|Palm and LJ for WebOS
||[Mar. 4th, 2010|01:14 pm]
I worry about Palm sometimes. They recently lowered their guidance for this quarter, analysts don't seem too upbeat, and their stock price for the last year looked promising when they released the Pre in July, but has dropped dramatically since then.
More concerning is the fact that, 9 months after releasing the first WebOS phone, Gartner estimates that 0.7% of smartphones are running WebOS. Hopefully this will improve now that they're on Verizon (and rumor is they'll be on AT&T sometime this year) and once they launch in more countries.
The good news is that the mobile phone market isn't quite like, say, the social networking website market, which has a very strong network effect. If all your friends are leaving Friendster for Facebook, then Friendster is less valuable to you, and you'll probably switch to Facebook. But I can still use the web just fine from my Palm Pre even if the rest of the world switches to iPhones and Droids and Nexus Ones. There is somewhat of a problem that if fewer people use WebOS, fewer people will write apps for it, but this is more of a slow process. Also, at least in the US, most people are under contract for their phones and so they only have an opportunity to switch cheaply every one or two years. I'm really hoping Palm can keep turning things around - just yesterday they released an update to the Facebook app that makes it much nicer.
Speaking of apps...
LJ for WebOS has been doing pretty well since my last update - as of this very moment I've sold 72 copies. It seems fairly random how many copies are sold a day - thanks to the new app My WebOS Apps I have a nifty graph on my phone with these totals for the last week: 5, 1, 1, 2, 4, 2, 0. So...who knows?
One of the frustrating parts has been seeing bad reviews indicating that it just isn't working for a few people. Most of these reviews came early and I'm pretty sure I've fixed the bugs since then, but most people don't go back and edit their review when problems get fixed, and I have no way of contacting them to ask them if it's working for them and to try to diagnose their problem if not. I've been trying to make it more and more obvious how to contact me to the point that if you can't load the posts a dialog comes up with a button to email me the problem. We'll see if this helps at all. Encouragingly, more of the recent reviews have been good than bad, bringing the average back up to 3.5/5 stars.
I spent a lot of last week working on a new feature that I really wanted to add: the ability to browse other people's journals. I even wrote the parsing code before I got stuck on a problem that I've gotten stuck on before - the inability to properly authenticate so that the client can load friends-only posts. The API way to do this is to call the sessiongenerate method - unfortunately LiveJournal's cookie scheme has changed because of some security holes, and no one's gone back and updated or added a new API.
Every time I run into this problem, I spent some time trying to "fake it" by POSTing the right thing to the login page, essentially pretending I'm a regular user signing in from a browser. As in the past, I can't get this to work and I'm not sure why, and it's really frustrating to try to debug because it's all guesswork.
So I was a little down about that, and the last release (made it to the App Catalog yesterday) only had a few small features like deleting posts. This week I took a stab at some random not-quite-featurey things that have been on my list for a while, and everything just kinda worked. The next release will use a lot less bandwidth on the initial request (66% less in my test case!), and it fixes a bug with comments not posting by showing the CAPTCHA dialog that LiveJournal was returning. I was amazed that both of these basically worked the first time, so that was a nice pick me up :-) I'll probably submit it to the App Catalog tonight after a bit more testing.
I'm running out of features to work on that are actually possible to do, so I'm very open to suggestions!