You may want to have a look at the WURFL - an approach being adopted for WAP
devices - see
http://groups.yahoo.com/group/wmlprogramming/messagesearch?query=wurfl
Clive
-----Original Message-----
From: Nick May [mailto:nick@kyushu.com]
Sent: 25 February 2002 13:17
To: keitai-l@appelsiini.net
Subject: (keitai-l) Re: open source keitai tools (was Re: Re: western
phone, imode sit
keitai-l@appelsiini.net writes:
>So far, it sounds like there's not much out there to
>start with. Why would this be? It's an odd situation.
>
>Well, here's my diagnosis. It's a diagnosis from a distance,
>since I've done relatively little hacking in this area, sad to
>say. And it might be a view through the wrong end of the
>telescope.
Well, there are certainly some things out there - and to date they have
been "enough". The problem has been "manageable" (ignoring WML) and most
people have been happy to hack things out for themselves or use Mika's
imode classes as a basis for their own stuff (in other languages I guess).
Since many people have been using php or perl (both of which encourage
the ad hoc approach), it just has not been an issue. I bet that lots of
people on this list have there own little data hoards about phones. Bill
Pellowe certainly has his own basic stuff for various JPhone handsets that
is brought intermittently up to date. The rest probably don't care, or
don't sniff'n'serve.
Now the problem is getting less manageable as the number of phones and
phone type increases. Also more and more people are looking to use a
Cocoon type solution. Both militate towards a more co-ordinated response.
> If you don't have a
>simple, open architecture, it might as well be 6502
>machine code.
Granted, data architecture is critical. (Not that there is anything wrong
with 6502 machine code. Ee! - when I were a lad....).
> Data can have inaccuracies, but not bugs or features. So of
>course it's just consumed, and not contributed to.
I don't think this follows.
>So: committed developers, user-driven data, data-driven
>code, all on the same page. A simple formula, perhaps,
>but not an easy one.
Oh bollocks. I don't see THAT much commitment from developers being
required (in terms of time at least). This is an area where half a
solution is a glass half full. Miss out a few phones and it does not
invalidate the rest of the data one has for other phones. This is not
world domination we are planning or the "one ring to in the darkness bind
them" (although it does have certain similarities - perhaps you could play
Gollum...) - just a more co-ordinated hack than the ad-hoc private hacks
lots of us currently do.
My interest is entirely practical - if I can't persuade others to do it
with me/for me I will have to do it myself for my own use. I suspect that
if the contributors to this discussion had collectively spent as much time
hacking as we have in discussion, we would have a semi-workable framework
already. Mika Tuppola knocked up something basic but very useful for imode
in a weekend. Most people knock up their own little tools for
imode/jphone url, emoji/graphic and tel: handling in a couple of hours.
Granted, it is all more complex in a cocoon type environment, and much
thought should be given to datastructure - but I don't think it merits the
"doctrine of despair" feel of your last post.
Can we stop philosophizing about it and just get on with it?
Nick "Iji" May
This mail was sent to address Clive.vanHilten@Misys.com
Need archives? How to unsubscribe? http://www.appelsiini.net/keitai-l/
Received on Mon Feb 25 16:28:31 2002