Perhaps you've read about Google's Project Glass - a set of augmented reality glasses that will provide users with real-time information right before their eyes. Literally.
After waiting awhile for
Join Scoop.it for Free
Share ideas that matter on the social web and experience
the benefits of curating the world's best content.
I don't have a Facebook, a Twitter or a LinkedIn account
Already have an account: Login
About everything making software flawless. Verification, analysis, methods and tools.
Curated by
Petr Muller
 Your new post is loading...
Scoop.it!
No comment yet.
Sign up to comment
Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Companies love metrics. Numbers, whee! Easy to grasp, and 100% coverage feels so good, especially when written to the executive summary. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
There's never enough of interesting tools to try out. This time it is an app helping with creating a realistical load on a system. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Interesing shortie about a company utilizing pay-per-useful-bug contract testers all over the world. The concept seems to be quite viable: it's a fine compromise between the in-house testing and public beta testing. In-house testers inevitably get some knowledge and assumptions about the software, and miss some issues by simply failing to realize another wicked use-case exists. And users are good at finding wicked use-cases. So if you engage testers by trade, not familiar with your SW, and unleash them on your SW, that can nicely complement your testing. Engaging your use base to actually file proper reports can be hard, and this is solved by the greatest motivation of them all: voluntary work and money. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Being Red Hat myself, I of course know about this for some time. But this is a nice summary of how SA tools can be useful for the immense project like an enterprise Linux distro. Plus, it has a totally weird photo of my colleague :) Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
I do not do much web application development or testing, but I heard a colleague of mine mention this web load testing tool, and he spoke well about it. The tool is Python-oriented (tests are Python scripts), and the fact that it is free (beerwise and speechwise) makes it an interesting addition to web apps testing toolbox. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Robert Dewar (the CEO of AdaCore) blogging about the possibility of truly reliable software in the light of the recent Knight Capital Group SW frak-up. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Okay, seems like the light-weight static analysis has found it's way into all major IDEs out there: VS has it, Eclipse now has it for both Java and C/C++ (although they chose to implement the checks from the scratch instead of integrating with some established tool), XCode has it for some time, and now NetBeans 7.2 brings FindBugs integration. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
I've been neglecting this aggregator for a while: I'm, sorry. First I was really busy, then I really tried to not be busy (=vacation), and then I travelled like a policitian before election. I hope things get calmer now.
I really liked this small collection of software testing related jokes. This one is really hilarious:
Sign that you are dating a tester: When you ask him how you look in a dress, he'll actually tell you. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Brilliant rant about ignorant programmers, statistics and sound (formal) methods of performance measurement. It's actually quite old, but this week I need to review yet another performance testing framework technical report which is, of course, ignoring statistics in its measurement design. Shame. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
LLVM is the place to go when one wants to implement static and dynamic checkers and bug hunting tools (sadly, I must add, being a GCC guy myself). This small tool is one of the nice examples: it instruments programs to detect dangling pointer usage and buffer overflows. The nice thing is that there are lot of resources on the page descibing how the tool actually works, including a paper, presentation and an algorithm description. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Yeah, even bash scripts contain errors. And there is probably more code that matters written in bash then we would both expect and like. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
I really start to wonder. Reek is the third Ruby-specific lightweight static analysis tool I mention on this topic. And I have more in the buffer, and I find more and more (in addition to the ones I already mentioned, I know at least about three more!). I really wonder why Ruby has so many tools, compared to, say, Python? I'm not a Ruby guy myself, so if anyone knows what makes Ruby programmers to write all those static analysis tools, let me know :) It seems like Ruby tools focus on validating good design techniques in OOP instead of just finding bugs. I need to look if such tools exist for C++ and Python. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
|
Scoop.it!
Petr Muller's insight:
I generally like John Regegr's blog. For an academic, he keeps focus on the "real world" specifics of the computer industry. It's a shame that so many of the computer science people in the academia are so much removed from the industry itself... Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Yeah, I am guilty too: as a kid, I spent a lot of time gaming. As pretty much everyone, I dreamed about having a job as a game tester: imagine that, getting paid for playing games all day! I ended up being a tester, but not in gaming industry. I have this article in the buffer for some time now, and now it's time to put it here. Being a games tester is not what I dreamed, that's for sure, but can it really be *that* bad? Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
...because they are likely to freak out. Really. Spending time fiddling with that crap you call Release Candidate does not make testers most stable people in the world. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Sometimes testers are swamped in the land of holy test case. If one does not exist, that's bad. When it is created, things magically improve. You get the test coverage percentage, status report lines, new PASS results...
Have you thought about not running a test case and just report, for you were 100% sure it will just PASS? And if you did, what does it say about such test? Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
In his blog post, Adam Knight discusses the urge of the automation testers to "make tests work" instead of correctly thinking about why the test failed. I kind of recognized some of my (wrong) habits in the article, and therefore I find it worth sharing here. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Quite interesting opinion about the role of QE departments in the increasingly agile software development world, where much of the testing of the code itself is done by developers in unit-tests. Imagine a SW project, which is properly covered by unit tests, used as a commit gateway. Imagine static analysis tools are hooked into the version control system, and any bugs found are immediatelly reported and fixed. Would that lead to perfect software? What would QE dept do to justify its existence on such a project? Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
We are happy to announce the first release of GNATprove. This tool is used for formal verification of Ada programs and is being developed as part of the Hi-Lite project. We provide binary distributions for x86 linux, x86 windows and x86-64 bit linux. Via PolettiN Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
Dr.Dobb's article about deploying static analysis tools in the real world, and the implications it has on your development (and business) process. Quite thorough, easy to read, and very detailed. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
So you think you are a C programmer? Test yourself using this quick quiz, put together by John Regehr (John Regehr is a Assoc. Prof. with University of Utah, and with his students, he worked on several bug hunting tools). The quiz focuses only on work with integers, which should be fairly easy, right? Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
It's a common anecdote that it's easier to create static analysis checkers over LLVM than over GCC. Yeah, right. But why? The article deals with Google cancelling some of their development on GCC and moving on to LLVM, and it hints on the precise reasons behind that anecdote. For even more gory details, do not overlook a link where Deseley Hutchins (Google developer) explains what difficulties did they have with GCC. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
...especionally when done in a smart way. This blog post briefly describes the smart process of fuzzing Chromium browser where everything tedious is automated: test case analysis and minimization, regression diagnosis and later verification. All done on thousands of running Chromiums at once. And I like the name: ClusterFuzz :) Delete the scoop?
Are you sure you want to delete this scoop?
Yes
Scoop.it!
As I mentioned yesterday, I see a lot of lightweight static analysis tools in the Ruby world. Brakeman is next of these. It has an integration with Jenkins, it focuses on security issues, and is specific for a popular Ruby on Rails framework for building web applications. Delete the scoop?
Are you sure you want to delete this scoop?
Yes
|
Again, it's older (for some reason, scoop.it gives me either old, or totally offtopic suggestions lately :/), but good. I think this is actually very good technique when you are a professional tester: have you ever reasoned about how would you test stuff you use or read about?