Archive for October, 2007

Greenphone Official Word

Tuesday, October 23rd, 2007

There’s info about the Greenphone and Qtopia Phone Edition up on the Trolltech site now. Interestingly the discussion about the future of the Linux platform seems to have moved over to Vodafone Betavine. Interesting choice, I’m not going to comment.

As a hobbiest interested in open mobile platforms I’m going to have to downgrade the effort from “neutral” to “ignore”. It might be worth checking out at some point, but the effort is off the rails right now. Note that this is the hobbiest view: if you’re just taking time on your own to poke at something cause you’re interested in having one of the devices around and think you could make it do interesting stuff. If you’re looking to manufacture a load of devices and sell them to consumers that’s a whole other issue I’m not going to comment on now.

My current recommendation for the most promising long term project is OpenMoko. It just actually makes sense. There’s a kernel, a toolchain, core services, and I have info about all of them. I still don’t understand what’s in the Greenphone SDK. And my current recommendation for best way to get some instant gratification hackery going quickly is the Maemo platform that runs on the N770 and N800 internet tablets from Nokia. You can walk into Best Buy or CompUSA and get one, get home and be developing within a few hours.

And Then There Will Be Cake

Tuesday, October 23rd, 2007

A new video game store called Play N Trade just opened right along my walk from work to home in San Mateo. And in addition to posting about Team Fortress 2 Russ also passes along a video clip or two from something in The Orange Box pretty much every day. So while I was walking home last night and saw that the store was still open I dropped in to get a copy.

The Orange Box is actually a collection of games, two of which I was particularly interested in:

Upon getting home however I discovered that Team Fortress 2 is online play only, there’s no campaign mode to help you get used to the controls or learn how to use the skills your character has. And the folks online are well, generally what they are in every other online game. Maladjusted, belligerent, and generally intolerant of anyone who doesn’t have way more time to devote to gaming than I do. So scratch that for now, I’ll have to find another group of newbies to play with.

Portal however sucked me right in. As Yahtzee commented in his review of The Orange Box:

Absolutely sublime from start to finish and I will jam forks into my eyes if I ever use those words to describe anything else ever again. (….) Portal is great and if you don’t think so you must be stupid.

That’s a pretty fair evaluation. The dark humor of the computer voice prodding and prompting and insulting and encouraging and occasionally making nonsense comments is fantastic. And the puzzles themselves were really interesting I thought. Check out the Wikipedia entry for some examples of variations of the kinds of stuff you need to solve. It reminds me of Myst in a lot of good ways, only without the boring bits and a whole lot funnier. Plus, it’s easy to die.

If you grab a copy of The Orange Box and feel like trying out Team Fortress 2 without a bunch of whining 12 year olds complaining about what shots you should have taken but didn’t, send me a message (my gamertag is Bitsplitter).

Russ Attempts a “Pants-on-Head Retarded” Usage

Monday, October 22nd, 2007

Ever since we encountered the the Zero Punctuation review of Halo 3 every one of my friends has agreed that our new ultimate goal in life is to be able to slip the phrase “pants-on-head retarded” into casual conversation as if it belonged there. It’s a fantastic phrase! So evocative, so terse, so derisive! Even without the little stick figure with his shorts on his head, you immediately grasp the seriousness of the situation. Russ has made a public effort, including the iPhone zealots as the target audience. Doesn’t quite flow right, but maybe I have an overcritical eye, tainted by my own inability to slip the phrase in without getting caught. Does it work? I’m just not sure it works without the accent… it might need to be delivered spoken.

Intersection of Video and Gaming

Monday, October 22nd, 2007

Charlie posted about a full length version of Bloodspell created by Hugh Hancock using a technique called machinima. I’m downloading it now. What it is at heart is using an engine from a gaming system (or something not meant for movie production) and using it to create videos. Anyone following Halo 3 has probably seen a whole bunch of clips of in game play meant to just be amusing antics, sometimes with voiceovers for the characters added afterward.

Charlie says: go have a look at Bloodspell — it’s only a 900-1000Mb download. And remember: this isn’t about to replace Hollywood tomorrow, but if Marvell and DC Comics aren’t feeling the chill wind down the back of their neck, they’re asleep at the switch. Because as streaming internet media players become ubiquitous, this sort of thing — cheap, fast and out of control — could very well be the future of mobile entertainment.

Interesting, very interesting. What I was watching for, and haven’t really seen yet, was the use of spectator modes in online game networks like Microsoft Live to generate content. It follows right along with the whole user generated content set of ideas that’s been popping in and out of every aspect of internet services for the last few years. And although I know it’s possible to do some stuff like create fly-throughs and saved films (for instance saved films is one of the new features in Halo 3) sharing them outside the realm of the Live network seems to be clunky at best:

Q: How do I get my Saved Films onto YouTube and the like?

A: The simplest method would be to use a video capture card and appropriate software to turn the Saved Film into video data. At this time there is no software capable of doing this without video capture hardware. Machinima makers and amateur videographers already know lots of tricks and methods for doing this.

I was expecting it to take a lot less time for Live to run something like a “highlights reel” for users. They know what games I’m playing, which is likely what I’m going to be interested in seeing highlights for. Saved films at least allow you share with contacts. It sounds like they’re thinking about a collective intelligence style emergent network for recommending clips and letting them bubble up overall (check out the last few Q/As on that Bungie page). I would assume that stuff like this would make great content for the G4 channel, and would definitely be the kind of thing I would watch from my mobile. If Russ sent me a clip titled “Expert plasma grenade technique, beat that!” it’s the kind of thing I would pull out my phone to watch. Especially if he were challenging my plasma grenade sk1llz like that, immediate representation would be required.

Greenphone Killed Off?

Monday, October 22nd, 2007

Engadget Mobile reports that Trolltech has sold their last Greenphone and is recommending OpenMoko hardware if you want to try out their software. I don’t see anything about this on the Trolltech site however. Last time I spoke to the folks at Trolltech they were talking about further iterations to the hardware and really refining what they were working on. Still, consolidating under the OpenMoko hardware platform feels somewhat like “standardizing on a platform”, so I’m not going to bitch too much.

I ran into someone at one of the MoMo events (at Trolltech interestingly enough) who had a hacked up version of Qtopia phone edition running on some commercial hardware. I think it might have been the Netgear Skype handset. Sure would be cool if they would release info about how to get something like that going. Ultimately the Greenphone didn’t match up with what I would like to see, with a few of the core components being closed. Maybe Motorola will release a Qtopia handset that helps out us open source developers (I’m not holding my breath for that, I hear Motorola mentioned at the same time as Linux on a daily basis but I’ve yet to figure out what the hell it is they’re working on). But it would be nice to see Trolltech giving their existing developers an easy way forward rather then dropping them off to some other project to fend for themselves.

Modifying docpurge

Sunday, October 21st, 2007

The N800 comes with a package called docpurge that deletes files which for most users are just cruft. I end up installing some packages that install stuff I actually want in places like /usr/share/doc. I want to keep them around, but I also don’t want them sucking up core memory either. Fortunately the script that cleans up the files is just a shell script that gets called by the package manager, its located at /usr/sbin/docpurge. I modified mine to check if the internal MMC card is present, and if so copy the files from /usr/share/doc out to the card before removing them. Good enough for what I want.

AdMob Stats Remix

Sunday, October 21st, 2007

We released a bunch of data to share a view of our particular slice of the mobile web based on the requests we see coming through AdMob, and Dennis from WAPReview went ahead and added some additional data to some of our tables to take a view of the market by screensize. Fantastic, thanks for contributing Dennis!

There have been a bunch of questions about the data coming back in emails and in postings on discussion boards. Sometimes the quirks are things related to the fact that we’re measuring this across a publisher network which doesn’t map 1:1 against the normally carrier based (or survey based) metrics that commonly appear for us mobilists. And then there was the issue with us having called the United Kingdom “Great Britain” in the original report. Sorry, fixed that. Obviously, we don’t get out too much :-)

Some of the other numbers that have been called into question are a bit harder to pin down. Like the overall reported size of markets when compared to the numbers we see at AdMob. Are our numbers wrong or skewed by something we don’t have great visibility into? Or are the existing outside numbers skewed in some way? We released our numbers because the environment is very information poor at this point, it’s hard to find concrete sets of numbers pulled from hard data. We were hoping that feedback from outside would help us pin down places where maybe we’re doing things wrong. But also hoping that it would encourage others who have hard data to open up some of what they have so that the community can start to build a larger understanding of what exactly is happening on the mobile web.

It’s very much like the open source maxim that “with enough eyeballs all bugs are shallow.” With enough interested folks watching the numbers and enough sources of information we should be able to get to a real shared understanding of just “how big the mobile web is” and how different areas compare. We’re planning to release numbers each month, tuning that report and refining the way we collect our data as we go. Being transparent with stuff like this means that sometimes you end up airing your mistakes for others to criticize. Thankfully the release has gone fantastic so far. I’m just hoping that when the time comes that we have something that we actually screwed up we don’t have to kill it off for fear of the negative PR.

NFC for Mobile Payments

Sunday, October 21st, 2007

Enrique posted his slides about “NFC in Mobile Commerce” from the Texas Wireless Summit. It’s an excellent overview of the technologies, but what I’m always more concerned about is the backend billing infrastructure. How are the transactions going to be cleared through to the end user account?

Is the carrier going to expose their billing relationship out through the NFC transactions? If we base the expected charges on “how valuable the carrier relationships with their customer” is we can expect a what? 30 percent cut for the carrier? Maybe 25? What convinces the carriers in this case that they should invest in a new set of infrastructure with a non-trivial amount of risk when they can just sit on their current monopoly? Also, I’ve got absolutely no knoweldge of economics or consumer behavior and stuff like that, but does this mean that the carrier would have to start extending additional lines of credit if we want these things to act like credit cards? Or can you bundle the account with a credit card easily enough that it would seem transparent to the end user? Cause that’s what we would need for mass adoption. I need to be able to walk into either my bank OR my carrier and say I would like a new credit card and walk out with a phone that works for NFC payments.

Or are the credit card companies the ones who end up clearing transactions for this stuff? It would seem like the carriers would be unwilling to allow that to happen across their network without someone allowing them to wet their beaks a bit. So will NFC transactions bear a higher cost to complete than the existing card based transactions? Who’s going to bear that cost? The credit card company? Why would they go through all the trouble to get NFC rolled out just to migrate transactions from a higher margin line to a lower margin line? One would assume they would start to capture more transactions, things done with cash today would be done with NFC tomorrow, and the credit card provider would get incremental revenue from their expanded role.

But money doesn’t get instantiated from nothing in either case, where does that money come from? Aren’t we really talking about sucking additional money out of the consumer in the name of a convenience they might just be mentally detaching from the actual cost? Ask them if they want to pay with stuff using their cell phone and I’m sure they’ll say yes. Ask them if they want to bear an average of 5 percent increase in their spending for the same goods in order to use their cellphone I’m not sure all of them would still say yes. The possible source of the funds would be in fraud prevention. If a handset based instead of a card based authentication mechanism would reduce credit card fraud by 20%, and the cost of deploying and running the service amounted to 15% of fraud loss, we would have something besides a zero sum game going on. But I haven’t seen any meaningful discourse on that yet.

I’m very much down with using tags/barcodes to extending URLs into the real world by associating physical items with online resources. No argument for that happening right away and all over. I’m also very much down with the mobile becoming the major source of transactions in the long run. Korea and Japan have proven that the model works. But any system like this isn’t run in isolation. When you’re booting up a payment system from first principles with very little existing infrastructure and almost no entrenched players it’s much different than revolutionizing an existing system, even if it is an antiquated and broken system. I see it being a while before I’m able to pay for most of what I do using my phone, even as an early adopter in the US.

Mobile Portals

Friday, October 19th, 2007

This is something that was mentioned at Mobile 2.0 a few times as well: there should be a mobile portal of the same form as the early Yahoo directory when it was still pages of links. We’ve been working on a mobile portal called TapTap at AdMob. It’s probably the right phase to have something of the sort, but I could be wrong. Only time and consumer usage will tell. Thing is, there are so many mobile portals already I wonder if what the world needs is really another one. Here’s a list of the ones I’ve been poking at, in the hopes that a little cross-pollination will help out a bit:

If you’ve got others (particularly others of note) please leave me a comment.

OpenMoko on Existing Hardware

Thursday, October 18th, 2007

Something I hadn’t realized until I went to a presentation last week was that OpenMoko is up and running to varying degrees on hardware other than the Neo1973, including the Treo 650. Cool, I think I have one of those kicking around somewhere. I also didn’t realize that all the info I was really looking for was over at OpenMoko.org (the OpenMoko.com site being the commercial part mostly for selling hardware). The wiki has a ton of info, almost none of which I’ve gotten a chance to read yet.