kwillets

19 ms·

181.
by kwillets 7y ago
The ACM had a nice article debunking the Wald story: http://www.ams.org/publicoutreach/feature-column/fc-2016-06
182.
by kwillets 7y ago
The drama on this is just getting out of hand. Let he who has not posted a picture on social media cast the first stone.
183.
by kwillets 7y ago
It was already pretty obvious with Echelon and so forth; IDK why people didn't get it before. I was disappointed in the NSA's security though.
184.
by kwillets 7y ago
There's another crane index, which is how many tower cranes you need to build one building. Google's is 6; I think they're among the world's least efficient builders.
185.
by kwillets 7y ago
And there I was hoping they would solve all the world's problems from their campus in Mountain View.
186.
by kwillets 7y ago
With SSE it's 2.6-3.4 cycles/byte to unpack UTF-8 into UTF-32 (ie integers). See https://github.com/KWillets/fastdecode-utf-8 . Selection (extracting the i-th element) is actually quite fast, but there'
187.
by kwillets 7y ago
McSherry's COST paper ( http://www.frankmcsherry.org/graph/scalability/cost/2015/01/... ) is a good reference on the advantages of SISD graph processing.
188.
by kwillets 7y ago
Has anyone figured out a way to manage these automatically? In many cases they're delayed commits; one branch of the code is retained, and other paths are deleted, including switching code (eg "if <switch> then A else B end
189.
by kwillets 7y ago
Azure storage has a visibility timeout of up to 7 days. https://docs.microsoft.com/en-us/rest/api/storageservices/pu... Like you I wish this functionality were first class; it seems like it will work, bu
190.
by kwillets 7y ago
Sadly, this was LookSmart back in 1998. It was an in-house edited directory, but I told the CEO we should build a pagerank that diffuses to unlisted sites as well. A pretty good web directory at the time was cheap, about $12 million a year
191.
by kwillets 7y ago
>Japanese made products used to be synonymous to 'cheap' and 'poor quality.' It was not until July 1950 when Dr. Edwards Deming (1900-1993) introduced SQC to the Japanese industry audience, using the trademark colored
192.
by kwillets 7y ago
>The next task coming up is that they want users to be able to ad-hoc schedule an email to themselves with the fancy graphs attached as a pdf, so we've got that to look forward to. This is the endgame of every OLAP tool. But execs
193.
by kwillets 7y ago
Ha, SQL hacks as patents.
194.
by kwillets 7y ago
How close is real time? Seconds? Minutes? Also, what type of aggregate?
195.
by kwillets 7y ago
MapR has been reduced. Hadoop was the floppy disk of big data. Ubiquitous, but always beat by other solutions.
196.
by kwillets 7y ago
Most OLAP has that exponential falloff in data use with time. I once extracted all the date strings from recent queries on a data warehouse and found the same distribution. Zynga once built a kind of time-series database with very similar
197.
by kwillets 7y ago
It's worked out into a particular kind of doublespeak where "liberal" officials cite a litany of reasons which always add up to NIMBY: neighborhood character, "affordable housing" as a tax on homebuyers, "gree
198.
by kwillets 7y ago
This must be censored to protect the public from antisocial thoughts.
199.
by kwillets 7y ago
Wow, that article is great for people who like a lecture with their news. Crime doesn't exist, apparently, and we don't have to do anything about it. News: Ring is aggregating local crime content captured on their doorbell camer
200.
by kwillets 7y ago
Mode missed the boat by not including ETL capabilities. I worked for a time on the tool that Mode was based on, and its ETL capability was the hidden hand that got Data Scientists to build and maintain the data pipeline.
201.
by kwillets 7y ago
99% perspiration, 1% inspiration, and the rest is ML.
202.
by kwillets 8y ago
We did a migration from Vertica to ADW a couple of years ago. ADW was slower and about 2x more expensive than Vertica on 8-year-old hardware, and it hit its node limit at around 60. Loading was also a PITA; there was no json support.
203.
by kwillets 8y ago
Micro has always seemed like the weak point of this game; it relies a lot on making the UI clunky for humans.
204.
by kwillets 8y ago
Prefix sum was my first thought as well. One approach I haven't benchmarked is to vpmulllq (64-bit in-lane multiply) by 0x0101010101010101. That produces an 8-byte prefix sum in each lane, so then you need to prefix-sum the high bytes
205.
by kwillets 8y ago
A colleague once told me an Autonomy rep was sleeping with the client he worked for previously. It was all based on hooking the primary approver and covering up the actual usage of the "product".
206.
by kwillets 8y ago
It's about time. Autonomy was like the Donald Trump of tech companies.
207.
by kwillets 8y ago
I did a lot of troubleshooting on a system like this a year or two ago, and most of it came down to making sure that global state transitions are atomic, and making communications as robust as possible. We had the basics of execute-once usi
208.
by kwillets 8y ago
SF has the same thing, except the cops are off-duty (moonlighting). That's why it's safe at the Getty mansion even while the rest of the city goes to hell.
209.
by kwillets 8y ago
Often per-click stuff ends up with dozens or hundreds of data points from different parts of the code -- heartbeats, feature usage, funnels, experiment entry, etc.
210.
by kwillets 8y ago
You need a graph database to track all the levels of funding.

More ›