well don't know why I couldn't reply to the initial thread..

yep Tyler is right I have columns with trailing null chars 
mycolumn, mycolumn\u0000\u0000 prints the same in cassandra-cli, but not if you write the output in a file
I haven't find where they come from, but no matters the architecture we are using is hopeless, will rather try to make sure to trim column names before insertion

2012/9/21 Tyler Hobbs <tyler@datastax.com>
If you're seeing that in cassandra-cli, it's possible that there are some non-printable characters in the name that the cli doesn't display, like the NUL char (ascii 0).  I opened a ticket for that somewhere, but in the meantime, you may want to verify that they are identical with a real client.

On Tue, Sep 18, 2012 at 4:03 AM, aaron morton <aaron@thelastpickle.com> wrote:
They are. Can you provide some more information ? 

What happens when you read the super column ? 

Aaron Morton
Freelance Developer

On 18/09/2012, at 5:33 AM, Cyril Auburtin <cyril.auburtin@gmail.com> wrote:

First sorry but I'm using an old version 0.7.10

and recently I've come up seeing this

=> (super_column=MYMED_embrun.maire@gmail.com,
     (column=permission, value=1, timestamp=1347895421475))
     (column=email, value=embrun.maire@gmail.com, timestamp=1347894698217)
     (column=id, value=MYMED_embrun.maire@gmail.com, timestamp=1347894698217)
     (column=permission, value=0, timestamp=1347894698217)
     (column=profile, value=e24af776b4a025456bd50f55633b2419, timestamp=1347894698217))

as a part of of a supercolumnFamily

I thought supercolumn was meant to be unique?

Tyler Hobbs