Blanket Statements about WAP

[Arved Sandstrom <Arved_37@chebucto.ns.ca>]

Hi,

Tim Bray said something a while back (on the 12th of this month, actually) 
that has bothered me a bit ever since.

<quote>
Except that WAP is a miserable multibillion-dollar failure... some
have argued convincingly is that they crippled the design based on
the assumption that the devices would have to be stupid and the
connections slow; leading them to, among other things, binary XML.
</quote>

I work for a company that's been writing stuff for cellphones (and now PDAs) 
for over 2 years. Our observation so far is that the devices _are_ stupid 
and the connections _are_ slow; we'd be delighted if the circumstances were 
otherwise. I don't know if the WAP designers crippled the spec or not - I do 
know that we pay attention to the size limit on the binary, and curse it 
sometimes, but when it's exceeded then it's because a ridiculous amount of 
content was pushed out for that tiny screen anyway.

Actually, come to think of it, of course they crippled the spec, and for 
good reasons. Give people a richer markup and they abuse it - the norm in 
the industry for pushing content to browsers on PDAs is to do screen 
scrapes; these look atrocious. Give people something more than WML for the 
phone and they'd abuse the hell out of that, too.

Sometimes it's hard to tell what speakers mean when they say WAP - all of 
WAP, just WML, who knows? Some of our people went to a conference and came 
back shaking their heads over a Simon Phipps rant about (guess what?) WAP; 
evidently he had never used the technology but had picked up some nice 
slogans about how crappy it was. Point being - we use it every day, and 
comments like this leave us scratching our heads: exactly what is it that we 
can't do? Demonstrate to us why WAP/WML is so terrible. We're being told 
that we are handcuffed, but I'll be damned if I can figure out how. One of 
our junior guys had a telling comment about this kind of sweeping 
assessment: "do they actually write this stuff?"

We bump into device limitations and gateway limitations and carrier 
limitations all the time. Nobody here is screaming for an enhanced markup 
language so they can do shopping cart apps on cellphones - we are entirely 
satisfied with the functionality subset that WML has right now. It's just 
about right. The deck/card metaphor works very well. We write stuff for PDAs 
also, and that's a whole different story. But for phones, WML is just about 
bang on target. That's the crop of phones we have right now in North 
America, have had for years, and will have lots of for the foreseeable 
future. In real life we don't have the luxury of using the latest device 
with the latest markup, not if the client has an installed base of 25,000 
phones with UP.Browser 3.0 or 3.1. And with HDML it's not XML anyhow - I 
guess Mr Bray hasn't had to struggle with HDML lately. Must be nice.

I don't know enough about the nuances of the binary coding inherent in WAP 
transmission of WML to comment, except to say that with all the other 
latencies that occur in the typical roundtrip, and considering the device 
capabilities, the encoding is probably unnecessary, but I wouldn't take a 
religious stance over it.

Overall point being, the assumption alluded to by Tim isn't just an 
assumption; that is, in fact, exactly how things are: dumb devices and slow 
connections. Maybe in 2005 things will be markedly different, but I'm not 
holding my breath. So, at present, and for years to come, WML works very 
nicely. I suppose I'm curious as to what the better alternative might be?

Regards,
Arved Sandstrom

Fairly Senior Software Type
e-plicity (http://www.e-plicity.com)
Wireless * B2B * J2EE * XML --- Halifax, Nova Scotia