wladimir

25 ms·

151.
by wladimir 14y ago
I agree that nothing solves all possible problems. But it's a huge jump from "there is no silver bullet" to "changing the language doesn't bring ANYTHING". Likely, it is quite a bit more subtle in practice. Preventing "obvious" errors (whic
152.
by wladimir 14y ago
You get downvotes because your reply contradicts itself. A safer language results in more mistakes being found at compilation time, so would result in less mistakes in the product, not more. Also I would expect having to specify annotations
153.
by wladimir 14y ago
Exactly. Internally, the data structure serialized as "LLVM bitcode" (or the textual equivalent) is a list of functions, each function represented by a Control Flow Graph of basic blocks. Each basic block contains runs of SSA (static singl
154.
by wladimir 14y ago
BTW I looked around a bit and found another similar system, based on QEMU instead of Valgrind, claiming to be faster, ie less runtime overhead while tracing, and also supporting kernel mode: http://wiki.nss.cs.ubc.ca/Tralfamadore Edit: ne
155.
by wladimir 14y ago
"The Chronicle framework provides complete register and memory-level recording of Linux processes and efficient queries over the recorded traces." Wow, I was looking for something exactly like this a while ago, but was unable to find
156.
by wladimir 14y ago
If you can formulate your problem in terms of tensor math (for example, neural networks) you can use Theano [1]. It is a very high-level approach. You give it mathematical expressions and it generates and executes GPU code for you. When I l
157.
by wladimir 14y ago
My point is that you're focusing on low level kernels only, which obviously need to be simple and highly optimized. However, the number of people actually writing lowlevel HPC code (the super-optimized number crunching inner loops, usually
158.
by wladimir 14y ago
With "in the future" you mean "currently, in a restricted subset of C that runs on massive numbers of very simple cores". Massive parallelism is also possible with somewhat less simple cores; CUDA and OpenCL also started from the "bare sub
159.
by wladimir 14y ago
Having some experience porting software to embedded architectures, I can tell you that cross compilation can be a hell in autotools projects just as well. In theory it's easy, but all the projects handle it slightly different, or have hacke
160.
by wladimir 14y ago
"LIKE_TRANSACTION", "TWEET_TRANSACTION", on the lowlevel IBinder RPC interface in Android also made me laugh a bit: http://developer.android.com/reference/android/os/IBinder.ht... . "There is no response returned and nothing about the sys
161.
by wladimir 14y ago
Indeed, I read that book last year and it felt really relevant, Vonnegut sort of nailed it with the second/third industrial revolution and the kind of social problems it would cause.
162.
by wladimir 14y ago
Those arrows are indeed pesky. In my fixed-width fonts (at least at the font size I'm using), → (->) and ⇒ (=>) look the same.
163.
by wladimir 14y ago
Having only two is great for developers, though, at least as long as native apps are preferred by users. There's a limit to the number of different platforms one can support, especially if they have their preferred programming languages lik
164.
by wladimir 14y ago
What I found an interesting development in syntax highlighting is "semantic highlighting". Some IDEs support this by integrating directly with a compiler (such as clang). At a primitive level it allows, for example, coloring nams of variabl
165.
by wladimir 14y ago
I wouldn't recommend using that one. There is a newer plugin [1] that uses "after" in combination with conceal syntax to achieve the same effect but without buffer hacks on load/save. [1] https://github.com/vim-scripts/Haskell-Conceal
166.
by wladimir 14y ago
Perhaps putting a social stigma on the people that do 'cooporate with the war machine' could avoid that, similar to that on people that work even though a certain group is on strike. I'm pretty pessimistic as well, but I don't think it's im
167.
by wladimir 14y ago
Yes, I think the underlying reason that drones are brought into these kind of discussions is because people are terrified of armed drones, especially autonomous ones. They make war cheap in the same way as CCTVs make surveillance cheap. But
168.
by wladimir 14y ago
He's not trying to illustrate what HN readers think, but the mindset of people that try to use technology for purposes of control. Either by having programmers build software for surveillance tech, or by making them build backdoors into oth
169.
by wladimir 14y ago
I kind of agree. Personally I have no problems with drones being used where police helicopters are now. It's cheaper, safer, and probably more effective too. However , and this is a big issue, drones are thus cheap making it possible to us
170.
by wladimir 14y ago
Well, dropping to the interactive prompt is a statement or two in Python. You could do that when something goes wrong (for example, in the inner, or outer exception handler of your application). There's also tools for Python where you can b
171.
by wladimir 14y ago
I think that if you have a good grasp of the commands in an editor such as vim, having "real-time" feedback isn't as important as with other UI paradigms (as long as it's not too slow, you do need a few fps). Though I'm not sure. It will ta
172.
by wladimir 14y ago
Well not really. I had a vague sense that there are extensions that add typing, but not more. My experiences with Lisp are mostly limited to SICP. Thanks for the link.
173.
by wladimir 14y ago
"primitive" is an awkwardly chosen word with negative connotations... I suppose it is in the sense that it doesn't do type inference (or reasoning about types at all, for that matter). Like a Lisp it simply executes what you write without a
174.
by wladimir 14y ago
Or filled with get rich quick scams. That reminds me of a Charles Stross novel, I think Accelerando.
175.
by wladimir 14y ago
A big part of the problem is trying to do this with C in the first place. The explicit memory operations create an extreme complexity in reasoning about a program exhaustively, not far from analyzing generic binary code. C is so permissive
176.
by wladimir 14y ago
It seems that Phoronix is going to new depths here in sensationalizing titles. They don't mention any reason why it would be crap as a driver. Reasons it'd be 'crap' is if it is much slower than the binary one, or buggy, or lacks features.
177.
by wladimir 14y ago
I have a few Mali devices next up on the chopping block after I finish reversing the VideoCore IV Hehe. Reversing a GPU is a lot of work . From my experience it usually involves zillions of commands, state bits that subtly change handlin
178.
by wladimir 14y ago
I certainly know about Nouveau. I even contributed to it, albeit indirectly (I'm the guy that wrote decuda). But reverse engineering GPUs is a thankless process, which has to be repeated for every new release of the hardware (in a certain s
179.
by wladimir 14y ago
AFAIK everything that runs on the ARM is now open source. The firmware of the GPU itself could still be a blob. That's very cool, I hope NVidia follows suit wishful thinking .
180.
by wladimir 14y ago
Ok, I misunderstood then. I thought it would subtly return no value instead of a value, and you'd only discover this at run time. That would be bug prone. If it simply generates a compilation error, no problem.

More ›