Slashdot Mirror
Another Angle To WAP And Linux
An anonymous reader pointed us to an article running on LinuxDevices.com talking about
Supporting WAP in Linux and why this should be a priority. WAP has taken a lot of (deserved) heat, but this is a good argument on the other side.
The issues raised include:
--
Many people here seem to be confusing a big issue here. This post is regarding the entire of the WAP situation. WML, which is what one would use to provide content for the WAP gateways, is one of the *SMALLEST* parts of the article. The WAP protocol is what he's talking about. WML is the display mechanism that you can spit out. WAP is the raw protocol in and of itself. WAP = IP.. WML = HTML.. The gateway converts WAP requests to HTTP. Apps could, theoretically, spit out WML direct to the phone, instead of going thru the HTTP 'gateway'.
-- I'm the root of all that's evil, but you can call me cookie..
Low bandwidth
Text-only interface
No advertising
No banner ads
Sounds like Lynx on Linux, I know....
Hmm - your comment appears to be both right and wrong to me. You're right about what the issue should be, but you mention "getting it done, tested, deployed". That's why a WAP browser that runs on Linux is important at the moment.
I develop for WAP and I use Linux as my OS of choice. Until very recently I had to boot into Windows (or upload and use my phone) every time I wanted to check that the WML my application is outputting. I think that an open WAP browser project is a great idea - just like the various open source WAP gateways are great ideas (and hopefully the various web tools that will spit out ready to run WML will be).
"Give the anarchist a cigarette"
A little planning goes a long way...
-
Security -- Firstly, even though the WAP protocol has a security mechanism via WTLS, it is not possible for the content provider, using HTTP or HTTPs, to know whether there was security over the air unless a proprietary header is placed in the HTTP header. Secondly, when a user connects with a WAP Gateway it authenticates with it, but never authenticates with the content provider. Thirdly, when the user has authenticated with the WAP Gateway, the gateway decrypts the secure content and might re-encrypt the message to be sent over HTTPs. This implies that the secure content that was intended for the content provider flows as clear text through the WAP Gateway. Fourthly, the WAP Gateway represents the browser and means that any cookies or security information is kept on the Gateway. Information such as credit cards, address, names and all sorts of other information usually kept in cookies by the browser are kept by the Gateway.
If you would read before posting you would appear much less stupid.I'm with a research project that used Wap extensively. Not myself, but my fellow students. Pity them. They kept swearing and swearing and swearing.
Note: This is is not my personal experience. I am only rehashing what they told me:
Poor device support. Constantly crashing Nokia phones. Very limited page sizes (there is a fixed maximum in the most common Nokia phone). Every phone provider offeres a completely different Wap gateway. Each of them has its own set of bugs. Try to type text in a WAP phone when you search something without making you want to scream out loud. Try to fit some actual informational response on a typical phone's screen. And of course, the costs are astronomical here in Germany.
So no, they told me that it does not live up the hype.
------------------
------------------
You may like my a cappella music
Gateways already exist for linux (from phone.com) but installing and running your own gateway for a commericial or hobby venture is retarded; unless you are planning on a private wireless intranet. If you are planning for that, then there are solutions that exist already that are 100% java by Nokia. Both options are expensive though...
More then that, WAP is incredibly lame compared to other technologies that exist now: voiceXML, Palm PQAs (web clippings), NTT Docomo's iMode phones, etc. My guess is that the need for WAP/WML as it exists today is going to be eclipsed by telephony applications because they let you get to the information you want faster and with less hassle. A combination of voice and display would, naturally, be killer.
I was at phone.com's Unwired Universe conference in San Francisco this summer and I walked away with a bitter taste in my mouth. Gateways are controlled by telcos, getting your WAP application listed on the gateway costs $$$ (unless your a major like Yahoo), and most US phones don't let you enter in a new address for an alternate gateway so you're SOL if you want to switch. Beyond that WAP is a horribly thought out spec, WML being a crippled lame piece of shit utilizing a horrible metaphor for handheld app development.
I do this for a living.
-earache
This is like saying:
Write apps for Windows because it's already there
Support the RIAA because online models of distribution aren't fully realzed yet
etc.
Look at the mess this attitiude has created in the current implementation of TCP/IP. The entire net is run on a hideously patched together and crufty protocol because everyone just decided to support it because, hey, it was there.
AFAIK, every last one of the problems noted in that article is being worked on as we speak{1} at many levels, from working on the next round of standards to implementing the current ones properly. Including end-to-end security, which, for example, is now being worked with by a PKI text-signing/crypting solution with the keys stored in the SIM cards, and Toolkits that will let WAP developers simulate and experiment with this.
It is too bad WAP wasn't a complete solution where everything was taken care of from the moment it was specced, just like the WWW had forms and https and dynamic content generation the instant it came out in 1993- oh wait, it didn't, did it? I am sorry, but these things just take time to get them right.
WAP standards themselves are exploding to take care of the objections voiced. We telco developers - Nokia here - moan and groan after each round of the WAP forum about what our browsers and Toolkits and gateways have to implement this time, yet we understand that the market needs these solutions, and we get to work. You will get push, you will get interoperability with your voice-mail and phone-calls and address-book, you will get styles and DOMs for your phones, current and 3G, you will get easy gateway-provisioning when you switch networks, you will get end-to-end security, and you will get it in a way that is not haphazardly cloned from the wired world but a way that makes sense for the networks and the devices and the way the bills get paid.
We also know that we aren't dealing with just one or two players like in the Netscape vs. IE world, but with a massive amount of browser- and gateway-makers, and that this time we have to make absolutly sure everything works together. Guess what: testing all this shit takes time. Believe me, you don't want to see our QA-matrix, nor the esoteric this-gateway-and-that-handset bugs we are chasing. We did learn from our first releases.
I am sorry you couldn't get it all in one fell swoop yesterday, a perfectly finished standard that did everything you needed, all with these 100% bug-free implementations. But we are working to get it to you now.
FJ!!
{1}Well, maybe not, it is time to sleep now in Hong Kong.
Now, I'm going to explain how the breakthrough technology of IP (Internet Protocol) solves all these problems:
Reliability: IP was designed back in a time when even leased line connections were unreliable, let alone the computers that were linked up to them. I would argue modern wireless communications is not significantly more unreliable than wired communications were back in the early days of IP. Furthermore, HTTP was originally designed as a stateless protocol, and as such, most "web sessions" are persistent for a limited period of time. That's right, I can pull out my ethernet plug, wait for 10 minutes, plug it back in, and then bingo, my Dell shopping cart is still there!
Variety of devices: Considering how many different kinds of devices support IP and even the Web, it seems insane to me to suggest that IP-based technology doesn't already provide enough capability to get the job done. Indeed, XML/CSS/XSL/XSLT etc. are all designed to address this issue. Indeed, before HTML became bastardized it was supposed to address this issue. The only thing that has kept the Internet from supporting a wide variety of devices has been market forces, not technological limitations. Wireless communications will hopefully balance that out.
Wide variety of networks: The original intent of IP was to bridge together a wide variety of networks. As such, IP can already be embedded on top of (and used to bridge) DecNET, NetBIOS, IPX, etc.... even itself!.
Limited bandwidth: Wireless networks today typically have 14.4kbps bandwidth, and those numbers are expected to climb significantly in the years to come. When IP was first being developed 2400bps was a lot of bandwidth. So, don't tell me that IP can't be used in low bandwidth situations.
The WAP guys have developed a huge set of protocols and technologies that mimic their IP counterparts. They've done this seemingly without considering how to use or extend the existing IP protocols to support their needs. I think it's pretty clear why this is happening.
sigs are a waste of space
This article seems to not mention the most important point- Geoworks claims to own a software patent that applies to all WAP implementations.
Thus, anyone who distributes WAP software is open to being sued.
Thus, no Linux distribution (at least in the US) will contain WAP. And no open source programmer in the US should be contributing publicly to WAP stuff unless he/she is willing to take a considerable risk...
I haven't actually heard anything too negative towards WAP... am I missing something?
/. article on it... unless I was very bored...
I've used it on my cell phone - and it's cool enough for looking up quick info (stock quotes, e-mail addresses, phone numbers, etc)... but I don't think I would try and read a
BlackNova Traders
Just wondering - I think WAP is a great technology, but places like here in Denmark, its really expensive to use the cellphone, even for data transmissions. We know it won't be long until we see advertisments on WAP sites, which will just make you spend more time "online" at the cellphone. Shouldn't we fight for cheaper telephone rates instead of making more and more flash and candy which only will make the users spend more money. On WAP and Linux, if people need it, and WAP is an open standard (correct me here?) - If people or a coder needs it, they will eventually make it. If for instance Nokia goes and makes a WAP browser for *nix/BSD, as binary only, would ya all go and beg them to go open source? If yes, why don't you go make something alike?. (sorry if it sounds like flamebait). My 2 cents..
-Stskeeps, http://unrealircd.com
WAP has taken a lot of (deserved) heat, but this is a good argument on the other side.
WAP is a crappy protocol that will be mutilated beyond recognition to make it (slightly) less crappy before it is popular. So any work done on it's current incarnation will need to be redone within a year.
1 Alpha7
Live to be Moderated
Some people will say that you shouldn't support WAP because of it's shortcomings. I would argue against that viewpoint.
An OS (ie Linux) should be able to support as many protocols/standards/etc as possible of each type. The ideal is for people to be able to choose each aspect of their system, without it breaking other parts, or requiring X so I can use Y. Everything should be as interchangable as possible.
WAP should be supported simply because it's there. You don't have to use it; just accept that it's there. That way, everyone can see it and try it, and the same for any alternatives that are offered. Then they can pick the one they like, or even use a mixture of them all. Interchangability.
The philosophy here is to make all the options available. If everyone can use whichever option they like best, and have the chance to make the judgement freely, they will for the most part use the best option. Of course, "best" means different things to different people, so you may still end up with several standards, but the poor ones will fall away, and the good ones will survive.
(Spudley Strikes Again!)
Linux does have some userland-uses. Five or six stories ago on slashdot, they mentioned that IBM's working on a Linux-based wristwatch. On top of my TV is a consumer electronics device running Linux. Linux may not be to the point of being able to pander to the point-click-drool crowd, but it's certainly more than just a server OS.
The most advanced is currently Kannel, followed by Ophelia and GNUws.
WAP is a userland phenomenon. Developers who target WAP vs. HTML vs. whatever, are developing in Perl, PHP, etc. -- this stuff has nothing to do with Linux, it's going to run the same whether they run it on a Linux box or a Solaris box (or even a Windows box, if a portable language is installed on it).
The real issue should be of getting the open source world's portfolio of web service tools to dish out WAP in a friendly and easy-to-configure way -- and getting it done, tested, deployed, and grabbing market share before Microsoft starts raping and pillaging that part of the market.
--
Tired of FB/Google censorship? Visit UNCENSORED!
'Course. I could be wrong :)
-Vercingetorix
-Vercingetorix
"Necessitas non habet legem." -St. Augustine