HackerTrans
TopNewTrendsCommentsPastAskShowJobs

arnarbi

1,538 karmajoined 14 jaar geleden
Icelandic. SWE @ Google. Security, authentication, authorization. Work on WebAuthn, FIDO & passkeys.

[ my public key: https://keybase.io/arnar; my proof: https://keybase.io/arnar/sigs/lf4Zbpd04PZPXu_eZyGhPVoef2BuJ0p5VUuzV8ZXD4I ]

comments

arnarbi
·16 uur geleden·discuss
Restaurant owners interviewed in the media here in SF are directly quoted saying they can’t do that because “customers would notice”, or think “oh that’s expensive, I can’t eat out twice a week”.

These are arguments for including the fees that make the customer __still pay the same higher price__, implying that the whole point is that they won’t notice. And reporters don’t seem to even register the absurdity of those remarks or question them in any way.

https://www.sfgate.com/food/article/sf-restaurants-junk-fees...

https://www.kqed.org/news/11992412/californias-junk-fee-ban-...
arnarbi
·2 maanden geleden·discuss
Scorched Earth was also my first hacking target. Found out that your cash balance and weapons inventory was all stored in a mysterious .ini file and you could just edit it.
arnarbi
·2 maanden geleden·discuss
Best "friendmaker" hobby I know is sailing.

It's in general a very newcomer friendly hobby, which is both important as a newcomer yourself as well as for meeting new people once you are into it. It's naturally collaborative so you have to communicate, not very intense so there's a good amount of chill time, and in the cases where small-talk doesn't turn up interesting topics you can always talk about sailing itself.
arnarbi
·3 maanden geleden·discuss
It's more like workers on a large oil tanker using bicycles to move around it, rather than trying to use another oil tanker.
arnarbi
·5 maanden geleden·discuss
> why they're attributed to AI?

I don’t think they mean scrapers necessarily driven by LLMs, but scrapers collecting data to train LLMs.
arnarbi
·9 maanden geleden·discuss
We did not, no. Just wrote up the report and moved on.
arnarbi
·9 maanden geleden·discuss
Back in college (~2008) we implemented this with a 7 foot tall back-projected screen and a couple of Wii remotes after seeing Johnny Lee’s video. The nice thing with that screen was that you could stand so close to it you couldn’t really see the edges.

We had as many people come test as we could, and we found that 90% of them didn’t get a sense of depth, likely because it lacked stereo-vision cues. It only worked for folks with some form of monocular vision, incl myself, who were used to relying primarily on other cues like parallax.
arnarbi
·9 jaar geleden·discuss
It's the number of processes either running or waiting in the queue for CPU time. This is averaged over 5, 10, and 15 minutes.

Load of 0 means idle (nothing to run), load of 1 means e.g. one process running and nothing in the queue.

https://en.wikipedia.org/wiki/Load_(computing)
arnarbi
·14 jaar geleden·discuss
> So you agree it wouldn't work for specs to be "written as executable reference implementations in a well defined programming language"?

No, then I would be disagreeing with myself. :)

> But sometimes you want to specify performance characteristics.

For a Markdown processor, hardly. But...

> ... cannot tell you what aspects of its behavior are "specified" and which are just implementation details.

That is a fair point indeed. I was merely disagreeing with the linked article in that a spec should be in a natural language, which I still hold to. Such specs generally have problems with under specifying, leaving too much up to implementation, which often ends with a few main implementations sort of becoming the actual practical spec. Look at browsers over the history for many examples. Your point seems to be that reference implementations might overspecify instead. This is true, but is easier to compensate for IMO. For example with clear types, natural language parts or even by conformance test suites being part of the specification.
arnarbi
·14 jaar geleden·discuss
Of course not. Such a spec would state what are its inputs and outputs, either as natural language descriptions or as types (say, if the reference language supports it).

Everything outside that (e.g. timing) would be fair game to optimize. Moreover, as I hinted at, a spec should not be optimized for speed or efficiency, but rather readability and unambiguity.
arnarbi
·14 jaar geleden·discuss
I do not agree that a spec has to be written in English. Like most natural languages, it is very easy to be ambiguous and to underspecify. Resolving this can make English just as human un-readable as anything else.

Specs like this should be written as executable reference implementations in a well defined programming language. This can very well be human readable, and should be done without regard for efficient execution. It's less ambiguous, amenable to automated conformance testing, and is easier to evolve than a natural language document.