draaglom

10 ms·

31.
by draaglom 6y ago
Original data here: https://httparchive.org/reports/loading-speed?start=earliest... The degree to which desktop load times are stable over 10 years is in itself interesting and deserves more curiosity than just saying
32.
by draaglom 6y ago
this is very cool! good job on "getting out of the way" + making it easy to enter data. How do you know who to text (who is on shift, that is, as shifts change)? Or do you just text everyone periodically?
33.
by draaglom 7y ago
People are speculating this is for e.g. when your coins are at an exchange; the currency hard forks; and the exchange does not (yet) implement the fork so you can't access the coins.
34.
by draaglom 7y ago
You're not far off, I'd analogise as "redis, but compiled into your app, with all the necessary bits to make that work like you'd hope (consistent hashing/replication), (native data structures)"
35.
by draaglom 9y ago
That's a good point - the distribution over time should affect the answer. I've heard this rule of thumb applied before to the same problem in a different way - rather than assuming you're somewhere in the middle of the time
36.
by draaglom 9y ago
I think the charitable reading is: In the _absence of information placing you somewhere specifically on the timeline_, assuming "I'm somewhere in the middle" works surprisingly well.
37.
by draaglom 9y ago
Bonus: you can use pg_amqp[1] to publish to RabbitMQ with transactional semantics 1: https://github.com/omniti-labs/pg_amqp
38.
by draaglom 9y ago
There's a good reason for this, and it's not (directly) to get you on the phone & up-sell you. These landing pages are optimised for conversion, which means they're targeted at the niche who are most likely to convert alr
39.
by draaglom 9y ago
This is really cool!!
40.
by draaglom 9y ago
I'm not saying it's not hard, but success is very much possible. E.g. Shopify handled $15.4 billion worth of transactions[1] in 2016, and that's just one platform. [1]: https://www.shopify.com/2016
41.
by draaglom 9y ago
Prime-time TV adverts apparently cost about $75k[1]; $45k for a 5-minute show doesn't seem so unreasonable in that context. [1]: http://www.nielsen.com/us/en/insights/news/2014/advertising-...
42.
by draaglom 10y ago
stock symbols: https://www.google.co.uk/finance?cid=13772865 https://www.google.co.uk/finance?cid=704805 https://www.google.co.uk/finance?cid=730771
43.
by draaglom 10y ago
1. It's not in HAProxy or nginx, at least 2. Random / round-robin are "good enough" in a lot of cases - most people aren't running at 70%+ utilization
44.
by draaglom 10y ago
Unix is almost 50 :)
45.
by draaglom 10y ago
From the looks of things, both are "fast enough". What I'd like to hear more about now is ease of use - I'm told that with Kafka you have to be quite careful around planning the number of topics / partitions you&#x2
46.
by draaglom 10y ago
Assuming here you mean something like a classic REST-alike "/events" endpoint which returns a bunch of stuff that's changed since the last time you requested it. In that case, as the number of events grows, the HTTP tran
47.
by draaglom 10y ago
I like what Shopify does here - because your app is tied to a partner account, they can email you saying "this payload has failed 20 times in succession". If it fails too many times then the webhook is uninstalled. Not to be snark
48.
by draaglom 10y ago
This is correct, and important. Webhook payloads need to be logically monotonic[0]; this probably means either: - having a lamport-clock timestamp for each payload so you can entirely discard older payloads in favour of new ones - a well de
49.
by draaglom 10y ago
This is what I've found: https://archive.org/details/whatcdcrawl&tab=about Can't seem to get the actual data though.
50.
by draaglom 10y ago
"More compelling than Yammer" does not strike me as a high bar.
51.
by draaglom 10y ago
It does say something about medical breakthroughs -- since they're happening all the time and not shifting this upper bound very much. This is not to say that a medical breakthrough which changes the limit is impossible, but it has to
52.
by draaglom 10y ago
The relevant section: DHS proposes to add the following question to ESTA and to Form I-94W: “Please enter information associated with your online presence—Provider/Platform—Social media identifier.” It will be an optional data field to
53.
by draaglom 11y ago
Intuitively, a calculator is not superhuman. A person who could do mental arithmetic as well as a calculator? That would seem to be superhuman. If you find out that they were using a calculator under the table the whole time, they're
54.
by draaglom 11y ago
I know "P.H." is "Public House". Not sure about the others.
55.
by draaglom 11y ago
Funny seeing the difference between then and now: https://news.ycombinator.com/item?id=4529484 Next time, let's try being this supportive before someone dies as well as after.
56.
by draaglom 11y ago
To "use something in anger" is a phrase, meaning to use something "for real" - in production, etc. rather than just to try it out.
57.
by draaglom 11y ago
As to your first point: I agree with you, both redistribution of wealth and regulating companies can be very good (even necessary) things. Acting on climate change is one of those necessary things. W.r.t the OP though: anecdotally the point
58.
by draaglom 11y ago
While the overall issue is not about redistribution in the traditional sense, the battle lines often are. On the one hand you have capitalist-optimists like Thiel who basically focus on increasing energy output from non-carbon sources (whic
59.
by draaglom 11y ago
I don't think you have to assume terrorists are stupid, good OPSEC is hard. Particularly re. phones / internet, it only takes one mistake to link a phone number used for nefarious activities with your identity. (I would expect the
60.
by draaglom 11y ago
I shared the same confusion before, but I do have some ideas: I think you could establish a communication protocol with an alien species with something like this: First, establish the meaning of a set of mathematical symbols by choosing rec

More ›