Google Chrome OS

Google Chrome OSSo… what I told yesterday evening to a good friend of mine has just been announced: Google is to release their “own” operating system called Chrome OS. In short: Linux kernel, simple interface, all apps on the web, better security, very fast boot. This is likely to be much like gOS actually. (I’m sad for them, it seems they have missed a Google buyout they were obviously expecting). So in the future it’s very likely that Microsoft Windows and Mac OS are going to be challenged the hard way on the mainstream market. But I was wrong yesterday because I predicted to my friend that the Google OS would be nothing else than Google Android. No: instead they have choosen a luxurious way to get the best from Google’s own development team by entering into a kind of self-competition, or maybe better to call that internal competition. Now, here are a few things that come to mind:

  • Linux kernel, fast boot: OK, there are a few people who already know how to do that, and actually do that for years
  • (Web) Applications on the web (say “in the clouds”, it’s more fashionable): OK… what else? Windows applications ? no. Linux applications? maybe since that’s Linux kernel. Hmmm.
  • Minimal user interface: ouh ouh! IceWM?
  • “redesigning the underlying security architecture of the OS so that users don’t have to deal with viruses, malware and security updates”: I think there are a couple of guys around here on the planet who have been using such operating systems daily, for at least… 15 years (ouch… white hair is not far).
  • What else…? … hardware support! plug and play your XYZ too cool €15 device, and see what happens. I’m curious to see what they do about this part, really curious.

So, OK, I’m a bit harsch, rather amused actually, with this announce, because you know, to me it’s just a minimal Linux system with a browser on the top of it, and I understand that this announce is important and is catching a lot of attention because there is a company with 6 letters starting by a G behind it. And it’s really possible that they catch a significant part of the OS market, for sure. Anyway, I’m not too convinced about the lack of support for Windows apps, I think they may be a little too confident about the capabilities of web-based applications, but this may change in the next three to five years. And of course I can’t wait to see what they do about the support of zillion hardware devices that come with a Windows driver on a CD to make them work. Stay tuned!

IP Traffic Shaping on Linux

Recently I got interested in traffic shaping to simulate various bandwidth capacities. It was a headache to find a working software in that field until I realized that 1) it was easy 2) it was straighforward on Linux kernel since version (2.2.x?).

First of all, you need to modrobe a few kew kernel modules:  cls_u32, sch_cbq, ip_tables

Then all you have to do is to use the “tc” utility which is part of the iproute package.

For instance, let’s assume that you want  to limit incoming and outgoing traffic to 256kbits/s on your local host, and assuming that you have a 100Mbps capable network interface on eth0, what you have to do is:

#  tc qdisc add dev eth0 root handle 1: cbq avpkt 1000 bandwidth 100mbit
# tc class add dev eth0 parent 1: classid 1:1 cbq rate 256kbit allot 1500 prio 5 bounded isolated
# tc filter add dev eth0 parent 1: protocol ip prio 16 u32 match ip dst 0/0 flowid 1:1

Then if you want to change the limit, use “replace” instead of “add” in the second command. For instance:

# tc class replace dev eth0 parent 1: classid 1:1 cbq rate 64kbit allot 1500 prio 5 bounded isolated

You will notice easily that it’s doing the job very well.

Anyway,  I went into some troubles when I started to monitor the traffic: the bandwidth that I set with tc doesn’t fit at all with the actual limitation. For instance, when setting 50kbit in tc, I get a real limitation of around 24 *kBytes* per second, which is about 200kbps. At first, I thought it was a problem with “knetdockapp” that I’m using to monitor the traffic. So I used Bandwidthd which shew similar results, and finally, I transferred a big file during 60 seconds and calculated the real rate from the number of bytes that were received. The results were still the same.

So I’m still wondering why there is such a difference between the figure provided to tc and the real shaped bandwidth.

Ulteo unveils the first Open Source virtual desktop, providing businesses with quicker, cheaper deployment and easier applications management

Following its commitment to desktop virtualization solutions, Ulteo, an Open Virtual Desktop Infrastructure company, announced today that they were releasing the first version of their Open Virtual Desktop solution for enterprises. Delivering faster deployment times and ease of management for the IT department, this first release can be integrated easily into an existing professional Linux or Windows IT environment. The solution can be up and running in a few minutes, delivering rich desktop applications to corporate users.

Read more >>>

Google Chrome is out

Google Chrome, a new web browser – based on Apple WebKit – aimed at surfing the world wide web, has been released by Google today. Only available for Windows for now, it should be available for other plateforms such as Linux later. What is funny, in my opinion, is that normally, releasing a new web browser should have made not even a news wavelet in the IT world. But Google is releasing it, so it makes a lot of noise. It’s clear that even if they were to release a toilet bowl, that would generate a lot of press. The good news is that they are releasing this product as Open Source, because (they)”owe a great debt to many open source projects, (…) We’ve used components from Apple’s WebKit and Mozilla’s Firefox” (which are open source projects). The question to Google now is: why don’t you improve Webkit and Firefox features and performance instead of releasing your own web browser? Other question: licensing. According to Wikipedia, Google Chrome is covered by BSD licensing but as far as I know, WebKit is covered by the LGPL. So, find the bug (if any). Regarding Mozilla, I don’t know enough about its licensing, so maybe it can be converted to BSD. I’ve tried to find information about source code licensing on the Chrome web site, but I couldn’t find any in 5 minutes. And by the way, why no antialiasing for text rendering? What else? Hmm… OK, we have a new web browser around there. Let’s dance.

About personal data security

[this is a repost of my answer to someone who wondered about security of personal data on Ulteo forums]

Let me explain my way of thinking about this issue.

At first, let’s assume that data integrity and confidentiality are the two needed requirements:

  • we want to be able to retrieve data, as long as we need it, whatever happens on the earth (bombs, earthquakes…)
  • we want to ensure that no one but authorized people can read the data, use it or modify it


Now there are different cases. Let’s take these two cases to simplify:

  1. your personal data : most of time they are stored on your computer. As a result, they are totally unsafe for several reasons: someone can break into your computer and steal your data, or you can be stolen your harddrive (laptop), or your house can go into fire etc. A slightly different case is your online data. For instance, Gmail, Yahoo! mail etc. They won’t guarantee anything but “doing the best” to secure the data. It means that it’s likely that they have advanced security systems (but who knows), that they have redundant servers around the planet etc. So, it’s more likely that your data get more secured if they are stored online in this case. Anyway, they are not really confidential: Gmail reads your emails to generate ads for instance. Additionally, it happens that they close accounts, for any reason. I know people who got their Yahoo! mail account closed because the Terms of Services weren’t respected (without any further detail). Later, they were unable to get in touch with someone at Yahoo! to get it back and lost all emails. Maybe in some cases that’s a bug. Worse: laws permit that your data can get accessed by government agencies anytime for any “good reason” (as far as I know that’s the case for Google in the USA and Blackberry in the UK). So there is still a risk to have your data vanish in the air, even if they are stored online on a big service.
  2. data within a corporation (ie “sensitive data”). Here, everything depends on the corporate’s policy about data security. Most of the time, I think there is a good level of integrity for the data, assumed that there are mechanisms to get the data replicated to other geographical places for instance. Confidentiality is certainly worse because security cannot be perfect, and also because many people within corporates use Gmail, Blackberry and other services intensively, apparently even for sensitive transactions/discussions. This is a real (known) issue for strategic corporations that need a high level of confidentiality.

Now, what I think, is that the key answer to data integrity and confidentiality is:

  • redundancy to address the integrity problem
  • heavy encryption to address confidentiality


For instance, with tools such as GPG and Thunderbird Enigmail (which are provided and installed by default on Ulteo), you can encrypt your sensitive emails very easily. The only constraint is that you first need to import your receiver public key first, but that needs to be done only once. Then, all you need to do is to select “encrypt message” when writing your email. With a 2048 or 4096 bits encryption key, this even removes the need to have any security or encryption “on the line” (TLS, SSL…).

In this case you can even add a personal gmail account in CC: as a safe backup! You won’t be able to read the email content within Gmail, but if you happen to need it, you can retrieve the email and decrypt it locally. And Gmail won’t be able to read the content of these archives in any way.

In the same spirit, Ulteo also integrates the Kopete “Silc” plugin that provides a totally secured IRC chat.

Now, there is the question of data that are stored at Ulteo. Right now, I can’t tell you more that “we’re doing our best to secure your data”. This means security measures on servers, and replication. But I agree that it’s not an ultimate solution.

We plan to provide an encryption feature that would permit us (and you) to store *only* encrypted data, that could be used/decrypted only by the owner of the data, using his credentials.

In this case, you would have a local secured directory where you could put all your sensitive data, and this would be the same on Ulteo online services. So in the bad case where you would be stolen your harddrive, or in the case Ulteo servers would be cracked, nobody but you couldn’t read your secured data.

Gaël.