Twitter’s new massive-scale twitter search backend. Sharding galore
‘I went over to Japan right around the time Takeshi was deciding which emoji were going to make it into the first cut of Gmail emoji. The [PILE_OF_POO emoji] was absolutely one of the necessary emoji that Takeshi said we have to have. There was actually conflict because there were people back at headquarters who had no idea what emoji were, and thought that having an animated [turd] in their Gmail was offensive.’ ‘[The poop emoji] got very popular when a comic called “Dr. Slump” was broadcast in Japan back to the ‘90s. Such poop was not an object to be disliked, but it had a funny meaning. This was a very popular comedy animation where a girl played a trick on other people using the poop. The poop was this funny object to play with. It was never serious.’ ‘In Japanese that’s called “unchi.” It’s a child word with a benign meaning. ‘
We plan to send an unmanned robotic landing module to the South Pole of the Moon – an area unexplored by previous missions. We’re going to use pioneering technology to drill down to a depth of at least 20m – 10 times deeper than has ever been drilled before – and potentially as deep as 100m. By doing this, we will access lunar rock dating back up to 4.5 billion years to discover the geological composition of the Moon, the ancient relationship it shares with our planet and the effects of asteroid bombardment. Ultimately, the project will improve scientific understanding of the early solar system, the formation of our planet and the Moon, and the conditions that initiated life on Earth.Kickstarter-funded — UKP 600k goal. Just in time for xmas!
Unlike the (excellent) Typescript, it’ll infer types:
An extremely good explanation from Marc Brooker that exactly-once delivery in a distributed system is very hard.
And so on. There’s always a place to slot in one more turtle. The bad news is that I’m not aware of a nice solution to the general problem for all side effects, and I suspect that no such solution exists. On the bright side, there are some very nice solutions that work really well in practice. The simplest is idempotence. This is a very simple idea: we make the tasks have the same effect no matter how many times they are executed.
This topographic map represents Ireland. It is designed for “hillwalking”. The contour lines are extracted from SRTM public data provided by NASA. These files contain a digitized ground represented by points. The sample rate defines a grid resolution for Ireland around 90m in northing and 60m in easting. In major cases, digitized points do not correspond with summits. Carrauntoohil (1039m, the highest summit of Ireland) does not appear in SRTM data. The altitude reaches only 1018m. Data were obtain from space with a radar. Because of the relative position between the radar and the earth, a shadow appears in some conditions (along ridges, behind summits…). This shadow matches with a gap in data (Imagine you with a flashlight in a dark room. It is hard to see what is in shadow). To close these gaps, you need other data or you can do interpolation. The second solution is applied in our case. There is one square degree per SRTM file with a sample rate of 1200×1200 points/square degree at Ireland latitude. […] All in all you obtain contour lines pretty sufficient for walking.
Marcus Ramberg says: ‘If you have a chromecast and you’re not using castnow, I don’t know what is wrong with you.’
John Allspaw with an interesting assertion that we need to ask “how”, not “why” in five-whys postmortems:
“Why?” is the wrong question. In order to learn (which should be the goal of any retrospective or post-hoc investigation) you want multiple and diverse perspectives. You get these by asking people for their own narratives. Effectively, you’re asking “how?“ Asking “why?” too easily gets you to an answer to the question “who?” (which in almost every case is irrelevant) or “takes you to the ‘mysterious’ incentives and motivations people bring into the workplace.” Asking “how?” gets you to describe (at least some) of the conditions that allowed an event to take place, and provides rich operational data.
hooray, no more uberjars or monster commandlines!
‘From 19 Nov, 2014 00:52 to 05:50 UTC a subset of customers using Storage, Virtual Machines, SQL Geo-Restore, SQL Import/export, Websites, Azure Search, Azure Cache, Management Portal, Service Bus, Event Hubs, Visual Studio, Machine Learning, HDInsights, Automation, Virtual Network, Stream Analytics, Active Directory, StorSimple and Azure Backup Services in West US and West Europe experienced connectivity issues. This incident has now been mitigated.’ There was knock-on impact until 11:00 UTC (storage in N Europe), 11:45 UTC (websites, West Europe), and 09:15 UTC (storage, West Europe), from the looks of things. Should be an interesting postmortem.