WebApp Sec mailing list archives
Re: Canonicalization
From: exon <exon () home se>
Date: Fri, 21 Apr 2006 16:43:51 +0200
Peter Conrad wrote:
Hi, On Thu, Apr 20, 2006 at 10:22:18PM -0400, Rossen Raykov wrote:Is that ?simplest form? achievable? One can perform many and different encodings making the task of decoding them very difficult and resource consuming. Usually it is cheaper and safeties to do semantic checkup and treat the input as erroneous if it does not confirm to the expected input format.you're comparing apples with oranges here. You must perform canonicalization *before* you can match the input against the expected format.For example if you are expecting number anything different than a number is error.Here are some different representations of the same number: 11 +11 11.0 11.00 011
This is just stupid. If the user is supposed to input an unsigned numeric integer it's just dumb to accept dots, plusses, commas and any other char than what fits in 0-9. Leading zeros can be stripped, although if the application itself never adds them you shouldn't allow that either. Even if you accept negative numbers or floating point numbers there's still a very limited range of characters to accept.
All user data that's truly free-form should never have any impact on the program's behaviour and should thus pass through unmolested and unvalidated. This is OK so long as the author knows about it (free-form data designated for use in database queries are *not* free-form).
Most web-application security holes come from programmers attempting too much dwimmery, making not only the application slower and more cumbersome to maintain, but also significantly less secure. Userfriendliness should be mainly in error-messages, not in logic code.
/exon ------------------------------------------------------------------------- This List Sponsored by: SPI DynamicsALERT: "How A Hacker Launches A Web Application Attack!" Step-by-Step - SPI Dynamics White Paper Learn how to defend against Web Application Attacks with real-world examples of recent hacking methods such as: SQL Injection, Cross Site Scripting and Parameter Manipulation
https://download.spidynamics.com/1/ad/web.asp?Campaign_ID=701300000003gRl --------------------------------------------------------------------------
Current thread:
- Canonicalization susam_pal (Apr 12)
- Re: Canonicalization Yann (Apr 12)
- Re: Canonicalization Rogan Dawes (Apr 12)
- <Possible follow-ups>
- RE: Canonicalization PPowenski (Apr 12)
- Re: Canonicalization Andrew van der Stock (Apr 12)
- Re: Canonicalization Rossen Raykov (Apr 20)
- Re: Canonicalization Peter Conrad (Apr 21)
- Re: Canonicalization exon (Apr 21)
- Re: Canonicalization Jason Murray (Apr 23)
- Re: Canonicalization exon (Apr 24)
- Re: Canonicalization Rossen Raykov (Apr 20)
- Re: Canonicalization Yann (Apr 12)
- Re: Canonicalization Eoin (Apr 21)
- Re: Canonicalization Andrew van der Stock (Apr 22)
- Re: Canonicalization Rogan Dawes (Apr 14)
- Re: Canonicalization Jason (Apr 14)
- Re: Re: Canonicalization Mariusz Pękala (Apr 14)
- Re: Re: Canonicalization Peter Conrad (Apr 18)