[IPython-user] Any future integration with pydbgr planned?

Fernando Perez fperez.net@gmail....
Tue Jan 26 01:31:25 CST 2010

Hi Flavio,

On Mon, Jan 25, 2010 at 10:44 PM, Flavio Coelho <fccoelho@gmail.com> wrote:
> quote from http://www.oss-watch.ac.uk/resources/gpl.xml#body.1_div.4
> It is also commonly believed that GPL v2-licensed software cannot be bundled
> with non-GPL v2 software in a collection of software, or that if it is, then
> all the software must be relicensed under the GPL v2. Again, this is false.
> Section 2 of the GPL v2 states that such mere aggregations are not governed
> as a whole by the conditions of the licence.

We're not talking about aggregation here, we're talking about actual
use via "import".  As best as I understand the GPL, that very much
triggers the copyleft requirements of the GPL (it's 'viral' effect).
The GPL is an 'infinite leverage' license, with no proportionality
whatsoever: you use one line of GPL code in a 'derivative work'
manner, you MUST GPL your entire codebase, all 50 million lines of it.
 In that regard, the LGPL is a much more balanced license, as it
demands only that changes to the original code are contributed back,
*not* third-party code.

There are many great things about the GPL: amongst other things, for
whole applications (e.g. the linux kernel) it makes a great license.
But I do dislike this 'infinite leverage' property when used for

> I haven't checed but I am pretty sure that python must link to some GPL C
> library...

Oh yes, readline... See more below if you think this makes anything
clear or simple :)

> more from the GPL FAQ:
> If a program combines public-domain code with GPL-covered code, can I take
> the public-domain part and use it as public domain code?
> You can do that, if you can figure out which part is the public domain part
> and separate it from the rest. If code was put in the public domain by its
> developer, it is in the public domain no matter where it has been.
> BTW gnu readline, which I think ipython uses, is GPL software.another link:

Well, we use readline indirectly, via python itself.  We import
python's readline bindings:

>>> import readline
>>> readline.__file__

which are themselves python-licensed.  So we appear to be OK. But
whether *those* bindings are actually legally OK for the PSF to
distribute is an entirely unsettled question, and a highly loaded one
at that.  I'm sure you'll find this thread highly informative:


As you can see, there has been in the past at least a very real
discussion of Python removing readline altogether for this very
reason.  If they ever do that, ipython will lose readine support,
period.  If the entire Python core team isn't going to fight the FSF
and RMS on this one, I'm not about to, believe me.



More information about the IPython-user mailing list