simonbyrne

15 ms·

91.
by simonbyrne 8y ago
Not generally: most compilers make use of the SSE instruction set which doesn't have extended precision. Occasionally old versions of gcc on 32-bit machines will still use the old x87 instructions (which do have extended precision), bu
92.
by simonbyrne 8y ago
No kidding: this is a problem with signed integers as well: C allows signed magnitude, ones' complement, and two’s complement representations, but no mainstream compiler supports anything other than two's complement. > In short
93.
by simonbyrne 8y ago
Any details on how it works? There isn't a lot of technical info on the website: http://matibabu.thinkitlimited.com/
94.
by simonbyrne 8y ago
You technically don't even need to have your licence on you when driving (though you can be required to show it within 7 days).
95.
by simonbyrne 8y ago
Thanks! My comment was not meant as a criticism of your work (I still like the paper, and do appreciate that you went to the effort of putting the amended results online), but more of the CS conference publishing model. I look forward to re
96.
by simonbyrne 8y ago
It is really annoying that CS conference proceedings don’t have a mechanism for retraction or post-publication errata.
97.
by simonbyrne 8y ago
Linus has long railed against the aspects of the C standard, and how they're been interpreted by compiler writers [0]. I'm not much of a C programmer, but the C aliasing rules are incredibly confusing, and seem to be interpreted d
98.
by simonbyrne 8y ago
Well, only approximately: as a double it is actually 10000000000000000159028911097599180468360808563945281389781327557747838772170381060813469985856815104
99.
by simonbyrne 8y ago
I take it from your GitHub profile that you're an experienced C++ programmer, in which case you're probably correct. Personally, I still find C++ to be effectively a black box, meaning that if I want to understand it or make chang
100.
by simonbyrne 8y ago
Oh no, it's the 90's: we need classes, right now!
101.
by simonbyrne 8y ago
No, I don't believe integers ever promoted on overflow (or at least not since 2012). If an operation involves two different integer types, they do promote to the larger one (i.e. an Int64 + a BigInt will give a BigInt).
102.
by simonbyrne 8y ago
I'm pretty sure the other two were as well.
103.
by simonbyrne 8y ago
It's also worth noting that Julia's JIT isn't tracing: it does all its compilation before the code is run (unless it hits a path which hasn't been run before, or wasn't inlined, in which case it runs the compiler ag
104.
by simonbyrne 8y ago
I agree that would be great. The crude answer is: make it easier for a computer to figure out what will happen when you run the code. The example I usually use is allowing integers to overflow, instead of automatically promoting to arbitrar
105.
by simonbyrne 8y ago
An interesting exception is the energy sector, which used Yahoo Messenger until it was shutdown: https://www.reuters.com/article/us-oil-traders-yahoo-idUSKCN...
106.
by simonbyrne 8y ago
Interesting. He is also mentioned in some more critical pieces about sheep farming in the Lake District: https://www.theguardian.com/commentisfree/2017/jul/11/lake-d...
107.
by simonbyrne 8y ago
Interesting background here: https://github.com/tensorflow/swift/blob/master/docs/WhySwif...
108.
by simonbyrne 8y ago
Originally, FFTs were the only time I missed 0-based indexing in Julia. But the neat FFTViews.jl package ( https://github.com/JuliaArrays/FFTViews.jl ) addresses this, and even improves on it by letting you use periodic
109.
by simonbyrne 9y ago
> I think we somehow took a wrong path when OO got embraced as wholly as it did back in the 90s or so. I think the particular paradigm that was adopted by C++, Java, etc. was a bad one. Having now worked quite a lot in a language with mu
110.
by simonbyrne 9y ago
The author of this post is the creator of the Julia visualization package Luxor.jl ( https://github.com/JuliaGraphics/Luxor.jl ). One thing I like about it is the nice turtle graphics feature ( http://juliagrap
111.
by simonbyrne 9y ago
Looks similar, except they recenter as well.
112.
by simonbyrne 9y ago
Thanks for the link, that paper was a great read: clever idea, and very well written.
113.
by simonbyrne 9y ago
Ah, thanks.
114.
by simonbyrne 9y ago
Can someone tell me what is going on here: the release notes [ https://golang.org/doc/go1.10#language ] state that > A corner case involving shifts by untyped constants has been clarified, and as a result the compiler
115.
by simonbyrne 9y ago
The post is quite old: while the technical arguments certainly had merit at the time, they have largely been addressed (the exception is probably error handling, but his complaint there is more subjective, and I still don't think any l
116.
by simonbyrne 9y ago
There may not be "correct" people on either side. The world is a complicated place, full of kludges and compromises.
117.
by simonbyrne 9y ago
He did elsewhere suggest teaching calculus by Big O notation: http://www.ams.org/notices/199806/commentary.pdf I would be excited to see someone try that.
118.
by simonbyrne 9y ago
see Chesterton's fence https://en.wikipedia.org/wiki/G._K._Chesterton#Chesterton.27...
119.
by simonbyrne 9y ago
Only got to do it once, but the upper deck was nice: it felt like your own private plane. But as far as economy goes, the A380 wins, especially seat 48D.
120.
by simonbyrne 9y ago
I think the point is more to do with savings: in a non-deflationary environment, you have the incentive to invest your savings in potentially productive use. In a deflationary environment the incentive is simply to hide your money under you

More ›