cannam

8 ms·

31.
by cannam 4y ago
I idly wonder how these compare to the arbitrary-precision implementations I worked on for REDUCE, exactly 30 years ago as it happens ( https://github.com/reduce-algebra/reduce-algebra/blob/master... ) - in its
32.
by cannam 4y ago
Yeah, I like Chomp. As I recall, I first installed it because it had an optional emoji pack with the old blob emojis in it, which I was miffed to have had taken away from me in an Android update.
33.
by cannam 4y ago
> From my experience I can say that the “old IBM machines were reliable and well constructed” is not true. They were bulky and over time, everything breaks. Too right. I know the T40 series Thinkpads quite well (the 2004-5 era) and they&
34.
by cannam 4y ago
> Nonsense it's as old as the tech industry, and the defence industry it grew out of. I was curious about this, because I also think of the staff engineer as a newish thing, and I don't believe I've ever actually met one.
35.
by cannam 4y ago
It can also output to C, similarly. But vectorisable code isn't recognisable as such in the output. I'm pretty sure the compiler that handles MLton's output never sees anything to vectorise.
36.
by cannam 4y ago
> What's the FFI overhead like? I don't have measurements to hand, but vaguely speaking, low. Low enough to be worth using for even quite modest calls into vector functions or optimised libraries. > And can you pass SML func
37.
by cannam 4y ago
> By the way, wasn't there a "ML for the Working Programmer" aeons ago? There was. It was written by the author of this very submission and you can download it free from his site (follow the link trail from author's n
38.
by cannam 4y ago
MLton can't do SIMD auto-vectorisation, which unfortunately gives it a ceiling significantly below C++ compilers for the sort of thing I do. I generally use FFI calls for chunky, frequently-used vector stuff. It's a good compiler
39.
by cannam 4y ago
I also use .mlb files exclusively and agree with your general point - but I do make use of Poly/ML and occasionally SML/NJ against the same sources, using scripts to drive them from .mlb files: https://github.com/c
40.
by cannam 4y ago
I didn't know about Realtime Trains ( https://www.realtimetrains.co.uk/ ). I'm definitely going to be using that! It shows a lot of technical detail, but the exciting bit is that it has the platform a train is booke
41.
by cannam 4y ago
> GIMP development started in 1995. The GNU Objective C frontend (the one that has stayed in use) started in 1993. I would hardly consider the GIMP developers (who became the early developers of GTK) to be rash in their option not to ado
42.
by cannam 4y ago
This is a little odd. There are at least three top-level comments here wondering why anything should be different now from how it was "back in the day", as if that were what the article is about. But the article doesn't appea
43.
by cannam 4y ago
A good point. I just checked - the CRT draws about 120W and the 40" LCD about 70W. The latter is more than I expected, given that it's not particularly big by modern standards - though it is a few years old itself by now.
44.
by cannam 4y ago
I can't imagine loving a monitor the way I loved my iiyama Vision Master Pro. 1600x1200 at 71Hz driven from a Matrox Millennium (4MB VRAM and no 3D acceleration whatsoever) with soft-edged pixels and no visible grain or mesh. It was by
45.
by cannam 4y ago
> everyone considers me a deviant for just wanting to start my day with cereal and milk My ideas about breakfast were turned on their head many years ago when a German friend (his typical breakfast: rye bread, cheese, cured ham) remarked
46.
by cannam 4y ago
Miyazaki is a fascinating person to read about, because of his combination of luminous creativity, clarity, and lightness with plain ordinary hard work, grumpiness, and pessimism. His two books of essay-ish fragments, Starting Point and T
47.
by cannam 4y ago
Perhaps also worth adding that the linked article (the one titled "Citation Needed") itself originally had hundreds of comments beneath it, many quite argumentative, which seem to be now missing - I guess lost in a site update. It
48.
by cannam 4y ago
I agree with the content of this article, but all the same my advice is usually the opposite - for reasons that its author sets out rather well in the table in his follow-up article ( https://blog.robertelder.org/signed-or-un
49.
by cannam 4y ago
Yeah, I imagine as long as Apple Music runs on Android, so will Shazam. It's a gateway.
50.
by cannam 4y ago
The Shazam algorithm was genuinely novel and invented by one of their founders. They patented it to great effect.
51.
by cannam 4y ago
That's interesting - I had a vague recollection of having heard of them before launch - I guess they were hiring from the pool of developers being laid off from the dotcom bust? I have an image in my mind of my boss at the time going a
52.
by cannam 4y ago
It's obviously a cracking algorithm, but what made Shazam doubly remarkable was how efficiently they turned it into a working product. It wasn't just a case of developing an algorithm that could in theory be used to match an aud
53.
by cannam 4y ago
Quite right! But approximate experiments and lightweight automation are really useful in deciding where to go and then making sure you stay there. I'm all for test-first, but I'd find it very hard to argue that it's a more im
54.
by cannam 4y ago
Well! This is interesting. I like the write-up. The acknowledgements include "Project Savanna, a group who shares my desire in improving the tooling around SML", with a link to a Slack page that asks for a login. Can you say anyth
55.
by cannam 4y ago
This is a good article, with (for me anyway) quite a twist at the end. The author quotes a tweet expressing amazement that any company might not use TDD, 20 years after it was first popularised - and then writes "I’d equate it to shell
56.
by cannam 4y ago
As an ignorant person who occasionally repairs things, something I have often wanted is a handwavy guide to failure modes of components expressed in terms of their audible effect. What I mean is something that would help me to take a proble
57.
by cannam 4y ago
I think you're right to distinguish between kinds of tiredness, rather than levels. The article, good and thoughtful as it is, suggests that a dead-tired state is somehow just more tired than normally tired. But there is also a tiredne
58.
by cannam 4y ago
> There is really no need to get rid of all powered individual transport. That is basically an ideological position, not a scientific one. Nah. I used to hope that moving propulsion to the electric grid would allow us to use renewables t
59.
by cannam 4y ago
I'm not from Germany, but I just came back (to the UK) from a holiday spent partly in Germany and partly in Denmark. In Germany I paid cash for every casual snack-grade transaction, and some bigger ones like restaurant meals. It seemed
60.
by cannam 4y ago
Totally with you there. "Useless use of cat" is usually good practice, not bad. It's clearer - the structure indicates right at the front what it is going to do, namely read a file and pass it through a pipeline. There is no

More ›