Hugo’s blog

J’ai reçu ce matin un petit email de Monsieur Bayrou, qu’il a appelé Lettre à mes compatriotes expatriés. Voici l’email en question :

C’est le premier à appeler les expatriés à voter pour lui. Je me demande s’il sent qu’il a une chance d’être au deuxième tour et que ça peut se jouer à pas grand chose. Je dois dire qu’en étant expatrié, on ne sent pas forcément que sa voix compte pour grand chose. En 2002, j’avais voté alors qu’ils avaient déjà annoncé que Chirac et Le Pen seraient au deuxième tour. Du coup, ça fait presque plaisir de recevoir du spam.

Je me demande quand même comment il a eu les adresses emails des expatriés. Deux solutions : la base de données de l’ambassade, où bien celle du Trésor Public. J’imagine que c’est la première, mais je me demande qui peut mettre la main dessus et sous quelles conditions.

Ce qui me rappelle que Chirac, en 2002, s’était également adressé aux expatriés pour leur demander leur soutien, comme il dit. Lui l’avait fait par la poste. Mais, comme c’est Chirac, il l’avait non pas fait à ses propres frais, mais par l’intermédiaire de l’ambassade. On ne se refait pas.

One thing that I really like about having lived in several different countries is that I have been exposed to a lot of different music, and that great artists that everybody knows in certain places are sometimes completely unknown in others. This has allowed me to see Ani DiFranco in a venue of at most 200 people in Paris, Ben Harper in a small club in Cambridge, etc.

Yesterday night, I was introduced to the Fillmore, which fits about 1000 people I’d say, to see The Tragically Hip, which is a massively popular Canadian rock band… in Canada. Gerald had taken me to one of their shows years ago. Since then, I’ve been listening to The Hip.

It was pretty much the same crowd: I was surrounded by Canadians. Some came all the way from Calgary in order to see The Hip play in such a small venue.

The support act was great. The Sadies put up a great show. They somewhat reminded me of Tarnation, but with a “rockier” sound, if that says anything to somebody else. They’re themselves Canadian, and who knows when I’ll have a chance to see them again, but I’m definitely going to get some of their CDs.

Then The Hip did a great performance. As I often realize, I appreciate a concert much better when I’m close (I remember not fully appreciating PJ Harvey at the Zénith in Paris, while I loved the bootleg I got of the show). Here I was at the second row, and I was able to witness the energy that Gord Downie puts into the show.

He basically doesn’t hold back. His shirt was soaking wet after the second song. He broke a guitar string during the second song, jumped all over the place for 1h45, gave away guitar picks, broken strings, towels, his microphone stand (yes, you read well). It basically was one of the best shows I had seen in a while.

The setlist is available in the HipBase forum. I have some photos on Flickr. And I tried to take a video of Blow at High Dough as you can see below:

I edited out the audio, as it was horrible, and you could not even recognize a melody. The result, without the sound isn’t great. You can watch the whole song on YouTube from another concert.

After a great performance, we were given when we exited the venue a pretty big poster for the concert, made of very thick paper. All I can say is that The Hip gives you plenty for your money, and even more. Don’t miss them.

The change of Daylight Saving Time in the US actually had positive aspects: when our GPS showed the wrong time, I started looking up for a firmware update, and discovered that it had a somewhat ancient firmware. I got this geek feeling of excitement that you feel that you are going to upgrade the software on a device to the latest version. Come on, you know the feeling.

I noticed in the changelog the following:

• Improved Traffic Interface

Until now, I had been pretty disappointed by the traffic interface: basically, at some point while driving, it would say “Caution, traffic ahead”, and display an icon. At this point, you might already be stuck in traffic and unhappy. Clicking on the traffic icon, you could ask it to avoid the traffic problem in question, with no idea of how much a problem it was and whether it really was a good idea to do so. Basically, it was useless, and frustrating.

Version 5 changes all that. It evaluates the delay that the traffic has on your itinerary, and takes this into account to choose the best route. It then displays, in the traffic icon, the number of minutes that you’ll waste because of traffic. And you don’t need to ask it to avoid anything, as it has already taken it into account. Basically, it’s traffic in a GPS system the way you’d like it to work.

So the question was: how efficient is it? Well, I did a real life test last night. Between Sunnyvale and downtown San Francisco, at 6pm, took me 50 minutes without ever hitting any traffic. Though you can argue that Mondays are usually not as bad as other weekdays, I was impressed by the route it made me take, and by the result.

The bottom line is that I’m now very happy with the traffic system. Kudos to Garmin.

Look at what’s happening in France:

I had been complaining for the past few months that my wireless card did not work right on my MacBook Pro. The simptoms were the following:

• At home, my Mac would not detect automatically my network, even though it was in the preferred networks list. I always had to pick it by hand.
• Regularly, it would not connect to certain networks, giving me an error. Clicking on Try again would sometimes work, and sometimes my only solution was a reboot.
• I would get poor performance. Pinging a maching would get me sub-second delays with a very regular pattern:
  

  PING xxx (x.x.x.x): 56 data bytes
  64 bytes from x.x.x.x: icmp_seq=0 ttl=58 time=3.927 ms
  64 bytes from x.x.x.x: icmp_seq=1 ttl=58 time=1805.443 ms
  64 bytes from x.x.x.x: icmp_seq=2 ttl=58 time=805.355 ms
  64 bytes from x.x.x.x: icmp_seq=3 ttl=58 time=4.002 ms
  64 bytes from x.x.x.x: icmp_seq=4 ttl=58 time=4.861 ms
  64 bytes from x.x.x.x: icmp_seq=5 ttl=58 time=3.775 ms
  64 bytes from x.x.x.x: icmp_seq=6 ttl=58 time=2000.225 ms
  64 bytes from x.x.x.x: icmp_seq=7 ttl=58 time=1000.141 ms
  64 bytes from x.x.x.x: icmp_seq=8 ttl=58 time=4.067 ms
  64 bytes from x.x.x.x: icmp_seq=9 ttl=58 time=1805.541 ms
  64 bytes from x.x.x.x: icmp_seq=10 ttl=58 time=808.282 ms
  64 bytes from x.x.x.x: icmp_seq=11 ttl=58 time=45.774 ms
  64 bytes from x.x.x.x: icmp_seq=12 ttl=58 time=3.702 ms
  64 bytes from x.x.x.x: icmp_seq=13 ttl=58 time=3.924 ms
  64 bytes from x.x.x.x: icmp_seq=14 ttl=58 time=1998.669 ms
  64 bytes from x.x.x.x: icmp_seq=15 ttl=58 time=998.551 ms
  64 bytes from x.x.x.x: icmp_seq=16 ttl=58 time=3.601 ms
  ^C
  --- xxx ping statistics ---
  17 packets transmitted, 17 packets received, 0% packet loss
  round-trip min/avg/max/stddev = 3.601/664.696/2000.225/778.601 ms
  

• I would occasionally get disconnected from networks.

IT looked at my Mac, ran hardware tests, and did not find the issue. I looked online, and did not find anything obvious, until I finally stumbled upon this lengthy article giving tips for any wireless issue that you might encounter on a Mac.

Near the end of the list (after many other things I tried), there is: Check for/quit third-party wireless applications (monitors/sniffers). And this was my problem: a simple widget showing wireless information, and nothing more. Not only that, but I installed this so long ago, and considered it so innocuous, that it took my a while to even realize that I had something in this category.

I’m amazed that polling the network interface would have such consequences, but it obviously does. Needless to say, I’m much happier about my Mac now, though I have noticed that Intel-based laptops are far less stable than PowerPC-ones: my PowerBook would never crash, while the MacBook Pro does occasionally. Unless it’s another harmless piece of software messing up the system…

Jonathan announces on his blog that WSDL 2.0 is moving forward to Proposed Recommendation (once the W3C Director approves it). Congratulations to Jonathan, who has been doing a great job maintaining the group on track to get there, and everybody who participated in this.

Looking back at the entire adventure, I think that a pivotal moment was about one year ago, in Mandelieu, when things were looking pretty grim and we managed to stir implementation interest around good French food.

Improving interoperability

WSDL 1.1 has been an endless source of interoperability issues. The WS-I Basic Profile has helped things a lot, however, it’s only constraints on top of a very loose language that are not enforced at the syntax level. Developers can still do whatever they want, and fall into a number of hidden traps. Along with the many changes and improvements over WSDL 1.1, WSDL 2.0 does one very important thing: it enacts, right in its language, constraints similar to the ones defined by the WS-I Basic Profile. That means that WSDL 2.0 can be used to describe current interoperable service descriptions, and pretty much only those, as will show the WSDL 1.1 to 2.0 converter.

As pointed out at the W3C workshop on Web services last week, interoperability is the number 1 issue that we face today with SOAP services.

The nearing completion of WSDL 2.0 is therefore good news for SOAP Web services: it should improve interoperability. Of course, in order for this to happen, WSDL 2.0 will need to be broadly adopted, and that will take a while. However, WSDL 2.0 is a move in the right direction.

And now on to the next interoperability issue, which is not limited to SOAP services, but also affects HTTP ones: XML Databinding.