bjwbell

For anything more than traffic tickets, predictable is not what I'd call it.

I've seen someone charged with kidnapping get it lowered to a misdemeanor (in that particular case it was appropriate).

I use Amazon Pantry and subscribe & save. And honestly it's nice but not a killer feature.

The trickiest thing is estimating what subscription frequency to use. I barely even think about optimizing for % discount.

Firefox, or maybe Gecko, has been on a downhill slide with end users for a long time. I've been hoping it'll get better for at least a couple years and it hasn't.

A good browser is hard, like really hard. It's in the same order of magnitude as writing an OS. To improve the user experience they need to hire guys like https://randomascii.wordpress.com (Bruce Dawson).

Chips today have thermal clock rate scaling. Thermal control must be explicitly disabled to not get clock rate scaling.

If you haven't read anandtech they have articles covering this, since clock rate scaling increases the difficulty of benchmarking.

Process isolation is easier with virtual memory. Gpus went through this transition recently or in the case of Linux very recently.

Yes, no qualification. I've been part of the hiring process several times. Unless you're exceptional, hiring is brutal for older devs.

I would worry about it and I'm only close to 30 but I'm self employed now.

I think it was Steve Sailer that first said preferences are important. In climbing it's commonly thought most people could climb 5.12 with intense training. But not many actually do because few enjoy training that hard.

I saw this first hand, I was naturally a much better climber than my friend but I didn't enjoy training or climbing as much. After a couple years he's putting up first ascents.

To give a shout out he's http://vividrea1ity.blogspot.com/

Brian Kernighan's recent lecture on language design, https://m.youtube.com/watch?v=Sg4U4r_AgJU, is an example (not sure if it was on hacker news). I'm working on SIMD in Go and don't want to make avoidable mistakes.

The newer (haswell or later) Intel gpus DO have ECC for the gpu caches. I guess marketing missed that one or engineering insisted.

Modern gpus (last 5 years) are optimized for GPGPU in addition to graphics. They also have preemptive multitasking on either a per batch or per workgroup basis.

Intel's latest gpu architecture has an embedded OS running on the gpu for scheduling command batches, I'm not sure what AMD and Nvidia do.

I still wouldn't write a general purpose OS for it.

Is there a good reason to disassemble the shader machine code instead of using what's needed from the open source Linux/Mesa GCN driver code?

1 out of 4 isn't that extreme, historically ~20% of individuals in Western Europe didn't have children.

A lot of people can and do get better. I can't comment on whether traditional treatment works for that or not. Personally I think treating the person is more helpful in the long run.

I'm curious, are there any studies that show antipsychotics not being harmful vs no medication in the long term (5-10yrs)?

I'm not sure it's comparable, the tool supply chain for one is very different. A Dutch company is the lithography tool supplier for both American and non American chip companies.

If Moore's law is over that could be less of an issue (old tooling will still be useful).

"Lisp" is easier on readers. If he said "s-expression language", I wouldn't immediately have a feel for the level of abstraction he was aiming for (webasm has an s-expression representation but is low level).

I agree with the rest of your comment.

Over time I've become fond of the phrase "show me the code" (obviously stolen from the film).

Not sure why the downvotes, I voted you up.

It comes down to performance per $ and quality. I'm not certain about AMD's perf/$ but Intel's Xeon Phi perf/$ is terrible.

AMD's software quality is not so great compared to Nvidia. Rich Geldreich's post, http://richg42.blogspot.com/2014/05/the-truth-on-opengl-driv..., is revealing.

I highly recommend Weka, http://www.cs.waikato.ac.nz/~ml/weka/. I used it for river flow forecasting (published a paper) and it was very good. We ran about 100 different models, various neural nets, SVM, logistic regression, multilinear regression, etc.

I spent about a month trying different tools and liked the understandability of Weka. Scikit-learn calls out to C for the heavy lifting, which is great but it's not exactly easy to read and understand the code.

Weird, I thought all dentists recommended sealing as a matter of course. Maybe I was just lucky.

Results don't scale linearly with effort. I don't know if it's worth it.

Yes, but that's not the common usage of the word. Or how most people understand it. With that usage you could say ivy league schools are NOT biased against Asians since Asian graduates aren't more successful than non-Asian ones, except nobody does.

A significant portion of our physiology is devoted just to breathing. When you've had sinus or lung problems it's very obvious or lived in polluted city e.g. New Delhi. Air is really really important.

It took me close to a year to recover from minor sinus issues.

Occam's razor strikes again or in the latimes case Occam's butter knife.

Just start contributing. There's no better way to learn than reading and understanding code. For me the most important item is a project that you personally connect with.

In my experience most projects are very kind to new contributors and someone will usually step up and gently guide. Quality code reviews have been most the effective mentorship I've had in Open Source and out.

I like to think programming is about being creative...but reality has stared at me too long. The large majority of it IS mundane and boring.

I just looked at the commit log for my current project, https://github.com/bjwbell/gensimd/commits/master (SIMD in Go), and ~90% of the commits are boring. I still feel very excited about it, perhaps I'm unbalanced.

You doubt university administrators sincerity ;)

I've noticed very few of these articles look at the recent past ('80s) when women were a significant percentage of programmers. I'm too young but I have the feeling working conditions were much better back then.

My guess it's the new CEO at Intel, Brian Krzanich, he was previously head of the manufacturing division. The previous CEO was from marketing.

I was thinking more about passing the reference test suites. Or in the case this case failing since GPUs won't all render identically.