teh

4 ms·

1.
by teh 2mo ago
benbye - do you have any intuition why the model only needs 4 frames to work? I played with LeWorldModel and I am mystified why the latent prediction frame number is so small. It's an almost Markov like property where a short history i
2.
by teh 3mo ago
I sometimes wonder how much of intelligence is being good with tools. I feel pretty averagely smart but give me some good tooling like a good editor, a good type system, semantic grep, good testing and some solvers and I can actually delive
3.
by teh 6mo ago
Another useful "Emergency exit" is CTRL+Z which stops the process and cannot be intercepted. It's often faster than hitting CTRL+C and waiting for process cleanup, especially when many resources are used. Then you can do e.g.
4.
by teh 6mo ago
Chrome uses sandboxing and a lot of safe tooling (wuffs, rust) for untrusted data. curl is heavily fuzzed and you still mostly control what you are downloading unless the target is compromised. With logs the attacker controls what goes into
5.
by teh 6mo ago
Super useful tool but need to be aware that this is reading potentially untrusted input (e.g. in the case of http request logs) and written in c++, so a possible attack vector. I use lnav where I trust the logs, but do wish a safe implement
6.
by teh 6mo ago
There is precedent: with type checkers like pyright you can opt into specific checks, or have a basic, standard, strict setting, each expanding the set of checks done. How would dependencies work in this schism? E.g. if serde starts using n
7.
by teh 8mo ago
I can't fully answer your question but I did once spend about a week porting plain internal configuration to cue, jsonnet, dhall and a few related tools (a few thousand lines). I was initially most excited by cue but the novelty fricti
8.
by teh 2y ago
I feel you (like many people) got burned by the steep learning curve. Empirically some pretty high powered companies use nix successfully. It's of course always difficult to know the counterfactual (would they have been fine with ubunt
9.
by teh 2y ago
I feel the same way. Excited to see another attempt. But it's a c++ engine so not something I would want to expose to the internet really.
10.
by teh 4y ago
I've looked into this but saw hugely variable throughput, sometimes as little as 20 MB / second. Even if full throughput I think s3 single key performance maxes out at ~130 MB / second. How did you get these huge s3 blobs in
11.
by teh 6y ago
This is a great book! After reading it I found it much harder to enjoy movies showing bad security though (such has heists, nuclear anything, ..). E.g. from the book I learned about the IAEA recommendations for safekeeping nuclear material
12.
by teh 7y ago
I just spent some time implementing a lazy VM: Note also the push/enter VS eval/apply implementation change in GHC described in [1]. [1] https://www.microsoft.com/en-us/research/publication/make-fa..
13.
Why someone is offering to buy your disabled private SSH key(twitter.com)
36 points by teh 7y ago|3 comments
14.
by teh 7y ago
I think this is a common misunderstanding. The images are in the public domain. Nothing stops Getty (or you, or anyone) from selling them, even though you can just use them for free. The value-add service that Getty offers is legal indemnif
15.
by teh 7y ago
More often than not recruiters (external and in-house) make up large numbers to get you to reply. Here is a verbatim quote from a mail I got last year: > For the right candidate year 1 comp will be up to £500k. After going through a thre
16.
by teh 7y ago
The original SES doesn't seem to do anything to prevent meltdown/spectre attacks [1] This version removed direct access "Date" [2] but I'm not sure I'd trust any code running in the same process space given how
17.
by teh 9y ago
I think this is the UK's implementation of EU PSD2 directive (e.g. [1]) so may not survive brexit. Looking forward to what'll come out of it though! [1] https://www.tsys.com/news-innovation/whats-new/Art
18.
by teh 10y ago
Hi, drop me a line at tom@wearewizards.io :)
19.
by teh 10y ago
We're mostly using Protolude at the moment. It has some nice properties, e.g. * "id" renamed to "identity" so you can use id in your own code * panic functions like "notImplemented" * a generic string conv
20.
by teh 10y ago
I do agree but I also remeber that Haskell is 27 years old. Newer Haskell-inspired languages like PureScript don't have a built-in list any more. There's a lot of old stuff in Haskell, e.g. String which is a list of char. We have
21.
by teh 10y ago
I've been joking with friends that my next job will be AWS efficiency guru. I've somewhat optimized our own use, but I think I could use similar, simple rules to get 20% out of a 500k / month budget. Give me what I save you i
22.
by teh 10y ago
I'm not sure you are arguing against any point the article made, rightly or wrongly?
23.
by teh 10y ago
That's the weirdest argument I have heard against Haskell so far. Re-factoring is the single best thing in Haskell IMO. 90% of cases are 1/ change type 2/ fix fall-out 3/ done.
24.
by teh 10y ago
I just migrated a complex extension and with Firefox 52 most features are there. It was a bit painful but the new API is much better IMO. Also, we now have one code base less to maintain.
25.
by teh 10y ago
That's just a list of the positives. I think it'd be useful to your argument to also provide some "real world" nuclear experience. E.g. * The tax payer in Germany is picking up most of the estimated 100+ billion Euro cle
26.
by teh 10y ago
That still leaks a number, let's say it's 0-5h and resolution is 1 minute, that's 60*5 == 300, so roughly 8 bits of entropy. A `onBatteryLow` callback might do the trick?
27.
by teh 10y ago
I've been running on Wayland for a year with only minor issues. Which distro are you using?
28.
by teh 10y ago
What kind of data do you have? Is this mostly text, or more like compressed time series?
29.
by teh 10y ago
t2 instances have credits that are replenished at a constant rate, and used up when you use the CPU - is this sustainable for more than 1h? That's ~25MB of data per second for an in-memory workload over 6 machines. I think they missing
30.
by teh 10y ago
We had a look at first commits from various projects a while ago [1] - I like how the first commits get cleaner the newer the projects are, presumably because version control became more ubiquitous. [1] https://blog.wearewizards.

More ›