Monday, December 24, 2007

Almost done painting

The walls are finished. Only the window frames and doors need work now. This allowed me to do the flooring. Here is a picture just before I finished the floor.

Now I am waiting for some furniture I ordered to arrive so I can setup my office again and get back to work. I need to do some testing of the Angstrom release candidates on the OMAP starter kit. I also need to get Angstrom running on the Xilinx ml403 evaluation board in the next few weeks.

Monday, December 17, 2007

Painting

I hate painting.

I'm remodeling the room in our house I use as an office. This needed doing several years ago when we moved in, but things were to chaotic at the time. The room has old wood panelling, the kind used to finish a house when you were running out of time and money in the late 50's.

Painting over this panelling will brighten up a room dramatically, but the finish tends to bleed through latex primers badly, this explains why the project is taking so long. I am up to four coats of primer now and ready to apply the final coats.

After painting, I am laying down fake wood floors. This should go much better than the painting work.

I ordered a cube style desk with peninsula this should give me plenty of desk space to fiddle with small embedded systems :) Or more space to clutter.

Hopefully, after Christmas I am back to full speed and have found some work to pay for all this.

Friday, December 7, 2007

Via Ferrata

I'm scared of heights. So is my friend Dave. See Dave climb a ladder on a via ferrata in Italy.



But we like to do exposed things anyway. There are no shots showing how high we are, because I can't look down. That is how I cope in "airy" situations.

Wednesday, December 5, 2007

Defense

This sums things up.

Sunday, December 2, 2007

Defense

Tuesday December 4 at 10 AM (EST) I have the defense for my MS degree.

Cart/horse issues

This came around in an email a couple of weeks ago:

I would like to take a moment to congratulate and thank everyone for
all the hard work on the 0.6.2 release. During the meeting today Ben
had suggested that we do something to recognize this momentous
occasion. As head of the "attend wine and cheese" group I would like
to propose that we have a technical meeting tomorrow (Thursday) at
PKs to celebrate the release.


Of course, the software release the celebration was for did not actually happen. Apparently, the celebration occurred anyway. Eventually, the software release was rescheduled for Dec 8.

Wednesday, November 14, 2007

SDR Forum Technical Conference

I spent last week at the SDR Forum Technical Conference. I spent the week talking with people I know about various SDR ideas and "networking". Highlights of the week were our Radio Challenge Team received an award for beat design, I received an Outstanding Technical paper award (I am not really sure why). Overall a very good week.

There were several presentations on partial reconfiguration of Xilinx FPGA's. It looks like the technology is close to being usable. When I get a little more time, I need to learn FPGA programming.

After the Conference, I went skiing at Arapahoe Basin with my friend Mike.

And yes, I need to create more blog entries.

Wednesday, October 24, 2007

I'm still here

Currently finishing off "the thesis". The abstract should be finished this afternoon, then another review by my advisor. Then I'll give it to my committee and schedule the defense date. Of course, one of my committee members left during my eternity here, so I need to replace him.

I'm procrastinating by running builds for the SFF SDR board including some fixes I made in OSSIE, and a build for the gumstix verdex + netwifimicroSD so I can test this hardware setup. I need to get the configs for the SFF SDR board split from the Davinci EVM and push the Verdex+netwifimicroSD changes.

The week of November 4-9 I will be at the SDR Forum Technical Conference in Denver. Big plans there involve looking at skis and doing some skiing after the conference. Oh yeah, probably some work related stuff also.

Thursday, October 18, 2007

Lyrtech Small Form Factor Board

I have had a Lyrtech Small Form Factor SDR (SFFSDR) board since May 9. I had about 8 weeks to get Linux running on it for the SDR Forum Radio Challenge contest. Sadly I could not make it run Linux in that time frame, so we switched to plan B for the contest. Since my thesis writing is nearly done, I have some time to see how good the davinci support committed to u-boot a couple of months ago is, I took another stab at getting Linux on the board

The davinci support in u-boot is good, I had the board working in a day. Some minor kernel config changes have the kernel booting, and this morning I set up root over NFS and got minimal-image running. It was great seeing the Angstrom ASCII logo scroll onto the screen.

Now to get an OSSIE image built .... And get the NAND flash working with the kernel so I can create a self contained system.


root@davinci-dvevm:~$ cat /proc/cpuinfo
Processor : ARM926EJ-S rev 5 (v5l)
BogoMIPS : 148.27
Features : swp half thumb fastmult edsp java
CPU implementer : 0x41
CPU architecture: 5TEJ
CPU variant : 0x0
CPU part : 0x926
CPU revision : 5
Cache type : write-back
Cache clean : cp15 c7 ops
Cache lockdown : format C
Cache format : Harvard
I size : 16384
I assoc : 4
I line length : 32
I sets : 128
D size : 8192
D assoc : 4
D line length : 32
D sets : 64

Hardware : DaVinci EVM
Revision : 64460000
Serial : 0000000000000000

Thursday, October 11, 2007

Back from Berlin

The final day of OEDEM was highlighted by the dinner sponsored by Tarent. I had a very good pork dish and dessert. I am not sure if the entree was "German Food", but people suggested the dessert was Austrian.

After dinner I walked back to my hotel along Bernauer Strasse, paralleling the former Berlin Wall. Walking past the wall gave me time to reflect on the changes that have ocurred during my life time. I was born in 1961, shortly before the wall was built, I am very to glad to see Berlin recovering from this tragedy.


The trip to Berlin for OEDEM has been a very good thing. The OpenEmbedded developers are excellent people, both on a personal and a professional level. The level of technical skill work ethic rivals anything I have seen in my various careers. I strongly recommend people find an open source project to participate in . Even if you are not a coding god, open source projects need users, testers, people to help track problems and most importantly, people that can write.

Many of the issues discussed at OEDEM could be traced back to inadequate documentation.

Wednesday, I had to catch a plane at 0930, this meant getting up at 0600. The airport was a combination of a subway and bus trip that went smoothly. The connection at Frankfurt was short, but went smoothly. I arrived at the gate as the plane was boarding. There was some chaos as the USAir staff rechecked our passports and walked us through the security questions again. Then I had to get new boarding passes. Finally made it on the plane.

The Airbus A330 has an on demand inflight video system. This made watching movies on the plane tolerable.

Back in Charlotte I had dinner and found out there is free wireless in Charlotte now. This made up for the ridiculous security line I had to go through after clearing customs. Note, if you plan to go through security in Charlotte, add one hour.

Before going home, I stopped by my office to pick up my thesis after my advisor made comments. And finally headed to bed.

Tuesday, October 9, 2007

News Flash

Real German food located. Koen bought me a Currywurst at lunch.

OEDEM Day 3

Today the hard work started started. The OpenEmbedded project is growing and attracting new user communities. Many of the issues discussed revolved around the problem of supporting and integrating new users into the OE community. We really should try streaming video for people who cannot be here, but all we have are minutes.

In my late update last night, I covered yesterday's eating details. Still no German food. There are calls to attempt a German lunch. Possible my German food experience will be a currywurst in the Frankfurt airport tomorrow on my way home.

Tonight we have a dinner sponsored by Tarent, a German company that uses OpenEmbedded. I am not sure where this is (I heard the name last night), but maybe it will be a German restaurant.

On the bicycle situation, our Dutch representative assures there are not very many bicycles in Berlin and I should go to the Netherlands to see bicycles. I also saw a recumbent when I went for coffee this morning. Our Dutch representative also try to report that he lost his wallet and passport on Friday in the train station. Apparently, the German police told him that him being an idiot was a problem for the Dutch authorities, not the German police.

Monday, October 8, 2007

Food update


Lunch: Italian.

Dinner: Mexican.

The Ad

I see my blog advertises Prism Tech. I am familiar with the company through my SDR work. They seem to hire all the big shot PHD's in the field, who attend all the right conference's and workshops. I find this moderately amusing. From the outside, I think they are top heavy with too many PHD's and not enough guys to get any work done. But this is probably not true.

OEDEM Day 2

First, the food update: Lunch Turkish, dinner "Restaurant Patagonia". Dinner was a combination South American Steak house, that served pizza and pasta. The steak was good.

I spent most of the day working on the gumstix computer I have. I am pretty good at the "real" embedded stuff (no display), but the screen stuff is all black magic. So I have learned a lot about this so far. The gumstix computers consist of a processor board with connectors on the front and back. So you can attach up to two other boards. Given you have peripherals requiring drivers spread across three boards, figuring out how to build a kernel and what modules are needed across three boards can drive you crazy.

Most of this is handled with kernel modules, but we think the wifi system uses different GPIO lines depending on which daughter board you have.


We have the display basically working and are trying to figure out what driver we need for the touchscreen. The wifi module loaded, but I left the antennas at home. Fortunately, Koen has a Neo he has taken apart and we took the GPS antenna from that. The GPS antenna marginally improved reception so I tacked a wire on the patch antenna board, so basically we use the GPS antenna for the connector and cable. Now we can see many access points. Still can not connect to the one we are using, but this will likely be easy after some sleep.

Sunday, October 7, 2007

OEDEM Day 1

Day 1 we met at Holger's office and worked on projects. I guess this is a form of "extreme programming". I worked on getting the display running on my gumstix computer. Since the board I have for networking is very new, this meant a lot of fixing hardware issues. We have the display running, we believe the micro-SD card is very fussy about which cards it reads. I sent a debug log to the gumstix list, hopefully that rings some bells with people.


For lunch we went to an Italian restaurant and (mostly) had pizza. For dinner, we ate at an Indian restaurant. So far the food has not been expensive. I suppose it helps we are not in an area frequented by tourists, unless they got lost looking for the nearby attractions.

Today there is more hacking. Hopefully, I can get enough software on the gumstix to draw something other than noise on the screen.

Saturday, October 6, 2007

Berlin

After arriving at my hotel, I showered and jumped on the Internet. Rapidly found Holger on #oe and we made plans to head to the train station to meet Koen. We arrived at the train station and found his express from the Netherlands was canceled. A phone call and we found he would not arrive for a few more hours. Then we called Graeme and Rodrigo to find they were on the way to the hotel from the airport and would meet us a Holger's office in a while.

Holger took me on a short tour of Berlin. The Central train station is brand new, built roughly where the former Berlin wall stood. We walked toward the Brandenburg Gate crossing the site of the Berlin Wall several times. There is a line in the roads marking the site of the wall in places. We passed the River Spree on the way to the Reichstag. Then on to the gate. As you can see, no more wall and the tourists


After this we walked back to the office, passing the American Embassy. Of course they blocked of the road in front of the embassy and had a fence with "No photgraphs" signs. I need to go back and take a picture, when I have my passport on me ... They also ask that you walk your bicycle past the embassy.

Walking their bicycles past the embassy must be really annoying to the locals. There are bicycles all over. Several times we have almost been hit by cyclists on the sidewalks. Generally, they announce they are coming in German. Since only Holger speaks German, the warnings are lost on the rest of us.

Dinner was at a Mexican restaurant. Graeme and I could both understand the menu. By now, I was near collapse from lack of sleep and a couple of beers. Holger and Harald explained how to get back to the Hotel via subway and also commented the street cars would be better. But the subway was easier. After looking at the large map with streetcars in the subway, I exited the subway after a stop and rode the street car back the hotel.

This mornings German lesson: Balzac is German for Starbucks.

Now it is time to shower and head for OEDEM 2007.

Flying

Flying to Berlin was good. No security horror stories. I did have to go through security again in Frankfurt. No shoe removal in Frankfurt, but the German security guys actually seem serious, lot's of hand wanding and bag picking apart. I was fine though, I pretty much avoid wearing anything metal while traveling and try to keep funny looking stuff out of my carry on.

Frankfurt Airport is a bit of a mess, I walked all over getting from Terminal C to A. The highlight was the tunnel from B to A, lovely rotating coloured lights.

Berlin/Tegel was fine. The buggage claim is at the end of the jetway. The strange airport experience: I saw a booth for an Iranian airline. Not quite as strange as when I realized why the "Vacation in Cuba" signs in Toronto looked funny to me.

Took the bus to the subway, and subway to my hotel. I should have got off at the first subway station ...

Coming from the airport to the Hotel, I must have crossed the former Iron Curtain a couple of times. More on this later.

Thursday, October 4, 2007

The Ad's

I've added the google adsense ads to the blog. Just in case this gets popular.

I also get a real kick out what ads appear on web pages. I find it very useful for keeping track of people doing work in my space who run ads on google. I use my gmail account for the gnuradio email list, it can be pretty surprising (to me at least) which companies are running ads.

Tuesday, October 2, 2007

Chumby

Yesterday, I received an invitation to buy a Chumby before it goes on sale to the masses. I'm not sure what you do with a Chumby yet. I'll be happy if I can get it to play NPR in the mornings and replace the clock radio.

I'll keep you up to date with the chumbification of my house.

Monday, October 1, 2007

This week

The report for the SDR Forum Radio Challenge is done. The "thesis" needs some review and should go to my advisor by Wednesday. Why Wednesday?

Well Thursday I leave for OEDEM in Berlin. I got mixed up with OpenEmbedded (OE) when I started looking at getting OSSIE running on the OMAP Starter Kit. At the time, it was an easy way to create cross compilers for developing software for the OSK. After the past couple of years, I have gotten better with OE to the point where I actually contribute a small amount of work to the project. The OE guys are a really great crowd of hard working nerds so this should be a really fun trip.

The hotel claims to have free wifi, so hopefully I can post from Berlin.

Saturday, September 29, 2007

Beer day on campus


Today is yet another home football game at Virginia Tech. I do not mind the concept of college sports, but at Virgina Tech, the game takes over the entire campus. People are actually encouraged to drink alcohol on campus, which is a 180 degree policy reversal from every other day of the year. The power of the almighty dollar.

Friday, September 28, 2007

Bitter grad student

This would be the blog of a bitter grad student, but instead I will finish my MSEE shortly and get on with my life. For the past many years I have been happily chugging along in grad school doing interesting software defined radio work. At times, it has been a lot of work.

Sadly, our group (for various reasons) has acquired a student who does not have a sense of community and focuses only on his personal goals and does not share in the larger project goals.

Recently, we made a commitment to support a company who did a presentation of our work at an industry meeting. I helped the company get our work running on a gumstix embedded system and we hoped to show the debugging tools this student is supposed to be working on monitoring the system. I had been told this would work, when it did not work, I was told this was not his problem. After some emailing people trying to figure out who should solve the problem, he appeared to show a little interest in the problem.

This email from the student in question to his advisor is why I am desperate to escape this lifestyle:

ADVISOR

My schedule for today:

8-10 In the lab
10-11:30 Engineering expo
11:30-12 Lunch
12-12:30?
12:30-3:30 class
3:30-4:00 available as long as COMPANY does not show up early
4-5 giving COMPANY a tour of the lab
5-6 Hughes information session
5-5:30 ?
5:30-7 Engineering Expo wine and cheese reception
7 Going home

I am obviously very busy, but I will do my best to help out during any breaks I manage to squeeze in.

Have you by chance asked SPONSOR about the error (or lack of an error for that matter)?

STUDENT