The Deliverator – Wannabee

So open minded, my thoughts fell out…

Archive for January, 2006

iunlock

Posted by Deliverator on 31st January 2006

Got my Nokia 6620 back from iunlock. They upgraded the firmware to the
latest “uncustomized” version. I have already noticed a lot of subtle
improvements to the user interface and the system feels quite a bit more
responsive. I am glad to have all the undeleteable garbage that AT&T
includes in their firmware version GONE. iunlock was very responsive in
the handling of my order. They flashed the phone and sent it back the
same day they got it. I would highly recommend them for all your phone
flashing/unlocking needs.

Posted in Mobile Blogging | No Comments »

Too Much Weekend

Posted by Deliverator on 30th January 2006

My weekend was very full and unfortunately it has caught up with me. I have a sore throat and my body aches and shivers. I spent a good deal of the weekend in damp clothes, which I am sure didn’t help. Seattle is a very wet place, but usually the rain follows the pattern of constant light drizzle which is just enough to make your hair frizzy all day. Most people here don’t carry umbrellas and normally I scoff at those who do. This weekend it seems like every time I needed to go outside it was pouring down rain hard enough to get me soaked just going from car to building.

Saturday I caught the ferry and headed to Keyport. Keyport is home of the Naval Undersea Museum, which this weekend graciously provided facilities for MATE to hold their kickoff event for the 1st Annual Pacific Northwest ROV Competition. Roughly 30 people (mainly educators) showed up to hear Jill Zande (from MATE) and Fritz Stahr (representing MTS/UW/Ocean Inquiry) talk about this year’s event and show footage from prior year’s. There was also a representative from the local naval base who showed footage and talked about how they use ROVs currently and how they intend to use them in the grim meathook future. The kickoff was definitely a low key event compared to a FIRST competition kickoff and frankly a lot of the information presented during the day was repetitive and redundant. They could have presented all the necessary information in an hour  or two instead of six (plus ferry rides/waiting at ferry docks). I know it sounds like I am bitching. I really had a lot of fun, I am just majorly crunched for time atm. The highlight of the event was when we all broke into groups to create small ROVs, designed to pick up PVC T junctions off the bottom of a large water filled cow trough. It was a lot of fun and I got a lot of pictures

I didn’t get home until about 7:45, at which point I met up with Ryan at my house and hauled the ROV over to Larry’s. We were intending to spend most of the time at Larry’s working on a nagging encoder issue, but got sidetracked when we realise that the right side drivetrain was locked in place. Turns out that when the gearbox was assembled a locking key designed to hold the driving gear in place had not been inserted. The driving gear was being held in place by a single small set screw. As FIRST co-founder Woody Flowers has noted, “set screws inhale audibly.” Of course, the set screw backed out and the gear it was holding in place slid down its shaft and managed to lock the gearbox up. I am just glad it did so now, rather than during the competition. Thankfully, no damage was done to the gearbox. Ryan (and larry the next day) worked on fabricating a new key to properly fix the gear in place. I disassembled the other gear box to check for the same problem. Thankfully the other one was done properly.

Sunday afternoon I headed to Larry’s and mainly spent time taking pictures. We had more hands than needed for a change. Most of the afternoon was spent fabricating and welding an outer frame and working on the shooting mechanism. The shooter is hitting a pretty small target now and all nay-sayers (including myself) now think this mechanism is going to work out. There is still some mechanical work that needs to be done in working out a good joint feed mechanism for the roller & shooter. The roller mechanism itself still needs to be built (easy as pi according to Larry) and the hopper needs to be extended in the front to the full 5′ limit. Still, the robot is definitely taking shape. We hauled it over to school, where Ryan and Kevin worked on the encoder issues while Erik and Ian ran a competition to determine this year’s human player. It looks like Yevgeniy Mikhaylov is going to be this year’s player. With the addition of Yev, this year’s drive team is now complete. Ian (last year’s arm controller) is taking over as driver. Cheuk is going to be handling the shooter/roller controls and feeding balls to Yev during the defensive period. I am pretty sure Erik Thulin is going to take over as coach, given that he is heading up this year’s strategy sessions. Erik and Ryan have been building the scouting program up for the last few years,  and are hard at work creating STAMP, a web based joint scouting application.

Posted in General, ROV Project, Titan Robotics Club | No Comments »

Wireless Happenings and Robots Stuff

Posted by Deliverator on 28th January 2006

The site for the new book, Wireless Networking in the Developing World, has gone live. Rob is one of the chief co-authors. I have been reading the free PDF version of the book, released under a Creative Commons license. I will be buying the dead-tree edition of the book thru Lulu as soon as all the kinks have been worked out, so that I don’t need a decoder-ring. Rob donates a great deal of his time, money and copious brainpower towards free culture and helping to technically empower people everywhere across the globe. I highly recommend reading this and his other wonderful books.

In other Seattle Wireless related news, the deployment of NodeAtoys on the UPN tower on Capitol Hill went smoothly (afaik) and is now linked into the network. Start pointing those high-gain directionals at the tower and see what you can see!

In robot related news, we finished wiring the control panel up, downloaded code to the robot and drove it around for the first time. We hauled it over to school and let the drive team have a go at it until 9pm. There is still an unresolved issue with the left quadrature optical encoder, but we managed to eliminate some possibilities with an oscilloscope. We fixed a problem with a bad pwm cable that was causing one of the two motor speed controllers for the left side to only operate intermittently, resulting in a power imbalance that was slowing down the speed of the robot. Once that was fixed, the robot moved smoothly and VERY swiftly across the floor. Seems to have plenty of torque as well, as we made the robot wax the floor, using Ryan as a mop! After wrapping things up at school, Ryan, Erik and I headed to Safeway, where Ian had reported seeing foam balls identical to the ones used in this year’s competition for sale. We left the manager somewhat befuddled when we purchased all 28 of the balls they had in stock…and two jars of peanut butter. We tried to explain that the balls were for a robotics competion, but decided not to explain the peanut butter….

Ryan will hopefully post some of the pictures and video from tonight’s goofy proceedings on his site. I left my digicam at home and couldn’t fall back on my Nokia 6620’s phonecam, as I am having the 6620’s firmware upgraded by iunlock. They have been very prompt and it is already on its way back, so I may have it as soon as tomorrow. I wish phone manufacturers would pull their collective you know whats out of their you know wheres and offer an easy way for consumers to patch bugs in their products…

Posted in General, Titan Robotics Club, Wireless | 1 Comment »

ROV Project Update – 012806

Posted by Deliverator on 26th January 2006

The ROV project has definitely been on the back burner for a while now, with the flurry of ongoing activity related to the FRC robot taking priority. I know people are going to be pretty wiped out after getting the robot finished, hosting the unofficial pre-ship event and finally shipping the robot off to Portland. Hopefully we will be able to start working on it again after spring break. While the project has been a bit idle of late, I have not.

This weekend I am hauling my butt out of bed to attend an all day ROV conference put on by MATE in Keyport at the Naval Museum. Among other things, the conference will be an organizing event for the 1st annual Pacific Northwest ROV competition, being held in Washington sometime in the Middle of May. I somehow doubt the TRC has enough stamina for another grueling build season this year, but you never know. At the very least, I hope to make some fundraising contacts. I might try and find a group of kids from one of the ther local schools that would be interested in giving it a shot.
Helena Wu from Lumileds emailed me this morning to inform me to expect a package in the mail shortly containing all sorts of wonderful high wattage LED emitters. It sounds like they were able to supply everything that i requested, except a few parts that had been discontinued, which they replaced with newer models that put out even more Lumens. It sounds like adequate lighting will not be an issue on our ROV. Many thanks Lumileds!

I need to find some Victor 884 or equivalent motor speed controllers for the ROV. We had two 884’s and an older 883 allocated to the ROV project, but we were running short. so I gave them to Larry to use on this year’s FRC robot. This year’s robot will be using SEVEN 40 amp motor speed controllers, if you can believe it. That battery is not going to last long! Anyways, if anyone has some 12v motor speed controllers to donate to the project, they would be much appreciated.

Posted in General, ROV Project, Titan Robotics Club | No Comments »

Is NPR trying to boost its google rank through comment spam?

Posted by Deliverator on 26th January 2006

My comment moderation que has been remarkeably free of spam of late, due to a rather long list of forbidden words that I have compiled. I am not a big fan of auto-censorship, as the chances of a false positives are rather high. I have tried to be careful in choosing words unlikely to have legitimate usage. So far, the only downside from this draconian censorship is that you will no longer be able to discuss viagra or party poker on my site. I know, it is a big loss :(

I still clear out my moderation que a couple times a week. Today, I noticed something rather interesting. I received a number of pieces of link heavy comment spam. Rather than advertising new and improved herbal viagra substitute, these pieces of spam were linking to Mercury New, NPR and other news outlets. Are the big media outlets soo afraid of blog coverage of news stealing their thunder that they are resorting to comment spam to artificially boost their rank?

Posted in Blogging, General | 2 Comments »

Seattle Wireless Hacknight – 012506

Posted by Deliverator on 25th January 2006

Headed to Seattle a bit early today. I was originally going to bring along Larry Barello to talk with Matt and Rob about various OEM wifi modules suitable for use in some very hush-hush embedded systems, but he was feeling rather tired after spending the last few weeks in his cold garage working on this year’s FIRST robot. I will have to drag Larry along next week. I arrived almost an hour early, so I spent some time talking to my brother Scott while waiting for others to arrive, which they soon did – and in large numbers. Tonight’s hacknight ended up being the most populous I can remember in recent history.

Rob showed up with copies of his new book, Wireless Networking in the Developing World, hot off the presses from Lulu. Lulu is one of the first companies to do “Just In Time” publishing. Got something you want printed? Send em a PDF and $10 and a week or so later a nicely bound dead tree edition magically shows up at your doorstep. That is the theory, anyways. Unfortunately, Lulu wasn’t too happy with one of the chain of free tools Rob used to create his PDF, in particular the embedded fonts. Rob ended up with a beautifully formatted book filled with gibberish text that bore a distinct resemblance to Kanji. Rob spent a good chunk of the evening using Lulu’s “live chat” technical support to get the problem fixed before anyone did something rash – like order 150 of the things :)

Rob also showed off his $100 Wi-Spy 2.4ghz frequency analyzer. It is now supported by Kismet in text mode, and graphical support for OS X is otw. Rob tried to get an early version of the graphical version going, but it kept exiting after running for a second or so. We ended up borrowing Casey’s laptop and played around with the official Windows software. It was pretty cool to “see” bluetooth and get a sense for just how much noise is floating around the 2.4 ghz band. At $100 it is a fun toy, although it would be much more useful with an external antenna port.

Matt has announced that the new node on the UPN (now UPN/WB?) tower will be put up this Friday at 8am. I am very enthused about this new node, as it is the first with real potential to link the Eastside into the network. I may even consider climbing the huge tree in the front yard to get line-of-site. Nothing is going to drag me out of bed at 8 am, though…

Casey showed up and showed off his paper wireless sliderule and spent most of the night hunched over his laptop working on a Excel spreadsheet inputting specs on various pieces of wireless gear. He seemed a little withdrawn. I hope he is doing alright.

Matt Wilson showed up in his brightly colored motorcycle garb and brought along some interesting retro hardware which consumed most of his attention for the evening. The chief focus of his attentions was an Apple eMate 300, a large clamshell computer designed for use at the gradeschool level.In concept it is somewhat similar to the Branium Wibook that I brought to Hacknight some weeks back. Matt Wilson brought the eMate in because it late model version of the OS powering the Apple Newton PDA. Matt Wilson is a big fan of his Zaurus PDA, for which a newton emulator is available. Matt spent most of the evening searching poorly organized Newton related archives in search of a debugger. He eventually found a suitable debugger, although it needed to be run from OS 9. Not to be outdone, Matt wipped out an old powerbook (upgrade to a g3 running at ~200mhz on a 33mhz bus). By the end of the evening he had managed to download the rom image to the old OS 9 running Powerbook, transfer it over to his much nicer OS X Powerbook and get it running in an emulator. Although he carried a very sizeable kit of tools on his motorcycle to pry the OS image off the eMate, he forgot his very pocketable Zaurus at home, so that part of the project will have to wait till later. I have collected quite a few obsolete…, make that “retro” pieces of hardware over the years, so I definitely can sympathise.

Spent some time conversing with Schuyler Erle, one of the co-authors of O’reilly’s Mapping Hacks and co-coder of the Nocat captive portal system. Schuyler is visiting Seattle for a week from Boston, courtesy of MSN. MSN has been trying to bolster its map related offerings, after taking a beating from Google Map/Google Local/Google Earth, etc. I should have taken the opportunity to pick his brain about cartography, a subject on which he is quite an expert, but had some interesting conversations with him about a variety of other topics instead. I enjoyed talking with him about an abortive project of his, a p2p filesharing system designed with the network limitations of wireless networks in mind. One of the chief goals of his project was to avoid the huge drop in bandwidth associated with many clients attempting to transfer large files simultaneously, resulting in frequency contention/packet collisions. His system was somewhat similar to Token Ring, implemented at the application layer. I introduced him to Eric Butler, who has been working on a WASTE like private p2p network called Meshwork. The rest of the evening was spent discussing MONO development and the virtues of various programming languages. Oh, btw, in case you haven’t heard, Microsoft Sucks :)

Pics here.

Posted in General, SWN Hacknight | No Comments »

Super Cheap 802.11a Access Points

Posted by Deliverator on 24th January 2006

802.11a is a wireless standard that hasn’t made as much of a splash as 802.11b & g gear. The 5 ghz band that the 11a standard uses has much less penetrating power than the 2.4 ghz that b/g equipment uses. Buildings love to eat .11a signals, requiring great number of access points to be deployed to cover a given area than with b/g gear. As such, 11a gear hasn’t caught on to as great an extent and is now available cheap through many surplus outlets. Given 11a’s inability to break out of a paper bag, why would you want it, even if it is available cheap?

There are a couple of good reasons, in my mind. For one, 11a gear is just as fast as g gear. Two, 11a gear uses frequencies that are not as heavily used as the 2.4 ghz band that is used by cordless phones, security cams and wifi gear. It is getting awefully hard to find free unlicensed spectrum in urban environments. I have personally been in places where my carputer has been able to see 30+ wireless networks simultaneously. That is a hell of a noise floor! Thirdly, because 11a gear uses such high frequencies, the wavelengths used are correspondingly shorter than b gear. That means you can create much smaller, high gain directional antennas. 11a gear might not be such a bad idea when you need to establish a high bandwidth connection in an noisy urban environment.

At last week’s Hacknight, Joe Towner brought in some corporate quality 802.11a access points made by Intel. He bought them for $4 each from surpluscomputers.com. I purchased three to play around with using my Jornada and internet access courtesy of t-mobile. Half the guys there were running packet sniffers and Ken was feeling particularly hacktastic after just getting back from Shmoocon (of which he is an organizer), so I wasn’t about to trust my credit card number to the unsecured wireless network at RedLine. Of course Ken just went to work for t-mobile (doing network security), where he will hopefully help close a few of their gapping security holes, so he could probably sniff my data anyways.

The access points arrived today, which was rather quick for such cheap shipping. I have been playing with it this evening. Initial thoughts:

  • Cool = software selectable antenna pattern (omni or 180 degree sector)
  • Not soo hot = need tftp server to flash the firmware. I used a free one from the aptly named weird solutions.
  • Cool = Has two mini-pci slots inside. One is filled by an Atheros mini-pci 802.11a card. The other can optionally be filled with an 11b card to turn the access point into dual mode ap.
  • Cool = Early reports show a high potential for hackability. The device has 16 MB ram and 8 MB of flash and a Motorola processor of some sort.
  • Not so much = kinda funny shape for an AP. Indicator lights unlabeled.

Posted in General, SWN Hacknight, Wireless | 1 Comment »

FRC 2006 Season – 4th Build Meeting – 012206

Posted by Deliverator on 22nd January 2006

Today was a day for prototyping shooting mechanisms. The shooter that Ian’s dad brough in at the end of last meeting was significantly modified since I last saw it. The biggest addition is a very effective rubber belt mechanism that conveys balls up to the shooting mechanism. We pulled the shooter outside the garage and used some electric drills to provide the power. The shooter could certainly chuck balls quite a distance, although not with a great level of accuracy at extreme distances. One of the rules of this year’s competition is that balls can travel no faster than 12 meters per second at exit, so we are more concerned with how accurate we can be given a lesser initial velocity than how far we could chuck a ball at full tilt. Larry and Dave were busy working on their “single wheel” shooter this morning, prior to my noonish arrival. They managed to construct a frame out of 1″ extruded aluminum and get the shaft and wheels mounted (they are actually using two small wheels acting on one side of the ball, so I am calling it the single wheel design). The belts and pulleys we need for utilizing the “big sim” motors have not arrived, but hopefully we will be able to test it soon. We will have to order some additional belts in order to test the two wheel design. Ultimately we will go with whatever works best, but I want ever reasonable idea to get tested prior to making any sort of irrevocable design decision. Larry welded the frame up and he and kevin bolted all the electronics to a temporary board. Kevin took the board and wires with him to school afterwards, so the electronics board may be assembled by now. Cheuk and I went to Loew’s hardware and picked up some screws and some speaker wire, neither of which was deamed appropriate, despite being exactly what was requested. I guess the speaker wire needed to be red and black to go with some wiring rule. Nobody bothered to tell me to get anything other than #12 gauge speaker wire, so I guess I am just out $12 for nothing. I am going to be ordering a bunch of ultrasonic sensors from Maxbotixs and some more belts to mock up the two wheeled shooter design that Ryan and I have been working up. I just hope I can get the particulars straight before placing the orders, as the ultrasonics and timing belts are significantly more expensive than a few screws and some wire.

Here are some pics of today’s meeting

Posted in General, Titan Robotics Club | No Comments »

3rd FRC Build Meeting – 012006

Posted by Deliverator on 20th January 2006

Today was the third build meeting for this year’s FRC robot. I arrived late and managed to pinch a finger pretty hard between my knee and a metal plate. I got a little shocky there for a bit. I was sweating for no apparent reason and my hearing was going in and out. I sat things out for a bit. By the time I was feeling ready to work, three more kids had arrived, so I let them do the hands on stuff and I just took pictures and generally got in the way. The kid’s managed to get the motor mounting plates attached and the motors in place before Larry called it a night. We made progress and the drive-train is now more or less finish. Still, I somehow feel we are moving more slowly than I would like. Hopefully we can catchup this Sunday with an extended session. Ian’s dad stopped by at the end of the meeting to show off a prototype shooting mechanism that he built. Here are some pictures of today’s meeting.

prototype shooter

Posted in General, Titan Robotics Club | No Comments »

Rockbox on my iRiver H320

Posted by Deliverator on 19th January 2006

I have mentioned Rockbox at least once before in a previous entry, but was always too scared of bricking my beloved iRiver H320 mp3 player to take the plunge. Well, I have been trolling the MisticRiver forums and it seems that the recent bootloaders are much improved and there is now very minimal risk of bricking one’s player. I went ahead and patched the 1.29K firmware for the H320 with the bootloader and installed the latest daily build onto my player. I powered up my H320 and it booted smoothly into Rockbox. The default interface is very minimalist, but is fully skinable. The MisticRiver crowd has gone ahead and created their own custom version – H300 Optimized which is updated often. There is even an automatic update tool to keep your player up to date. Rockbox on the H320 is definitely a work in progress, but it has a number of capabilities not found in the official firmwares including gapless playback and on the fly playlist creation. The new version 5 bootloader can dual boot into either Rockbox or an official firmware at powerup, so one can have the best of both worlds. So, give it a whirl and help the developers release a polished 1.0 version sooner by submitting bugs.

Posted in General, iRiver, Portable Computing/Gadgets | No Comments »