ntoshev

42 ms·

331.
by ntoshev 16y ago
Is anyone interested in working in this space? If so please shoot me an email. I have implemented a prototype of a service for sharing physical stuff. - it's a Facebook app to utilize the existing trust network - game elements (gaining karm
332.
by ntoshev 16y ago
I agree you have to make it work with the real and not ideal kernel :) Another thing PHK misses is that single-threaded approach solves the database concurrency problems for free, and people start to use it more - e.g. Stonebraker in his Vo
333.
by ntoshev 16y ago
Be sure to check PHK's comment (#5) madvise + mincore syscalls could be used to get the kernel to preload the pages asynchronously.
334.
by ntoshev 16y ago
I upvoted you, but the general problem with these links is that they affect author motivation, whether or not they realize it. This is especially true for subjective listings such as this one, they are simply a brain dump of whatever the li
335.
by ntoshev 16y ago
Python has it's share of warts. Its automatic memory management is not a modern GC, it involves reference counters that are slow, take up memory and don't play well with copy-on-write semantics of modern virtual memory managers. Its functio
336.
by ntoshev 16y ago
Now I feel less guilty for refusing to install Moonlight on my Ubuntu laptop and downloading Feynman lectures from isohunt torrents instead.
337.
by ntoshev 16y ago
The way Facebook's Graph API is structured can be applied to other datasets containing interconnected objects. It's like a REST service serving JSON, with a specific way to construct the URLs. I like it a lot. Here is a blog post explainin
338.
by ntoshev 16y ago
> strongly recommend against building your entire business around a single platform I see this statement a lot and I never understand it. If you're Zynga, should you build Farmville for OpenSocial and as an independent site too? The cos
339.
by ntoshev 16y ago
This is just spelling / grammar, I have no problems with these errors, especially when the main documentation is wiki-based (this may be because I'm not a native speaker). In my limited experience with Facebook docs before the Open Graph AP
340.
by ntoshev 16y ago
This is Bret Taylor's Graph API - the concept simplified the API a lot, and companies like Freebase copied it for their own data. Too bad it is not working...
341.
by ntoshev 16y ago
LUI = "Linguistic" UI, including CLI You're right it can be easier once you are familiar with it, but there's nothing to suggest how/if "up" works if you are not.
342.
by ntoshev 16y ago
Graphic UIs show you what actions are available right there, linguistic UIs rely on you to think up a valid expression in that program's particular language to do what you need. GUIs will always be easier for simpler stuff, but less expres
343.
by ntoshev 16y ago
I don't use Yahoo/Bing straight up, e.g. I re-rank, omit, etc. All re-ranking you can do is very limited with these services because they are black boxes: you don't see what factors went into ranking a page the way it happened, you can't
344.
by ntoshev 16y ago
So does he mean you shouldn't negotiate opportunities? I don't think this is valid in general.
345.
by ntoshev 16y ago
For a quote out of context, by "that information" he could have easily meant "the real-world social connections between people", which Google could probably infer from Gmail. Such articles are meaningless.
346.
by ntoshev 16y ago
You should really automate this feedback process by mining the web for examples where people complain about Google quality AND they mention the search queries themselves :-)
347.
by ntoshev 16y ago
What do you think about granularity of filtering: do not take all your customers' servers down, just filter the offending pages/services at network level until they sort it out.
348.
by ntoshev 16y ago
Is this better at anything than http://scholar.google.com ? I find it inferior: it's slower and I can't filter by time. Search results seem similar.
349.
by ntoshev 16y ago
Feynman's "Messenger lectures" @ Cornell begin with the example of the law of gravitation, and it's a really excellent lecture (for you, not for your son). I think your son doesn't need answers to the questions you ask here. He wants a simp
350.
by ntoshev 16y ago
Don't worry about the lines of code you write, I'm pretty sure even Dr Norvig can't write 50 of these per day: http://norvig.com/spell-correct.html
351.
by ntoshev 16y ago
It's not in the other half of the genome. All cells in an organism share the same DNA, but different cells develop in a very differentiated manner depending on the cells around them (e.g. should it be a neuron or a connecting tissue?). The
352.
by ntoshev 16y ago
> the decompression algorithm, i.e. the route from the genome to brain, is insanely complex, but it doesn't say anything about the actual complexity of the brain This is where both Kurzweil and you are wrong. A very complex compressio
353.
by ntoshev 16y ago
This is a good benchmark, but not perfect, since people can only upvote what they already see. Try bringing some schizophrenia to the site: show the "classic" version to the "classic" crowd by default for some time and see if the two versio
354.
by ntoshev 16y ago
I'm the opposite, e.g. I find Django templates too restrictive. I have used Rails, but my web framework of choice is Tornado; I really like their design choices.
355.
by ntoshev 16y ago
It would be more interesting to see this map corrected for land mass to see actual density.
356.
by ntoshev 16y ago
Generators actually allow you to decouple code in more ways than normal functions and classes let you. See the "Why functional programming matters" paper, or the more practical presentation here: http://www.dabeaz.com/generators/Generators
357.
by ntoshev 16y ago
On Unix I did this: cat /dev/urandom > /dev/dsp There is a free iPhone app too.
358.
by ntoshev 16y ago
Are there hedge funds or investment banks who treat quants/software engineers equally to the traders?
359.
Large Language Models in Machine Translation (Google's way of doing it)(acl.ldc.upenn.edu)
1 points by ntoshev 16y ago|0 comments
360.
by ntoshev 16y ago
I think it was more of a regret that his intuition about the problem area gets invalidated so quickly. The top technical guys are supposed to have good intuitions about what works and what doesn't; this is a major part of the value they can

More ›