Alpertopia

Friday, January 15, 2010

Hulu Desktop + Wiimote

I learned today of Hulu Desktop, a way to watch Hulu without opening the web browser. I had just bought a Bluetooth dongle with the hope of using it as an interface device, and this was the first opportunity to do so!

Using Jason Smith’s Wiimote Presenter, I just remapped some keys to make more sense for Hulu. The Hulu configuration file for Wiimote Presenter is here.

Windows 7 also comes with Bluetooth support in-box, so all you have to do (if you have a BT dongle or laptop) is right-click the Bluetooth icon in the system tray, go to add device, press 1+2 on the wiimote, click the wiimote in the device list, and select the option to pair without a code.

Sunday, January 10, 2010

Stupid Norton: IPv4 autoconfiguration error

I spent two hours trying to figure out why, whatever the configuration settings I threw at it, a laptop would not connect to a wireless connection. Turns out the culprit was Norton Internet Security. Only by running the Norton Removal Tool was I able to fix the computer.

Basically, ipconfig /all would give me the following message:

Connection-specific DNS Suffix . :
Link-local IPv6 Address . . . . . : fe80::ed42:547:5d1b:20ef%9
Autoconfiguration IPv4 Address. . : 169.254.32.239
Subnet Mask . . . . . . . . . . . : 255.255.0.0
Default Gateway . . . . . . . . . :

Ooooh, how fun, they say. 169.254.x.x is the IANA-defined black-hole. Oops. Of course, rebuilding the IP stack or turning off IPv6 didn’t help. Sigh. Probably recommending that everyone remove Norton now now that MS Security Essentials, AVG, Ad-aware, etc etc do a better job without destroying the computer while updating.

Thursday, January 7, 2010

Locked Out of Your G1?

One issue has been bugging the Android user community for awhile now, and it seems like Google has been mum on the issue (android bugs 3006, 4784).

The issue is that a phone can be locked out if too many tries of the unlock pattern fails to unlock a phone (usually a friend or a kid playing with the pattern maker). The fallback mechanism that the Android puts forth is to fall back to the google login screen, which prompts for the google apps account associated with the phone. Only problem is that even though you enter your username and password correctly, the phone refuses to let you in. This is with Android 1.6.

Here’s how I fixed it, based on feedback from the above two bugs.

Have someone call you.
Either you or the caller hang up.
Immediately start mashing the home button, this should bring you to the home screen, bypassing the lock. Be wary as inactivity will reinstate the lock (though a keyboard-open event will bypass it again).
Go into Settings > Security and disable the pattern lock by entering your pattern.
Put the phone to sleep.
Wake it up, and enter in your username associated with the phone.
Enter in ‘null‘ as the password (no quotes).
Rejoice.

I’m not sure if entering in null as your password in the outset will work, but that’ll be fun if it did!

The problem with the implementation here is that if the phone is locked out, the Android phone will check a local hash of the account password. At some point in the unlock process, however, this hash is screwed up and overwritten. The phone seems to have no inclination to connect to 3G in order to verify the password, though some users have reported persuading the phone to connect to the internet in order to verify the password.

Seems like a serious issue as the only solution before this fix was to hard-reset the phone. Any issue like this that threatens to corrupt/lose customer data should be prioritized to the top-level. That’s, of course, not to mention (“the feature”) that you can circumvent the security mechanism by mashing the home button! I hope this helps anyone else that runs into this problem.

Monday, December 7, 2009

Overlake Double-take

Can someone tell me what the Sounder is doing at Overlake Transit Center? I wonder how heavy rail will annoy people working around that area, heh.

Wednesday, December 2, 2009

Dec’09 UW Light Rail Station Meeting

This is the second or third of these meetings that Amanda and I have went to, and it’s definitely starting to get a bit interesting. According to Sound Transit’s project page and the presentation that they gave, final designs have been completed and the transit agency is preparing to go ahead with the first phase of construction right after autumn quarter wraps up at the UW (December 14th).

There were a considerable number of repeat questions after Sound Transit’s presentation, and it was definitely neat to see that the plans are a lot more concrete and the plans make practical sense (for the most part!). Questions such as bike access (grade-separation from the Burke-Gillman trail, bike runners for the stairs on the east end), trucking soil and spoils out of the excavation area (modifying the Montlake/Pacific Wy signal), and parking issues dominated the Q&A period.

Then came the UW presentation for the overhaul of Rainier Vista and everything went incredibly insane. The guy gave a decent presentation, although quite sparse on the comparisons with replacing Sound Transit’s arcing footbridge to the south end of C-12 with a 100-ft wide lid over Pacific Place (necessitating the need for regrading the street 17 feet lower) and using a separate signaled crosswalk to cross Montlake at the midpoint between Pacific Pl and Pacific Ave (think the Seattle Aquarium crosswalk). I’ve included a picture below.

It’s tiny at the moment, hopefully a better image will surface in the next couple of days. One of the most important debatable aspects of the design is the crossing at Montlake that’s partially hidden by the lower text bubble. Sound Transit is predicting at least 10,000 daily trips using the light rail at Husky Stadium by 2030. Which one supports this growth? Apparently 25% of riders are expected to make transit connections, and 70% of the remaining 75% will go north to upper campus. The remaining 22% will go to the hospital and points south.

Of course, most of the criticism that came after the UW presentation was the practicality of the change compared to the Sound Transit proposal ($18 mil vs. $6-8 mil) and the complete infeasibility of an at-grade crossing for over 75% of users. The poor UW guy eventually shut down while suggestions and criticisms were voiced against the UW plan, but some important points were raised. One gentleman was particularly concerned with the jaywalking tendancies and the lack of queuing of the university community and questioned the pace of the traffic light counter to pedestrian traffic (“…if the traffic light isn’t responsive to within 10 seconds, it’s not fast enough…“, to which the reply was “…[the light] is fixed to 30-second intervals.“) Many concerns were raised about efficiency; is constructing a 100-ft wide bridge only accessible from the station/bus stops by crosswalk really necessary? The Seattle Transit Blog has a little discussion about pedestrian feasibility in the area, though throughout the presentation it was becoming clear that a pedestrian bridge would not be constructed in conjunction with the Rainier Vista lid.

Neither option supports good connections to the hospital, and it came up through some discussions with other people at the meeting that the medical center didn’t support a dedicated pedestrian right-of-way from the station to the hospital (why? crowd control?). People that work/visit at the hospital will still have to make the two-hop jump to the hospital through Montlake and Pacific.

My suggestion would be to extend Sound Transit’s bridge idea. At the current time, transit planners are envisioning re-routing buses to take advantage of the light-rail transfer and utilize the north side of Pacific Pl, hence the inclusion of a stairwell and elevator in the center of the walkway span. Instead of (or in addition to), there should be a pedestrian ramp from that midpoint toward the center of the vista triangle. While this enables people to populate the triangle (it hardly gets any traffic and can be a decent gathering area), it allows for a one-crosswalk connection to the hospital, bike connection to street level, and access to future bus stops.

Wednesday, November 18, 2009

Transit hacking – polylines

After attending the King County Metro developers conference earlier in October, they awarded each of us with a CD of the current dataset and also a GTFS copy of the data. I had been having troubles trying to figure out metro’s shape data (apparently it was some form of GIS awesomeness), so I was sooo happy to find out that Google’s format has shapes defined by a list of lat/long points.

After adding all the data to a sql server instance, I hacked together a C# program to extract the lat/long points for each shape and encode Google Maps polylines, with the help of an implementation Mark Rambow did in java. That put info back into the database, so I exported it into mysql and put it up on Alpertopia.

It seems to work pretty well! There’s a wacky issue with some routes completely screwing up the map though; so if a route doesn’t display, you’ll have to reload the page. I haven’t pinned down the bug yet.. One of the coolest things here is jQuery, which everyone on the internet says is awesome. I really didn’t like the syntax at first, but my then-roommate Chris was using it in his research and was extoling the virtues of it. Looking at the code of my hacking above, I’ve gotta admit that it’s pretty nice

Those shape IDs correspond to individual trips that the buses take, so that a bus can have more than just two trips (allows for express, weekend, snow routes). That’s the next step, hopefully.

If there’s any interest in the code behind it to generate the polylines, I’d be happy to make it available.

Wednesday, September 23, 2009

New Research & Moving On

Though there’s been a little bit of a hiatus at Alpertopia, we here like to think we keep the ball rolling. We sort of do!

I have graduated from the University of Washington and am now working with Microsoft in Windows Core! It’s an interesting transition as I’ve been in academia for nearly all of my life as the expectations, responsibilities, and obligations are different, and free time is radically different (only in the evening, usually not on my terms!). I’m working with a great group of people, and they’ve been really accommodating to me.

For my senior thesis, I wrote a little bit about quantum computation. You can learn more about it by clicking on the image on the right-hand side or by browsing to the projects page. It went pretty well, but I’ve been rather anxious at how quickly it ended; I feel like I need to actually complete the research in order to tie it off!

I recently trundled down to the SDOT warehouse in SoDo to pick up some surplus signs. I took a couple pictures with my phone and some pictures of the spoils! Greenwood Ave., Sand Pt. Wy., Fremont Ave., and Newton St. all made the cut! See the pictures below or at the Flickr set.

I’m hoping to get a lot of backpacking in this fall, keep your eyes posted for pictures soon.

Tuesday, October 14, 2008

Moving Hosts

Hey all -

I’ll be moving hosts in the next couple of days, so this post will serve as bookmark whether the domain has propagated over to the new site or not. This way, I’ll know it has not if I still see this post on the front page!

Tuesday, July 15, 2008

Türkiye Timez

Welp, it’s been a fun two weeks here in Turkey so far – making the family rounds and visiting the southern Mediterranean coast. So far, I’ve been in and around Ankara having visited Anıtkabir (Atatürk’s resting place), the shops in and around Ulus, the villages (more like cities now!) of Ayaş and Baypazarı outside of town, and seeing many people that I have not seen for the five years since I’ve been here. We’re getting around mostly by taxi and bus, and it seems to be working out so far.

Sunday, May 25, 2008

Just starting to get busy?

Welp, the quarter’s over halfway done and the real work seems to be just beginning. After this weekend, I’m gonna have to hightail my research into aconitase, the fun-loving bioinorganic molecule that loves to use its Fe₄S₄ cluster to facilitate a stereo-specific reaction mechanism instead of just serving as an electron mover.

Not to mention that Folklife is tomorrow! I think I’ll be leaving home with Amanda’s group of friend (see my events on Facebook for that info), but if you’re down there, definitely give me a call and we can go jump in the fountain together. Read more…