Re: Who should security issues be reported to?
From: Duncan Booth (duncan.booth_at_invalid.invalid)
Date: 01/28/05
- Next message: Paul Rubin: "Re: Who should security issues be reported to?"
- Previous message: Diez B. Roggisch: "Re: Why do look-ahead and look-behind have to be fixed-width patterns?"
- In reply to: Paul Rubin: "Re: Who should security issues be reported to?"
- Next in thread: Paul Rubin: "Re: Who should security issues be reported to?"
- Reply: Paul Rubin: "Re: Who should security issues be reported to?"
- Reply: Fredrik Lundh: "Re: Who should security issues be reported to?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: 28 Jan 2005 12:50:10 GMT
Paul Rubin wrote:
> Duncan Booth <duncan.booth@invalid.invalid> writes:
>> In other words, I'm intrigued how you managed to come up with
>> something you consider to be a security issue with Python since
>> Python offers no security. Perhaps, without revealing the actual
>> issue in question, you could give an example of some other situation
>> which, if it came up in Python you would consider to be a security
>> issue?
>
> Until fairly recently, the pickle module was insufficiently documented
> as being unsafe to use with hostile data, so people used it that way.
> As a result, the Cookie module's default settings allowed remote
> attackers to take over Python web apps. See SF bug 467384.
SF doesn't seem to know about any such bug any more.
Google finds me
http://mail.python.org/pipermail/python-bugs-list/2001-October/007669.html
which appears to be SF bug 467384, but it says nothing about security or
the Cookie module, just that you wanted better documentation.
I think its a bit borderline whether this really was a security bug in
Python rather than just a problem with the way some people used Python. It
was a standard library which if used in the wrong way opens a security hole
on your machine, but there are plenty of ways to open security holes.
The response seems to have been to document that there is a security
concern here, but it is still just as possible to use python to expose your
machine to attack as it was before.
But thanks anyway, it does give me the sort of example I was asking for.
- Next message: Paul Rubin: "Re: Who should security issues be reported to?"
- Previous message: Diez B. Roggisch: "Re: Why do look-ahead and look-behind have to be fixed-width patterns?"
- In reply to: Paul Rubin: "Re: Who should security issues be reported to?"
- Next in thread: Paul Rubin: "Re: Who should security issues be reported to?"
- Reply: Paul Rubin: "Re: Who should security issues be reported to?"
- Reply: Fredrik Lundh: "Re: Who should security issues be reported to?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
- Re: Whats so funny? WAS Re: rotor replacement
... > party library in addition to Python, ... that's as bad as a crypto bug.
... internet is security related. ... If they're able to install external libraries,
... (comp.lang.python) - Re: OT: Whats up with the starship?
... the security advisory explains that the cause of the problem is a bug ...
in the source code used to implement reprfor 32-bit Unicode strings, ... Python
2.2 was released in 2001. ... (comp.lang.python) - Re: New to Python: Features
... Is that a limitation of the language (the security holes I mean)? ...
>done with Python. ... >You can, of course, design by contract. ...
(comp.lang.python) - Re: New to Python: Features
... > Now your encouraging me not to use Python? ... group development (security,
consistent naming scheme, ... give about a 5% speed boost, so the answer to your original
... object which interacts with string interpolation. ... (comp.lang.python) - Re: Need a compelling argument to use Django instead of Rails
... Ruby can also lock access to attributes. ... What about learning Python
instead of repeating arguments from clueless ... This is not "security", this is data-hiding.
... this is language-inforced access restriction. ... (comp.lang.python)