On Februrary 16th, in an interview with Noam Chomsky, the interview is one of Chomsky's best.
At one point he speculates about the potential strategy being applied to Iran:
[...] So it could be that one strain of the policy is to stir up secessionist movements, particularly in the oil rich regions, the Arab regions near the Gulf, also the Azeri regions and others. Second is to try to get the leadership to be as brutal and harsh and repressive as possible, to stir up internal disorder and maybe resistance. And a third is to try to pressure other countries, and Europe is the most amenable, to join efforts to strangle Iran economically. Europe is kind of dragging its feet but they usually go along with the United States.
This week Seymour Hersh publishes
on the New Yorker some of his findings. The article makes
the case that the US is now funding
terroristsfreedom fighters in Iran to
destabilize the regime (as speculated by Chomsky before).
The article also happens to match some of Chomsky's
For a quick overview you can watch this video interview with Seymour Hersh.
It’s very hard to predict the Bush administration today because they’re deeply irrational. They were irrational to start with but now they’re desperate. They have created an unimaginable catastrophe in Iraq. This should’ve been one of the easiest military occupations in history and they succeeded in turning it into one of the worst military disasters in history. They can’t control it and it’s almost impossible for them to get out for reasons you can’t discuss in the United States because to discuss the reasons why they can’t get out would be to concede the reasons why they invaded.
If you listen to Seymour's interview it seems that the strategy is going from "unimaginable catastrophe in Iraq" to "unimaginable catastrophe in the whole Middle East".
In the meantime, it turns out that the weapons that were presented a couple of weeks ago as being "manufactured in Iran" turned out to be manufactured in Iraq.
If that was not enough, the UN Calls US Data on Iran's Nuclear Aims Unreliable:
The officials said the CIA and other Western spy services had provided sensitive information to the Vienna-based International Atomic Energy Agency at least since 2002, when Iran's long-secret nuclear program was exposed. But none of the tips about supposed secret weapons sites provided clear evidence that the Islamic Republic was developing illicit weapons.
"Since 2002, pretty much all the intelligence that's come to us has proved to be wrong," a senior diplomat at the IAEA said. Another official here described the agency's intelligence stream as "very cold now" because "so little panned out."
For those of us too cynical to believe anything the Bush administration has to say on any matter, this is hardly news. But it is always nice to see them debunked in public.
Posted on 27 Feb 2007
Last week I described how international travel for an Internet addict can be very costly, 10 euros here, 12 dollars there 20 euros for 24 hours there and very soon you have paid a hundred dollars in a two-day trip on Internet access fees.
Now, instead of paying half a dozen people to get WiFi access, I have consolidated my "getting ripped off" with AT&T. Now they bill me an insane amount for getting "3G worldwide access".
Two problems though: "3G worldwide" actually means "3G in Boston, and 9600 baud modem speed in Europe". During the entire trip I could not get the "3G" light on the card to turn on, it consistently stayed in "2G" mode.
But just before you think "Well, GPRS/2G is not so bad", I want to point out that it took 25 seconds to load www.google.com.
Posted on 27 Feb 2007
Ted Haeger interviews Anna Dirks.
Anna leads the team at Novell that was in charge of improving the usability of the desktop. Her team launched Better Desktop, the Tango Icons and prototyped, tuned, improved, mocked-up and tested various elements of the usability elements in the SLED desktop.
Her interview is here in Novell's Open Audio.
She talks about the process used to design the new features of SLED and it also got a few hints on what is coming up for SLED 10 SP1 (pay attention to the physical weight of the Usability Labs).
Posted on 26 Feb 2007
After being ripped off by hotels and airports for my Internet connections (with some places charging as much as 20 euros per hour, and your standard 10 euro per hour at CDG), I have purchased a 3G card from Cingular and a plan to use it internationally.
I followed the instructions on this blog post that covers another card on OpenSUSE 10.2.
With SLED 10 and the Aircard 875, the only difference is that I had to use a different vendor and product ID:
# /sbin/modprobe usbserial vendor=0x1199 product=0x6820
Or alternatively dump that in the /etc/udev/rules.d/51-3g-datacards.rules file.
Once you configure the modem in yast as described in the above blog post, go to NetworkManager, and select the "modem0" serial connection. There seems to be a bug in NetworkManager, when the modem0 connection is active, the icon seems to go away (am guessing its not finding the icon for it, and hence the applet vanishes).
A few minutes later, you will be online.
Am still missing some tools to determine the upload/download speed, anyone have some good pointers?
Posted on 21 Feb 2007
Lutz Roeder has just released a new version of Reflector:
This time Reflector will work out of the box with Mono on Unix (no special handling or special flags) and will even detect the presence of your Mono libraries:
Recently, while implementing the second chunk of lambda expressions, I had to go through all the classes that derived from the Mono.CSharp.Expression and the Mono.CSharp.Statement classes to implement the "CloneTo" methods, and Reflector came in as a great tool to find those classes with minimal fuzz and without wasting any grep batteries in the process:
Reminder: the use of Reflector to look at third-party code is prohibited if you are planning on ever contributing to Mono. You are free to use Reflector to decompile Mono code though.
I would have blogged early this morning when the news came out, but I struggled betwen Consolas, Monaco and Tahoma for the screenshot. Solving the dilemma was only possible thanks to the extensive dinner discussion with Garrett and Jeff at Kashmir.
One great feature is the Reflector Analyzer to explore who exposes, uses, depends or inherits a given type:
There are no borders on the screenshots due to the way Compiz paints windows on the screen (a separate process paints the decorations).
Jackson has been fixing various issues in Winforms to improve the rendering. These screenshots are the result of his endless hours of work. Thanks Jackson!
Posted on 21 Feb 2007
This weekend am heading to Brussels to the FOSDEM conference. The last time I spoke there, Mono was in its infancy (2002). At the time the C# was about to be self-hosting on Linux, but was not quite there (0.8 and 0.9 releases).
Now, five years later Mono has gone through 60 releases and is made up of a few million lines of code and has spawned plenty of software projects.
The question is: What kind of things should I talk about in my two sessions at FOSDEM?
I have one general Mono session, and one session in the OpenSUSE track. Since Mono is so large, and there are so many interesting things being done with it, I find it challenging to find a good topic to discuss with a technical audience like FOSDEM.
I could do a general overview and give some high-level overview of what we are doing, but I feel that the FOSDEM audience is probably more interested in something more technical. So I should probably limit my presentations to two or three key topics.
Should I talk about what is available today, and what are people doing today, or should I talk about the future, and what we want to do?
Maybe the general session could be about the state of the project and where we want to go in the future; And the OpenSUSE session about the practical things that can be done today. Not sure.
What would you like to hear about?
Posted on 21 Feb 2007
Today we announced the support of Visual Basic.NET in Mono.
Visual Basic is still one of the most used languages in the Windows world, and with this release we hope to assist a large segment of the population to bring those applications to Unix with Mono.
Rolf developed the new compiler. The new compiler is fascinating because it is a VB 8 compiler (this means that it supports generics), but also because it is written in VB itself. This compiler was sponsored last summer by the Google Summer of Code (2006 edition).
The compiler on my laptop takes 12 seconds to compile itself (78,000 lines of code).
The new runtime (for 1.0 and 2.0) was developed by Boris Kirzner, Guy Cohen and Rafael Mizrahi at Mainsoft and just like the VB compiler it was written in VB, and it is made up of roughly 17,000 lines of code.
As some people have pointed out, the compiler and the runtime are not enough to run applications. A lot of the portability will be mandated by your API consumption. If your application is based on .NET 1.0, you should be ready to use Mono now. If you are using .NET 2.0, you will need to check whether everything your application needs is supported by using Moma the Mono Migration Analyzer.
Our 2.0 support these days for ASP.NET and the core is passable. It is not 100% there, but the ASP.NET Starter Kits should run, and so should applications like mojoPortal.
Posted on 20 Feb 2007
I was reading Rolf's post where he complains that people never get his name right. I am used to it.
English speakers typically pronounce my name as "Mi-koo-elle", or "Mi-goo-elle", while the actual pronunciation is more like "Mig-elle".
Bonus Update: Also most English speaking people miss-pronounce "de Icaza" as "di-eye-kaza", but it is pronounced "the-eekaza".
Posted on 20 Feb 2007
Ted Haeger discusses the changes done recently to the Main Menu. Link to the video:
I love the main menu that Anna's team designed and Jimmy implemented. The KDE folks then improved upon the design and on this new iteration some of the ideas from the KDE design are incorporated.
The code is available from Gnome's SVN
Update: Someone at the office told me story of how the "Search" input box at the top of the main menu ended up at there.
In the early days of the main menu the search bar was at the bottom of the main menu, similar in spirit to Vista, but this prototype existed before Vista integrated it.
During the routine usability tests that are conducted in the Cambridge office, one of the tasks that was part of the test was something along the lines of "Find the document that contains foo". But people would not use the built-in search, they did not notice it on the main menu. Instead people went to the file manager and started opening file by file to find the document.
When the search was moved to the top, the subjects in the usability tests immediately started using it.
Then Vista released their first menu with search integrated at the bottom of the menu. I have been wondering if the guys in the Vista team had conducted any similar tests with their start menu.
Then someone from the MacOS team weighted in and described the process used at Apple.
From Ted Haeger's video you can see that logout/shutdown also got redesigned with the new menu. I do not know the story behind how it got implemented (there are now two options: shutdown/logoff in the menu) and both bring up two dialogs with further options.
It would be interesting from someone on the Novell Desktop team to blog about that process to complement the Vista and OSX postings.
Posted on 18 Feb 2007
Alan McGovern has created a fantastic BitTorrent client library in C#. The effort was part of last year's Google Summer of Code. Alan continued to tune it and implement many of the protocol extensions after the SoC was over so it is now a very complete.
We have a very early Gnome UI that was created last year, but it has not been updated very much, and it could really use some work to get it updated.
The library these days is quite mature and the command line client works well, but we really ought to have a Gnome UI.
There is also a case to be made for a simple and clean UI for Bittorrents.
A Winforms UI should also be possible, and am sure our friends in the Windows world would appreciate it.
Alan posted some updates on the current state of the library:
uPnP support has been enabled in MonoTorrent using Mono.Nat. So all you people with uPnP routers no longer have to worry about manually creating the port mapping in your router. It'll all be done automagically (all going well ;) ).
Disk writes are now fully asynchronous, but now will automatically throttle download speed if you are downloading faster than your harddisk can write. So you won't ever get 100s of megs of ram being used and 100% cpu usage when exceeding your write speed.
Upload and download speed calculations have been improved drastically (ish) for torrents. What i did before was calculate each individual peers upload and download speed, then sum up that for all connected peers to see a torrents overall download rate.
Posted on 17 Feb 2007