Posted by Jonas Elfström
Mon, 05 May 2008 19:17:00 GMT
I've felt it and I've heard it from colleagues several times. Writing unit tests can be hard work. Especially adding unit test to an existing code base is, at best, cumbersome. Also it's one of those things with delayed gratification. Sometimes it's not even you that will benefit from them being there because the biggest win can be long down the road, when changes to the system has to be made.
Tests may seem to be isolated and it's even considered a good thing to keep them that way. Even so the tests of your application has a correlation to what the system aims to do on a bigger scale. This one of the things BDD focuses on. I think that one of the biggest advantages is that you in one process writes a specification and tests that ensures that the spec. is met. Testing becomes a natural part of the development process. This way it clearly shows that BDD and TDD are design processes and that it's certainly not all about adding unit tests.
Find out more about BDD on: http://behaviour-driven.org
It must be stressed that BDD is a rephrasing of existing good practice, it is not a radically new departure. Its aim is to bring together existing, well-established techniques under a common banner and with a consistent and unambiguous terminology.
For Ruby RSpec has almost become the de facto standard for BDD. The concepts Story, Scenario, and Test feels natural and the syntax is short and easy to read.
In languages like Java or C# the tests often becomes much more cluttered and some of that clutter is the extra code that comes with static typing. I believe that dynamically typed and overall dynamic languages like Ruby or Python could find a nice little niche here. They could become DSL's for testing.
RSpec is on it's way for .NET/C# via IronRuby and for Java via JRuby but don't hold your breath because they are still in alpha and beta.
.NET / C#
Testing .NET with IronRuby...
NSpecify => RSpec… well closer anyway
Java
Java Functional Testing with JRuby and RSpec
JtestR
Ruby
Slapp - A simple chat wall Merb tutorial. With nice exampes of using RSpec.
Behavior-driven testing with RSpec
ASP.NET MVC
ASP.NET MVC Test Framework Integration Walkthrough
MVC Preview - Testing
ASP.NET MVC Session at Mix08, TDD and MvcMockHelpers
Posted in Ruby, C#, Java | 3 comments
Posted by Jonas Elfström
Mon, 11 Feb 2008 23:26:00 GMT
In a recent release TrueCrypt now supports drive/partition encryption.
One reason to encrypt on disk instead of file level is that operating systems and applications sometimes accidently stores passwords on your hard drive. This can happen in a number of ways and one common mistake applications make is to not prevent to be put on disk by the OS. Modern systems have a page/swap file. If a program gets paged out while holding your clear text password in pageable memory your password will be written to disk. The problem is that there are password recovery tools that can scan your page file for passwords.
You can configure Windows (and surely most other operating systems) to clear the page file on shutdown which will give you better protection (and slower shutdowns). Be aware that if you simply turn off the power the page file will be intact.
Posted in Security | 2 comments
Posted by Jonas Elfström
Wed, 12 Dec 2007 16:15:00 GMT
I recently attended a session held by Marcus Murray. It seems it was kind of a compressed version of the session he held at TechEd earlier this year. Murray is witty, charismatic and has a broad and deep understanding of IT-security issues. He cracks jokes and practices a little social engineering to keep the audience attentive.
If you and your IT-staff wants to be briefed (and scared) with the latest in IT-security I could easily recommend Murray.
He demonstrated a couple of tools that both impressed and scared me. First he demonstrated how to set up a mail based attack using the commercial Core IMPACT. It's a very impressive tool and mail based attacks are only one out of many attacks this software has the ability to execute. Before seeing this I could never have guessed there are tools this advanced and this easy to use. The lists of exploits it can test, in an all automated fashion, were long and seemed to be up to date.
Murray also demonstrated ARP poisoning and hijacking of a RDP session by using the free Cain & Abel tool. You could feel the discomfort in the air as it dawned on the audience how easy this is to set up.
Posted in Security | no comments
Posted by Jonas Elfström
Mon, 26 Nov 2007 20:40:00 GMT
There's some buzz about a new trojan technique called "Man in the browser". The trojan plugs itself into the users browser and then it intercepts the HTML. This have all sorts of implications, for instance the SSL certificate will seem to be valid.
Even if your virus protection does not detect the "Man in the browser" there are still ways to be quite safe. If your bank uses a security token which you not only logs in with but also use to sign your transactions, the attack will most likely fail. One problem is that it is up to you, the user, to actually verify that you are signing the expected amount to the expected accounts. To get protection against someone who has complete control over your computer the security protocol must "communicate" with the security token both ways.
It is not enough to only let the user enter the total value of the transactions because the tokens answer to such a simple value could be reused (during a short period of time) also the attack could be crafted to create the correct amount but to other accounts. By asking the user to also sign every new account the attacker will not be able to hide a transaction to his account.
The challenge is to make the security protocol clear and simple enough so that the user can understand what he would expect the bank to respond and expect from him.
It's the usual three: have a firewall, an updated virus protection and a secure bank with a digital security token were you sign your transactions and maybe you can sleep a little better.
Posted in Security | 1 comment
Posted by Jonas Elfström
Thu, 15 Nov 2007 21:38:00 GMT
Sometimes you do not want to show the database id for a row in the URL. The reason could be that you do not want someone to be able to scan through all the data.
One solution is to use GUID's but they have drawbacks and one of them is that they add a considerable length to the URL. The shortest URL-safe representation of a GUID I've seen is 22 characters but usually they are 36 characters.
Depending on how your id's are implemented a much shorter way could be to simply to encrypt them.
Here's a Ruby-example that Blowfish encrypts, Base64 encodes and URL-encodes an integer value. You can get crypt as a gem:
gem install crypt
require 'rubygems'
require 'crypt/blowfish'
require 'Base64'
blowfish = Crypt::Blowfish.new("A key up to 56 bytes long")
plainId=123456
encryptedBlock = blowfish.encrypt_block(plainId.to_s.ljust(8))
idForURL = URI.escape((Base64.encode64(encryptedBlock).strip))
decryptedId = blowfish.decrypt_block(Base64.decode64(URI.unescape(idForURL))).strip.to_i
The .ljust(8) is because Blowfish is a 64-bit block cipher and the Ruby-implementation does not pad the data itself.
The id in the URL in this case would be c2PSXWgky40=. Its 12 characters long (11 if you skip the equal sign) and that's 10 or 24 characters shorter than a GUID. Also there is zero percent chance of a collusion and if you want to you can even decrypt it.
This is not a super safe implementation but if you start your id's at a random and not too low number you are making it a bit harder for someone to crack the 56-bit key. Actually a truly random and at least 64-bit big number would be a better choice as it would have no connection to the true id at all. You would have to check for uniqueness before storing those in the database though.
Posted in Security, Ruby, Cryptography | no comments
