Even with all the VC money, recent history is full of startups that tried and failed in mobile soft market. Successes are almost non-existent and those that are still alive, are zombies. Stay away from walled gardens... you cannot win. You cannot rule somone else's sandbox. The networks are their property and they hold all the keys. Even if you win the lotto and you enter their network, they can boot you and lock you out anytime they wish. And they will.
More here: http://opengardensblog.futuretext.com/archives/2006/06/opengardens_wal.html
Then there's WiMAX and the other "last mile" broadband initiatives. Carriers are no longer the only ones who hold the key.
Its not a matter of if mobile devices will be the viable application development platform, its a matter of when and how. Startups attacking these issues and their investors are not wasting their time.
Oh, I have. Big problem with those is that WiFi is treated as just another carrier channel and you still have to figure out how to install your software on a closed-off cell phone. First of all, carier are very picky about what they allow on their networks. They will tell you that they do that to protect their networks from malicious software and to protect their consumers' phones from draining their batteries too quickly.
"The combination scares the HELL out of the carriers. But market pressures are leaving them no choice but to accept this is what users want."
They're not scared. They're just playing the embrace, extend and extinguish strategy. Recent announcement by T-Mobile and UMA support shows what game their playing: http://gigaom.com/2007/05/03/t-mobiles-wifi-cellular-summer/ MetroFi deployments have been a failure so far. Earthlink has stopped responding to RFCs from cities. They realized that no one is willing to pay for WiFi. That means that free wifi spots are going to be the norm in the future. Problem is that these wifi spots are not very reliable and the range is awful. Look at the popularity of EVDO as an example why (metro)WiFi has been a disappointment so far.
"Then there's WiMAX and the other "last mile" broadband initiatives. Carriers are no longer the only ones who hold the key."
You still need spectrum for WiMax. Which startup has $100mm to buy some spectrum for WiMax?
"Startups attacking these issues and their investors are not wasting their time."
I'm not saying it's an impossible problem but what I am saying is that your chances of success are very low indeed. You have a much higher chance of success by making your software work over open networks such as Internet than trying to crack a walled garden.
The problem with mobile carriers is I think that people are not aware of what they should expect - they are not aware that they are denied the functionality "X", so they are not complaining to the carriers. But suppose one of the carriers would block for example Twitter, wouldn't people start to notice?
On my T-Mobile service, for instance, I have to send an SMS to "phone number" 500, with the email address, subject, and body of the message separated by appropriate delimiters. How does it work on other carriers?
What I need is a comprehensive, cross-carrier list for sending email from cell phones. I've found this information for a few carriers, but it's not complete by any means. For instance, it looks like Cingular (U.S.) and Rogers (Canada) use the same method as T-Mobile (U.S.), just using a different number (500 for T-Mobile, 121 for Cingular, and 0000000000 for Rogers).
I may have to devote a few days to digging out this information by hand from the carrier web sites. If I do, I'll add it to the Wikipedia article.
I think it's a fascinating space, in any case. Sooner or later people are going to start getting it right. I wrote up some notes on what I think "getting it right" means right now:
http://journal.dedasys.com/articles/2007/05/03/cell-phone-software-the-right-path
[1] I seem to recall reading about a certain well-known individual who has a neat idea for a language that still hasn't seen the light of day after years, although it's rumored to power some kind of news site.
we (heysan) are building a mobile messaging services, but the guys at iminlikewithyou have done a great mobile integration too.
I'm interested in bth 'pure' mobile startups and those who have mobile as just a feature in a bigger service (dopplr for example).
twitter and jaiku are mobile startups, right?
and also: http://paulgraham.com/fixrazr.html
With the right killer app, people will strive to buy the right phone.
Also mobile don't have another 'killer app' to come. Everyone wants something different from the phone, just like everyone uses the web for different things. It's about choice. Some want news, others want myspace or funny videos of cats. What mobile needs is to be easy to develop for and easy to find the apps that are right for a particular user. We need a yahoo or AOL that will guide the new users and show them what its possible to do on their phone.
(I can't find the reports that show the japanese usage of data currently. From what I remember most used voice and sms. Photos and email were also popular. After that usage for any function drops off into a 'long tail' distribution. The japanese users were all heavily using their handsets for different purposes. I'll see if I can dig up the report in the morning but it makes for fascinating reading)
Twitter and Jaiku do not _run_ on cell phones. You do not have to install anything to use them. They use SMS gateway which anyone can setup (if you have $50K+ a month to pay for the traffic, that is). My post was about apps which you have to _install_ on a phone! There's a huge difference between the two approaches.
I've noticed that most of the popular, succesful mobile startups use just sms/mms and a counterpart web site currently. Perhaps we shall see them 'evolve' as the years pass as mobile apps and then mobile web become familiar to 'regular' mobile users. Users will look back at the day you had to remember text commands and shortcodes just like we now look back at the command prompt.
For now I agree, apps are a pain and you should stick to the most well know interface of sms when you can. The 'fun' stuff needs apps but the chance of success is a lot lower currently.