobile-request@w3.org Subject: RE: Mobile phone capabilities list? Luca You may not be aware of situations where laptops and PDAs use wireless data but we are aware of plenty of developers who worry about such things. You mention WAP markups. WAP is not restricted to WML as you are probably aware but includes XHTML profile support. Laptops and PDAs also use such profiles and developers may wish to provide content targetted at these markup languages and devices and device capabilities knowledge is important to have optimal matching of content to device. Whether WURFL supports this case or not is separable issue. To me HTTP headers giving an indication of device capabilities, e.g. providing device type, URLs to appropriate profiles such as UAPROF etc. is a much more reliable way to achieve device capability indication than the IMEI which appears to only be useful in restricted cases, e.g. user agents integrated with the device containing the IMEI and domains where the IMEI is available Alastair J Angwin, Mail To : IBM UK Laboratories, Hursley Park, Winchester, Hampshire, SO21 2JN, UK Tel: +44-(0)1962-816817 ... Fax: +44-(0)1962-819777 ...Mobile: +44-(0)7703-128131 Email: Lotus Notes : Alastair Angwin/UK/IBM or ANGWIN@IBMGB Internet : alastair_angwin@uk.ibm.com angwin@uk.ibm.com External IEA : GBIBMX59@IBMMAIL "Luca Passani" <luca.passani@ope To: <public-ddwg@w3.org>, <www-mobile@w3.org> nwave.com> cc: Sent by: Subject: RE: Mobile phone capabilities list? www-mobile-reques t@w3.org 27/07/2005 09:07 >As has been said ... the IMEI can be used to know about the radio device >being used. But that does not assure knowledge of the client capabilities >e.g. laptop / PDA with GPRS card. >There is much more than is needed to be known. Really? where does this requirement come from? I have talked with hundreds of developers from very varied backgrounds over the past few years, yet I can't recall a scenario where wireless developers were going mad trying to identify if it was a laptop over a GPRS connections they were trying to send an MMS too. On the other hand, WURFL only covers WAP, markups, WAP Push, MMS, J2ME, EMS, ringtones, usability, wallpapers and streaming. You may be aware of other reasonably common scenarios outside of the domain of those technologies. In that case, I would really be curious to hear about them. Also, I think there is a misunderstanding here. Some may find it useful to use IMEI as a key to figure out device capabilities. This would be an extra feature. Nobody ever said that this should be the only way. In most cases, HTTP headers are much better for that purpose. IMEI is useful under a few conditions: - you have access to the IMEI code for the device of the subscriber you want to serve. - you don't have an HTTP request coming from the device Luca
Received on Wednesday, 27 July 2005 10:32:01 UTC