regehr

11 ms·

61.
by regehr 14y ago
Exactly. I've seen plenty of overflow detection libraries that do precisely this. The undefined overflow plus the lack of access to CPU flags makes a pretty good argument against doing this stuff at the C/C++ level.
62.
by regehr 14y ago
Awesome, I've added a link to the bottom of the post.
63.
by regehr 14y ago
To give a few more details, an inversion is where there's a layer of air where it gets warmer instead of colder as you go up. For example, last night at the SLC airport it was 4 degrees F whereas at the top of one of the 11,000 foot peaks n
64.
by regehr 14y ago
A professor in atmospheric sciences at the U of Utah has a great blog and has been writing a lot about the inversions lately: http://wasatchweatherweenies.blogspot.com/ I live in SLC. The air is indeed pretty ugly right now. On the othe
65.
by regehr 14y ago
"But the airport officials were scared. They didn't know whether the destruction of the launch loop was an isolated incident of terrorist sabotage, or maybe the beginnings of a revolution-no one seemed to think, ever, that it might have bee
66.
by regehr 14y ago
Reminds me of this old obfuscated C contest entry: http://www.cise.ufl.edu/~manuel/obfuscate/pi.c
67.
by regehr 14y ago
The right way to do it: 1. Present data showing there is a problem. 2. Propose fixes. But that's harder than doing it the wrong way: 1. Screed. 2. Propose fixes.
68.
by regehr 14y ago
When these show up on Earth they're often called "moqui marbles." Lots of them are found in Utah and other parts of the American southwest: http://blog.regehr.org/archives/703 In the top pictures you can see them embedded in the canyon w
69.
by regehr 14y ago
Speaking as the instructor for a class that is using raspberry pis: at this point the most important thing about the raspberry pi is the community and mindshare. It's going to be hard for a new board, however awesome, to replicate this.
70.
by regehr 14y ago
My university (Utah) does not allow this.
71.
by regehr 14y ago
I expect that Wilson will enjoy your writeup, definitely forward it to him. He's a Googler you should get to know anyway...
72.
by regehr 14y ago
I'm a professor. If my student unilaterally posted something on arXiv, I would strangle him/her, at best.
73.
by regehr 14y ago
Read the steps again. They are pure sarcasm.
74.
by regehr 14y ago
Csmith is much less like a random walk; most program generation steps result in forward progress. The problem is that at the end of many forward steps, we may have a test case that is, in some sense, very similar to all of the other program
75.
by regehr 14y ago
This does indeed work. The problem is that there is a very large number of possible goals that a random tester might aim for, and we do not know ahead of time which ones are most important. Also, for complex testers like Csmith, it is not c
76.
by regehr 15y ago
This is all true and Blackboard probably isn't even the worst of it. Our grad admissions is done using ApplyYourself which is a giant turd. The front office uses Peoplesoft and nobody has ever said anything good about it. This disconnect
77.
by regehr 15y ago
Somebody watched _Synecdoche, New York_ too many times...
78.
by regehr 15y ago
I'm a CS professor and for the last several years have refused almost every review request from a non-ACM/IEEE journal. I'm also an associate editor of an ACM journal and often have a very hard time getting people to review submissions. My
79.
by regehr 15y ago
I always find "50% F&A" to be misleading. You add 50% to the grant budget to figure out how much money to ask for. Therefore, the actual amount of the grant that goes to overhead is more like 33%.
80.
by regehr 15y ago
It's a nice ISA but has some painful aspects. 8-bit registers aren't big enough for many purposes. Shift instructions are one bit position at a time. Explicit stack pointer manipulation isn't atomic. The best thing about AVR is avr-gcc.
81.
by regehr 15y ago
Of course book-learning and in-class learning needs to be supplemented with a large amount of self-directed learning. As a professor I'm sure Matt is perfectly well acquainted with the limitations of a 4-year degree program.
82.
by regehr 15y ago
During college I was on a road trip with friends and we stopped at a rest stop on the interstate. After a while a trucker came over carrying a tire iron and looking ready to use it. He was so strung out on uppers he had decided we were all
83.
by regehr 15y ago
Wouldn't a virtual ISA be against the interests of hardware designers who benefit from ISA lockin?
84.
by regehr 15y ago
Historically, compiler certification is a bad joke. It amounts to concerted testing at best, and wishful thinking in most cases. Few (remotely useful) compilers have been verified in any mathematical sense, and even then there is generall
85.
by regehr 15y ago
"Why Take a Compiler Course?" http://blog.regehr.org/archives/169
86.
by regehr 15y ago
It's an early version of the Matrix. Code representing werewolves and vampires will be reused in subsequent iterations.
87.
by regehr 15y ago
No, that niche is already occupied by Blackberry.
88.
by regehr 15y ago
I have a compiler like that. It turns any code into "return 0".
89.
by regehr 16y ago
I had a couple of Indian grad students in my office last Fall who were seriously distraught that they were getting an A- in one of my courses. They thought this would prevent them from getting a job. I tried to disabuse them.
90.
by regehr 16y ago
Excellent recent essay on this topic: http://jonathanherzog.com/node/139

More ›